OceanofPDF.

com
About This eBook

ePUB is an open, industry-standard format for eBooks. However, support of

ePUB and its many features varies across reading devices and applications.
Use your device or app settings to customize the presentation to your liking.
Settings that you can customize often include font, font size, single or
double column, landscape or portrait mode, and figures that you can click
or tap to enlarge. For additional information about the settings and features
on your reading device or app, visit the device manufacturer’s Web site.

Many titles include programming code or configuration examples. To

optimize the presentation of these elements, view the eBook in single-
column, landscape mode and adjust the font size to the smallest setting. In
addition to presenting code and configurations in the reflowable text format,
we have included images of the code that mimic the presentation found in
the print book; therefore, where the reflowable format may compromise the
presentation of the code listing, you will see a “Click here to view code
image” link. Click the link to view the print-fidelity code image. To return
to the previous page viewed, click the Back button on your device or app.

OceanofPDF.com
Using Artificial Intelligence

ABSOLUTE

BEGINNER’S

GUIDE

Michael Miller

OceanofPDF.com
Using Artificial Intelligence Absolute Beginner’s Guide

Copyright © 2025 by Pearson Education, Inc.

Hoboken, New Jersey

All rights reserved. No part of this book shall be reproduced, stored in a

retrieval system, or transmitted by any means, electronic, mechanical,
photocopying, recording, or otherwise, without written permission from the
publisher. No patent liability is assumed with respect to the use of the
information contained herein. Although every precaution has been taken in
the preparation of this book, the publisher and author assume no
responsibility for errors or omissions. Nor is any liability assumed for
damages resulting from the use of the information contained herein.

ISBN-13: 978-0-13-535605-0

ISBN-10: 0-13-535605-9

Library of Congress Control Number: 2024947065

$PrintCode

Trademarks

All terms mentioned in this book that are known to be trademarks or service
marks have been appropriately capitalized. Pearson cannot attest to the
accuracy of this information. Use of a term in this book should not be
regarded as affecting the validity of any trademark or service mark.

AARP is a registered trademark.

No part of this publication may be reproduced, stored in a retrieval system,

or transmitted in any form or by any means, electronic, mechanical,
photocopying, recording, scanning, or otherwise, except as permitted under
Section 107 or 108 of the 1976 United States Copyright Act, without the
prior written permission of the Publisher. No patent liability is assumed
with respect to the use of the information contained herein.

Limit of Liability/Disclaimer of Warranty: While the publisher, AARP, and

the author have used their best efforts in preparing this book, they make no
representations or warranties with respect to the accuracy or completeness
of the contents of this book and specifically disclaim any implied warranties
of merchantability or fitness for a particular purpose. No warranty may be
created or extended by sales representatives or written sales materials. The
advice and strategies contained herein may not be suitable for your
situation. You should consult with a professional where appropriate. The
publisher, AARP, and the author shall not be liable for any loss of profit or
any other commercial damages, including but not limited to special,
incidental, consequential, or other damages. The fact that an organization or
website is referred to in this work as a citation and/or a potential source of
further information does not mean that the publisher, AARP, and the author
endorse the information the organization or website may provide or
recommendations it may make. Further, readers should be aware that
Internet websites listed in this work may have changed or disappeared
between when this work was written and when it is read.

Microsoft and/or its respective suppliers make no representations about the

suitability of the information contained in the documents and related
graphics published as part of the services for any purpose. All such
documents and related graphics are provided “as is” without warranty of
any kind. Microsoft and/or its respective suppliers hereby disclaim all
warranties and conditions with regard to this information, including all
warranties and conditions of merchantability, whether express, implied or
statutory, fitness for a particular purpose, title and non-infringement. In no
event shall Microsoft and/or its respective suppliers be liable for any
special, indirect or consequential damages or any damages whatsoever
resulting from loss of use, data or profits, whether in an action of contract,
negligence or other tortious action, arising out of or in connection with the
use or performance of information available from the services.

The documents and related graphics contained herein could include

technical inaccuracies or typographical errors. Changes are periodically
added to the information herein. Microsoft and/or its respective suppliers
may make improvements and/or changes in the product(s) and/or the
program(s) described herein at any time. Partial screenshots may be viewed
in full within the software version specified.
Microsoft® and Windows® are registered trademarks of the Microsoft
Corporation in the U.S.A. and other countries. Screenshots and icons
reprinted with permission from the Microsoft Corporation. This book is not
sponsored or endorsed by or affiliated with the Microsoft Corporation.

Please contact us with concerns about any potential bias at

https://www.pearson.com/report-bias.html.

Warning and Disclaimer

Every effort has been made to make this book as complete and as accurate
as possible, but no warranty or fitness is implied. The information provided
is on an “as is” basis. The author, AARP, and the publisher shall have
neither liability nor responsibility to any person or entity with respect to any
loss or damages arising from the information contained in this book.

Special Sales

For information about buying this title in bulk quantities, or for special sales
opportunities (which may include electronic versions; custom cover
designs; and content particular to your business, training goals, marketing
focus, or branding interests), please contact our corporate sales department
at [email protected] or

(800) 382-3419.
For government sales inquiries, please contact
[email protected].

For questions about sales outside the U.S., please contact

[email protected].

Editor-in-Chief

Brett Bartow

Executive Editor

Laura Norman

Associate Editor

Anshul Sharma

Marketing

Josh DeMatteo

Director, AARP Books

Jodi Lipson; Leah Miller

Development Editor
Charlotte Kughen

Managing Editor

Sandra Schroeder

Senior Project Editor

Tonya Simpson

Copy Editor

The Wordsmithery LLC

Indexer

Cheryl Lenser

Proofreader

Johnna Vanhoose Dinse

Publishing Coordinator

Cindy Teeters

Compositor

Bronkella Publishing LLC

Graphics Processing

tj graham art

OceanofPDF.com
Contents at a Glance

Introduction
1 AI: What It Is and How It Works
2 The Benefits and Risks of AI
3 Getting Started with All-Purpose AI Tools
4 Using AI to Find the Right Words
5 Using AI to Find Information
6 Using AI to Connect with People and Pursue Interests
7 Using AI to Create Art and Images
8 Using AI to Get a Job
9 Using AI at Work
10 Using AI for Travel and Transportation
11 Using AI for Health and Wellness
12 Using AI to Help Caregivers
13 The Future of AI
Index

OceanofPDF.com
Table of Contents

Introduction
AI Is Already Changing Things—and Will Change Even More
What’s In This Book
How This Book Is Organized
This Book Is About Generative AI
1 AI: What It Is and How It Works
What Generative AI Is ...
…and What It Isn’t
Understanding Predictive AI
Understanding Generative AI
Comparing Predictive and Generative AI
A Brief History of Artificial Intelligence
Artificial Intelligence Is an Ancient Concept
The Birth of Modern AI
The First AI Boom—and the First AI Bust
AI in the Twenty-First Century
How Does Generative AI Work?
Understanding the AI Process
What’s Necessary to Make Generative AI Work
How to Use Today’s AI Generators
How You Can Use Generative AI In Your Daily Life
AI in Art
 AI in Business
AI in Finance and Banking
AI in Gaming
AI in Healthcare
AI in Writing
AI for Learning
AI in Marketing and Advertising
AI in Music
AI in Programming and Software Development
AI in Transportation and Travel
AI Everywhere
Summary
2 The Benefits and Risks of AI
Understanding the Potential Benefits of AI
Automating Boring Manual Processes
Improving Business Productivity
Reducing Risk
Making More Things Available 24/7
Personalizing the User Experience
Enabling Better Recommendations
Improving Digital Assistants
Managing Messages
Improving Healthcare
Enhancing Learning
 Augmenting Creativity
Making Your Life Easier
Understanding the Potential Risks of AI
AI Can Spread Misinformation
AI Can Be Biased
AI Can Invade Your Privacy
AI Will Replace Some Jobs
AI Will Make Mistakes
AI Uses Significant Resources
How to Spot AI in the Wild
How to Identify AI-Written Text
How to Identify AI-Generated Photos and Images
How to Identify AI-Generated Videos
How to Identify AI-Generated Music
How to Identify AI-Generated Propaganda
Debating the Ethics of AI-Created Content
Sharing AI-Generated Content
Using AI for Schoolwork
Handling Plagiarism
Using AI Content for Malicious Purposes
Dealing with Built-In Bias
Dealing with Inaccurate Content
Protecting Privacy
AI and Copyright Law
 Does Using Copyrighted Material to Train AI Violate
Copyright Laws?
Is AI-Created Content Copyrightable?
Can AI Violate Copyright Laws with Content It Creates?
Summary
3 Getting Started with All-Purpose AI Tools
What Is an All-Purpose AI Tool—and What Can You Do With It?
How to Use a Generative AI Tool
Getting to Know Publicly Available AI Tools
Freestanding AI Tools
Embedded AI Tools
Using the Most Popular AI Generators
Using ChatGPT
Using Claude
Using Google Gemini
Using Meta AI
Using Microsoft Copilot
Using Perplexity
Using Pi
Using Poe
Which Generative AI Tool Is the Best for You?
How to Construct the Perfect Prompt
Using AI-Generated Results—with Caution
Summary
4 Using AI to Find the Right Words
When and How You Should (and Shouldn’t) Use AI for Writing
Using AI for Different Types of Writing
Writing a Short Note to Family or Friends
Writing a Longer Letter to Family and Friends
Writing an Email Message
Writing a Thank-You Note
Writing a Business Letter
Writing a Social Media Post
Writing a Memoir
Writing a Short Story
Writing a Poem
Using AI to Improve Your Writing
Using AI to Suggest Topics
Using AI to Outline Content
Using AI to Write—and Rewrite—Content
Using AI to Edit Content
Examining Popular AI Tools for Writing and Editing
Grammarly
Hemingway Editor
HyperWrite
ProWritingAid
QuillBot
Sudowrite
 Wordtune
Summary
5 Using AI to Find Information
How AI Differs from Traditional Web Search
Why AI Is a Useful Search and Research Tool
Things to Be Careful of When Using AI for Search and Research
Evaluating Popular AI Tools for Information and Research
Using All-Purpose AI Tools for Search and Research
Using Research-Specific AI Tools
How to Focus AI to Get the Information You Want
Crafting the Perfect Prompt for Research
Examples of Using AI for Research
Fine-Tuning AI Results
Using AI to Summarize and Understand Information
Summary
6 Using AI to Connect with People and Pursue Interests
Using AI to Keep Up with Friends and Family
Ways to Use AI to Connect with Your Friends and Family
Using AI to Share Memories
Using AI to Better Understand People
Use AI to Find and Pursue Hobbies and Interests
Finding New Hobbies
Pursuing Existing Hobbies
Finding Online Communities
 Using AI to Improve Your Social Media Interactions
Conversing with AI by Text or Voice
Talking with AI via Text
Talking with AI via Voice
Summary
7 Using AI to Create Art and Images
How AI Generates Images
How to Use an AI Image Generator
What Kinds of Images Can You Generate with AI?
Cartoons and Comic Book Art
Collages
Fantasy
Fine Art
Greeting Cards
Manufactured Reality
Portraits
How to Ethically Use AI Image Generators
Comparing Popular AI Tools for Creating Images
Adobe Firefly
DALL-E in ChatGPT
DeepAI AI Image Generator
Deep Dream Generator
DreamStudio by Stability AI
Google Gemini
 Hotpot AI Art Generator
Image Creator from Microsoft Designer
Midjourney
NightCafe
OpenArt
Crafting the Perfect Prompt to Generate Perfect Images
Summary
8 Using AI to Get a Job
Using AI to Help Find the Job You Want
Identifying the Types of Jobs That Suit You
Researching Potential Employers
Using AI to Write a Resume
Writing a Resume with an All-Purpose AI Tool
Writing and Fine-Tuning a Resume with a Resume-Writing
Tool
Using AI to Write Cover Letters
Using AI to Prepare for Interviews
Summary
9 Using AI at Work
Using AI to Improve Productivity
Generating Content
Managing Projects
Communicating and Collaborating
Creating Presentations
 Using AI to Make Meetings More Manageable
Scheduling Meetings
Taking Notes and Summarizing Meetings
Summary
10 Using AI for Travel and Transportation
Using AI to Create Travel Plans
Making Personalized Travel Recommendations
Booking Flights and Accommodations
Creating an Itinerary
Functioning as a Virtual Travel Assistant
Using Popular AI-Powered Travel Planning Tools
GuideGeek
Layla
Roam Around
Trip Planner AI
Wonderplan
Using AI to Prepare for a Trip
Researching Documentation Needs
Learning Local Languages
Getting Packing Assistance
Predicting the Weather
Enhancing Your Personal Security
Using AI During a Trip
Getting Personalized Recommendations
 Providing Cultural Insights
Navigating Your Route
Translating on the Fly
Keeping You Safe While Traveling
Using AI in Local Transportation
Getting Where You Want to Go
Optimizing Ride Sharing and Carpooling
Finding Parking
Optimizing the Use of Public Transportation
Summary
11 Using AI for Health and Wellness
Using AI to Create Personal Fitness and Nutrition Plans
Creating a Fitness Plan
Creating Nutrition and Meal Plans
Using AI-Powered Fitness and Nutrition Tools
Using AI for Your Mental Health
Providing Information and Resources
Encouraging Journal Writing
Guiding Cognitive Behavioral Therapy Exercises
Guiding Relaxation Exercises
Using Chatbots for Automated “Therapy”
Using Purpose-Built AI for Mental Health
Using AI to Understand Health and Wellness Information
Researching Medical Conditions and Diagnoses
 Researching Medications
Interpreting Medical Results
Communicating with Healthcare Providers
Summary
12 Using AI to Help Caregivers
Using AI to Help Caregivers with Health Tasks
Getting Smarter about Medical Information
Working with Doctors
Providing Personalized Care Plans
Monitoring Health Conditions and Identifying Trends and
Issues
Improving Nutrition
Using AI to Help Caregivers with Financial and Legal Matters
Managing the Financial Waters
Understanding Legal Issues
Receiving Useful Tips and Personalized Advice
Discovering Other Resources
Using AI for Emotional Support for Caregivers
Using AI to Provide Virtual Companionship and Assistance
Examining Other AI Tools for Caregivers
Summary
13 The Future of AI
What’s Next for AI?
AI Gets Smarter, Faster, and Less Expensive
 From Artificial Intelligence to Artificial Superintelligence
AI Gets Embedded in Other Devices
AI Merges with Other Technologies and Services
AI Gets Personalized
How Will AI Impact Your Life in the Future?
AI and You at Home
AI and You at Work
AI and You at Play
AI and You All Day, Every Day
What Risks Lie Ahead?
We Lose Oversight and Control
We Let AI Become Weaponized
The Ultimate AI Risk: Achieving the Singularity
Summary
Index

OceanofPDF.com
About the Author

This book is written by me, Michael Miller. I’ve written more than 200
books over the past four decades, including many in the Absolute
Beginner’s Guide series and almost two dozen books with AARP. My books
have collectively sold more than 1.5 million copies worldwide, so I must be
doing something right.

I have a particular interest in new and evolving technologies, so writing

about AI is right up my alley. My readers say I have a knack for explaining
complex technologies in easy-to-understand terms, and I have no reason to
argue with that. I hope I can help you better understand artificial
intelligence and how you can use it in your everyday life.

Oh, in case you’re wondering, I am an older gentleman, which only means

I’ve been around long enough to experience many different “new”
technologies. (I have been writing for four decades, after all; I wrote one of
the first books about the Internet when it was new!) I live with my wife and
random daily combinations of four stepchildren and eight grandchildren in
a suburb of the Twin Cities in the often-frozen land of Minnesota. In my
spare time, what there is of it, I play drums, entertain (and am entertained
by) my grandchildren, and write about music from the ’60s, ’70s, and ’80s
in my Classic Song of the Day blog (www.classicsongoftheday.com).
You can find out more about me on my website, located at
www.millerwriter.com. Feel free to use the contact form there to get in
touch with me. I’m open to criticism, faint praise, and the occasional
earnest question. I may or may not respond personally, although I guarantee
I’ll read everything you write.

OceanofPDF.com
Dedication

To my eight wonderful grandkids who will grow up in an AI world: Collin,

Alethia, Hayley, Judah, Lael, Jackson, Jamie, and Adelynn.

OceanofPDF.com
Acknowledgments

My thanks to everyone at Pearson who helped turned my words into a book,

including Laura Norman, Anshul Sharma, and Charlotte Kughen. Thanks
also to the good folks at AARP for their input and support, including Jodi
Lipson, Leah Miller, and Michael Phillips.

OceanofPDF.com
About AARP

AARP is the nation’s largest nonprofit, nonpartisan organization dedicated

to empowering people 50 and older to choose how they live as they age.
With a nationwide presence and nearly 38 million members, AARP
strengthens communities and advocates for what matters most to families:
health security, financial stability and personal fulfillment. AARP also
produces the nation’s largest circulation publications: AARP The Magazine
and AARP Bulletin. To learn more, visit www.aarp.org,
www.aarp.org/espanol or follow @AARP, @AARPenEspanol and
@AARPadvocates, @AliadosAdelante on social media.

OceanofPDF.com
We Want to Hear from You!

As the reader of this book, you are our most important critic and
commentator. We value your opinion and want to know what we’re doing
right, what we could do better, what areas you’d like to see us publish in,
and any other words of wisdom you’re willing to pass our way.

We welcome your comments. You can email or write to let us know what
you did or didn’t like about this book[md]as well as what we can do to
make our books better.

Please note that we cannot help you with technical problems related to the
topic of this book.

When you email, please be sure to include this book’s title and author as
well as your name and email address. We will carefully review your
comments and share them with the author and editors who worked on the
book.

Email: [email protected]

OceanofPDF.com
Reader Services

Visit our website and register this book at informit.com/register for

convenient access to any updates, downloads, or errata that might be
available for this book.

OceanofPDF.com
Figure Credits

Cover & UN Images_ Description: Cute Little Chicks:

Rozdesign/Shutterstock

Figure 1.1: Wikimedia Commons

Figures 1.3–1.5: OpenAI Inc

Figure 1.6: Deep AI, Inc

Figure 2.1: Deep AI, Inc

Figure 2.2: © Microsoft 2024

Figure 2.3: Freepik Company S.L.

Figure 2.4: Pixlr Pte Ltd

Figure 2.5: GPTZero

Figures 3.1, 3.2, 3.34: Google LLC

Figures 3.3, 3.4, 3.21–3.23: © Microsoft 2024

Figures 3.5–3.8, 3.33, 3.36, 3.39: OpenAI Inc

Figures 3.9–3.12: © 2024 Anthropic PBC

Figures 3.13–3.17, 3.37, 3.40, 3.44–3.48: Google LLC

Figures 3.18–3.20, 3.35, 3.38, 3.41–3.43: © 2024 Meta

Figures 3.24–3.27: Perplexity AI

Figures 3.28, 3.29: Inflection AI

Figures 3.30–3.32: Quora, Inc.

Figures 4.1, 4.4, 4.7, 4.10: OpenAI Inc

Figures 4.2, 4.3, 4.8, 4.11: Google LLC

Figure 4.5: © Microsoft 2024

Figures 4.6, 4.9: © 2024 Meta

Figures 4.12–4.14: Grammarly Inc.

Figures 4.15–4.17: Hemingway

Figures 4.18–4.20: HyperWrite

Figures 4.21–4.23: © 2024 Orpheus Technology

Figures 4.24, 4.25: QuillBot

Figures 4.26, 4.27: Sudowrite

Figures 4.28–4.30: Wordtune

Figures 5.1, 5.3, 5.6, 5.16: Google LLC

Figures 5.2–5.5: OpenAI Inc

Figure 5.7: © 2024 Meta

Figures 5.8, 5.9: © 2024 Consensus NLP

Figures 5.10, 5.11: Ought

Figures 5.12, 5.13: Scholarcy Limited

Figures 5.14, 5.15: Scite LLC

Figures 6.1, 6.8: © 2024 Anthropic PBC

Figures 6.2, 6.3, 6.6, 6.13–6.15: OpenAI Inc

Figures 6.4, 6.7: Google LLC

Figures 6.5, 6.9: © 2024 Meta

Figures 6.10, 6.11: © Microsoft 2024

Figure 6.12: Inflection AI

Figures 7.2, 7.11, 7.14, 7.15: Adobe

Figures 7.17, 7.18: Deep AI, Inc

Figures 7.19, 7.20: © 2024 Deep Dream Generator

Figures 7.6, 7.12, 7.21, 7.22: STABILITY AI

Figures 7.9, 7.23: Google LLC

Figures 7.24, 7.25: Panabee, LLC

Figures 7.1, 7.10, 7.26, 7.27: © Microsoft 2024

Figures 7.3, 7.8, 7.28: 7.32: Midjourney, Inc

Figures 7.33, 7.34: NightCafe Studio

Figures 7.5, 7.7, 7.13, 7.35, 7.36: OpenArt

Figures 7.4, 7.16: OpenAI Inc

Figure 8.1: OpenAI Inc

Figure 8.2: Google LLC

Figure 8.3: © 2024 Jobscan

Figure 8.4: Kickresume

Figure 8.5: Rezi

Figure 8.6: Teal Labs, Inc

Figure 8.7: Works Limited.

Figure 9.1: Monday.com

Figure 9.2: OpenAI Inc

Figure 9.3: Sendsteps

Figure 9.4: Reclaim.ai, Inc.

Figure 9.5: Otter.ai

Figure 10.1: GuideGeek

Figures 10.2, 10.3: Layla

Figures 10.4, 10.5: Roam Around

Figures 10.6, 10.7: Trip Planner AI.

Figures 10.8, 10.9: Wonderplan

Figures 10.10, 10.12: OpenAI Inc

Figure 10.11: © 2024 Anthropic PBC

Figure 10.13: Vasco Electronics

Figures 11.1, 11.8, 11.12: © 2024 Meta

Figures 11.2, 11.6, 11.7, 11.13: OpenAI Inc

Figure 11.3: © 2024 Anthropic PBC

Figures 11.5, 11.10: © Microsoft 2024

Figures 11.4, 11.9, 11.11, 11.14: Google LLC

Figure 12.1: SixD Incorporated

Figure 12.2: © Intuition Robotics Inc. 2024

Figure 13.1: (Based on Superintelligence: Paths, Dangers, Strategies by

Nick Bostrom. Estimates of the timing and duration of takeoff were not
included on Bostrom’s original graph.)

Figure 13.2: © 2024 Engineered Arts

OceanofPDF.com
Introduction

You can’t open a newspaper, load a web page, or scroll through a TikTok
feed without being presented with either inspiring or alarming stories about
AI.

Artificial intelligence—AI—is considered one of the most significant

technological developments of the current century. And its impact is just
beginning.

In its simplest form, AI is the idea of machines that learn and think like
humans. Long a concept in science fiction (think Frankenstein and “Star
Trek”), AI has only in the past years become a reality.

In fact, your life is already being touched by AI, and it’s likely you’ll be
affected even more in the months and years to come. It’s not just theoretical
or for scientists and researchers; AI is a technology that you can use in your
daily life.

If you think AI might be a passing fad, take a look at some recent headlines:

“AI Will Make Mental Healthcare More Human” (Psychology Today)

“Artists’ AI Dilemma: Can Artificial Intelligence Make Intelligent Art?”
(The Guardian)
“How to Protect Yourself (and Your Loved Ones) From AI Scam Calls”
(Wired)
 “Jamie Dimon Says AI May Be as Impactful on Humanity as Printing
Press, Electricity and Computers” (CNBC)
“‘Social Order Could Collapse’ in AI Era, Two Top Japan Companies
Say” (Wall Street Journal)
“Will Artificial Intelligence Boost Productivity? Companies Sure Hope
So” (The Seattle Times)

All those stories—and many more--were published in a single day. As you

read them, you may find it difficult to get a grasp on what AI is and its
potential impact on our world. Is AI the greatest thing since fire and sliced
bread, or will it cause the demise of humankind as we know it? How can it
enhance your life and make it better? What should you watch out for?

The answers, at this point in time, are obviously less than clear. What is
clear is that AI is here—and here to stay.

AI Is Already Changing Things—and Will Change Even More

AI is all around you, including places you may not even realize. AI powers
personalized content recommendations on your favorite streaming video
sites. It drives real-time navigation apps on your phone. It’s used in virtual
assistants like Alexa and Siri. It’s even a part of supply chain management
that puts the products you need on store shelves.

And that’s just the simple AI that’s been in use for the past few years.
Newer generative AI—what most of this book is about—doesn’t just power
things; it empowers you to create new things. Generative AI lets you create
stories, images, video, and audio with just a few brief prompts. You can use
AI to help you write letters and social media posts; people who have trouble
drawing a straight line (like me!) can use AI to create both fantastic and
realistic images; businesses can use AI to answer customer questions and
manage employee schedules. It’s all out there, and there’s even more to
come.

It’s clear that AI will have a radical effect on our lives. We can expect to
interact with AI in much of what we do—writing emails, articles, and blog
posts; creating art; finding information; managing finances; traveling; and
obtaining medical care. There’s little in this world that AI won’t impact.

That doesn’t mean AI is perfect or always safe or easy to use. AI is a

powerful technology that can be confusing and intimidating, even to those
who work with it every day. There are seemingly unlimited ways for
technology to incorporate AI and unlimited ways for people like you and
me to use AI to assist in our daily lives.

There are also a seemingly unlimited number of AI tools available today,

with more introduced constantly. Sorting through these options and
understanding how they work—and how you can get the most out of them
—are just a few of the things we as individuals need to master moving
forward.
It’s important to be aware of and able to recognize the limitations and risks
presented by this new technology. Because AI is trained on data from
human beings, it can represent both the best and the worst of our society. AI
can be used to spread misinformation. It can increase the sophistication of
scams. It can reflect societal biases such as preconceptions about age, race,
and gender. It can offer up intellectual property that belongs to someone
else. And it can also be seen as threatening to human security; some experts
believe that AI has the potential to replace humans in many areas, if not
completely. Whatever its ultimate impact, AI is part of our world and we
will all come into contact with it in our daily lives. Just as we’ve had to
learn how to use the new technologies of the past, from washing machines
and motor vehicles to personal computers and the Internet, we need to learn
how to use and get the most out of AI going forward. By getting out there
and experimenting with today’s AI tools, you can discover all the amazing
ways that AI can help you become both more productive and more creative.
And how to know where you shouldn’t use it, or where you should use it
with caution.

That’s what this book is about.

What’s In This Book

Using Artificial Intelligence: Absolute Beginner’s Guide is a book about

what AI is—the pros and cons, benefits and risks—and how to use it. It’s a
book about a highly technical subject written for a nontechnical audience in
easy-to-understand language. I try to make AI as understandable as possible
while focusing on the technology’s practical aspects. That is, this book
eschews all the technical mumbo jumbo in favor of showing you the many
ways you can use AI to perform everyday tasks.

If I do my job right as an author, this book will help you

Understand, at a very basic level, how AI works

Learn how AI can benefit you in your life
Recognize and manage some of AI’s risks, limitations, and security
implications
Discover commonly available AI tools and use them for specific tasks
Identify when AI can be helpful and trusted and when it can’t
Spot AI-generated content

I include lots of examples to show you how AI works and give you step-by-
step instructions so you can try AI for yourself. You’ll see how AI works in
different applications, from writing letters and creating art to managing
your health and helping you be more productive at work. The emphasis is
on the practical use of AI—how you might encounter and employ the
technology in the real world.

How This Book Is Organized

The information in Using Artificial Intelligence: Absolute Beginner’s Guide

is organized into 13 chapters and a glossary. Here’s how that information is
presented:

Chapter 1, “AI: What It Is and How It Works”: This first chapter digs
into the ideas and technologies behind AI, focusing on the newer, more
powerful generative AI. You’ll learn AI’s history, how it works, and how AI
is being used today.

Chapter 2, “The Benefits and Risks of AI”: Here is where I discuss the
great ways that AI can help you—and what to watch out for.

Chapter 3, “Getting Started with All-Purpose AI Tools”: This chapter

gets into the fun stuff: how to use free and publicly available AI tools—
such as ChatGPT, Google Gemini, Meta AI, and Microsoft Copilot —for
fun and in practical applications.

Chapter 4, “Using AI to Find the Right Words”: Learn how to use AI to

assist in writing everything from email messages and letters to articles and
business reports. You can even use AI to write poetry and fiction for your
own personal use (you can’t claim AI-written material as your own!)—and
edit your existing work to make it read better.

Chapter 5, “Using AI to Find Information”: AI is several steps beyond

today’s web-based search engines. Discover how AI can find information
faster—and help you understand what it finds—and when you need to
verify AI’s output (short answer: always!). Whether you want to figure out
how to fix your car or plan dinner for six with picky eaters, AI can help. It
can also be a great tool to help you summarize information.

Chapter 6, “Using AI to Connect with People and Pursue Interests”: AI

can help you connect with other people who share your interests. You can
even use AI to manage—or generate—your online conversations and help
you pursue your favorite hobbies. This chapter also covers how and when to
have helpful conversations with AI

Chapter 7, “Using AI to Create Art and Images”: Discover the easy-to-

use AI tools that let you create photorealistic images or pieces of art that
transcend your imagination.

Chapter 8, “Using AI to Get a Job”: Get a leg up on other applicants by

using AI tools to find the right job, craft the perfect resume and cover letter,
and prepare for job interviews.

Chapter 9, “Using AI at Work”: It’s not surprising that employers are

looking for AI to improve productivity. Learn how to use today’s AI-based
tools to automate routine tasks, collaborate with colleagues, and make
better business decisions.

Chapter 10, “Using AI to Manage Your Travel and Transportation”:

Discover how AI can help you create trip plans, find the best places to stay
(at the best prices), and prepare for your journey. You’ll also learn about
self-driving vehicles and using AI to map the most efficient routes.
Chapter 11, “Using AI for Health and Wellness”: Learn how you can use
AI to create fitness and nutrition plans, better understand what your
providers tell you, and improve your mental health.

Chapter 12 “Using AI to Help Caregivers”: If you are one of the more

than 37 million people caring for a relative or friend, you’re probably eager
to get all the help you can—and AI can help you become a better caregiver.
Discover how AI can automate daily tasks such as pill management,
improve personal health and safety, and even provide companionship.

Chapter 13, “The Future of AI”: The first dozen chapters of this book tell
you where AI is today. This final chapter predicts what’s next for AI—how
AI will evolve and how it will affect our lives in the future.

Glossary: This section provides a list of AI-related terms you need to know.
Learning these terms won’t make you an AI expert, but using them may
make you sound like one.

One last thing. As you read through this book, you’ll see a variety of notes
that provide additional information and warnings that alert you to AI’s
limitations. You’ll even find some extended sidebars that provide tangential
information and responses to frequently asked questions. These notes,
warnings, and sidebars aren’t essential for learning how to use AI, but you
may find them interesting or helpful.
This Book Is About Generative AI

As you’ll learn in the very first chapter of this book, there are two types of
AI in use today: the older predictive AI and the newer generative AI. While
predictive AI is interesting and useful, and the type we know best—asking
Siri questions, getting Alexa to turn on music, having your grammar and
spelling checked when you’re writing—it’s the more powerful (and, dare I
say, more “intelligent”) generative AI that promises to have the biggest
impact on society. Generative AI is so named because it can generate new
content—words, pictures, sounds, ideas—that didn’t exist previously. This
ability to create something from nothing is what gives generative AI its
promise to transform our world.

For all these reasons, this book focuses on generative AI—what it is and
how to use it. I won’t completely ignore predictive AI, but generative AI is
where it’s at today and where I devote the bulk of my attention and
coverage.

OceanofPDF.com
1

AI: What It Is and How It Works

IN THIS CHAPTER

What generative AI is…and what it isn’t

Predictive AI versus generative AI: what’s the difference?
A brief history of artificial intelligence
How does generative AI work?
How to use today’s AI generators
How you can use generative AI in your daily life

Artificial intelligence—AI—already touches our lives in innumerable ways,

and it will continue to do so in ways we can only imagine today. The newest
form of AI, generative AI (what this book mostly focuses on), promises to
help individuals like you and me create all sorts of new content. But to get
the most out of generative AI, it’s helpful to have an understanding of this
new technology—which is what you’ll get from this chapter.

What Generative AI Is ...

Let’s start with the basics. What exactly is artificial intelligence?

When I asked ChatGPT, one of the most popular generative AI tools today,
that question, here’s how it responded:
AI means making computers do things that humans do, like learning,
reasoning, and problem-solving. The aim is to create smart systems that can
do tasks that usually need human intelligence.

Not surprisingly, that’s not a bad answer. (You’d expect AI to know what AI
is.) Put in more human terms, AI is an intelligence not naturally born. It is a
simulation of natural human intelligence created by machines.

AI is all about computers that go beyond computing to also “think.” That

means they don’t merely regurgitate facts from existing sources but actually
come up with new ideas and processes. AI “learns” from enormous amounts
of data it’s exposed to, creating something new—just like human beings
can.

The goal of AI is to create machines that can think like humans and also
perform tasks that go beyond the capabilities of human minds. It happens
when computers ingest and analyze large amounts of data, identify patterns,
and extrapolate from those patterns to deduce solutions and make decisions.

As to why AI exists, I’ll let ChatGPT answer that question, too:

Overall, AI exists to enhance human capabilities, improve our quality of

life, drive innovation, and address some of the most pressing challenges
facing society today.

I couldn’t have put it better myself.

…and What It Isn’t

Note that AI isn’t a “smart” app on your phone—it’s much more advanced
than that. While there are some who seek to exploit the buzz around AI by
applying the term “artificial intelligence” to existing processes and
applications, true AI goes well beyond the technologies in use today.

For example, today’s web-based search engines—Bing, Google, Yahoo, and

the rest—might look like AI as they try to understand your queries and
deliver relevant results. But there’s not much intelligence behind what they
do; they use simple algorithms to deliver certain results based on defined
inputs. They don’t think or figure things out, they just follow a set of rules
to try to find the best results. (Although you may notice that Google
recently began adding AI’s best thinking to the top of your search results.)

Note

An algorithm is a set of step-by-step rules or instructions

designed to perform a task or solve a problem. Think of an
algorithm along the lines of “if this happens, then do that; if
something else happens, then do this other thing.”
Likewise, many online chatbots purport to use AI technology but actually
don’t. One chatbot I’ve used for online customer service claims to be “AI
powered” but, in reality, there is nothing intelligent about it; it simply
supplies prewritten answers based on predetermined customer questions.
That is, you pick a question from a list and it spits out the appropriate
answer. That may be artificial, but it certainly isn’t intelligent.

Understanding Predictive AI

Predictive AI is an older, more established form of artificial intelligence.

Until very recently, it was the only kind of AI you were likely to encounter
in day-to-day life.

Note

Predictive AI is also known as traditional AI, narrow AI, or

weak AI.

Predictive AI is so named because it is most often used to predict specific

outcomes or trends based on existing data. Predictive AI is used by today’s
web search engines to predict what web pages match your query; by social
media companies to predict what posts you’re mostly likely interested in;
and by streaming video and music services to predict what shows or songs
you might like.

Predictive AI is task-specific, typically operating with a predefined set of

instructions for a specific task. It uses the standard AI processes of data
collection, data processing, outcomes, assessments, and adjustments but is
based primarily on historical data. It analyses that historical data to
understand patterns and trends it can use to make predictions about future
trends and events.

You can find predictive AI in use today in a variety of applications, where it

predicts

Traffic flow
Social media likes/dislikes
User search queries
Viewer/listener likes/dislikes
Spelling and grammar checks
Need for preventive maintenance
Credit risk analysis
Customer demand forecasting
Inventory level predictions
Patient outcomes in healthcare
Stock prices
Understanding Generative AI

Generative AI is a newer type of artificial intelligence that can generate new

data similar to and based on existing information and examples. Unlike
predictive AI, which is focused on making specific predictions, generative
AI is capable of generating completely new data that has not previously
existed. As the name implies, generative AI can generate text, audio,
images, videos, music, and other content that is often indistinguishable from
that created by humans.

Note

Generative AI is also known as strong AI.

Generative AI can be used to create all sorts of things, from newspaper

articles and photorealistic artwork to web pages and software programs.
(That’s right, generative AI can create programming code—much faster
than can human developers.) It can generate hypothetical scenarios and
simulations for training and educational purposes. It can create new and
totally unique content beyond what it was trained on.

You can find generative AI in use today in

 AI text generators and chatbots, such as Anthropic’s Claude, Google’s
Gemini, Microsoft’s Copilot, and OpenAI’s ChatGPT
AI image generators, such as Adobe Firefly, DALL-E, DeepAI’s AI
Image Generator, Deep Dream Generator, Microsoft Designer’s Image
Creator, and Midjourney
AI video generators, such as Colossyan AI, DeepBrain AI, HeyGen AI,
and SundaySky

Note

Generative or strong AI isn’t the ultimate form of artificial

intelligence; rather, experts describe the currently theoretical
super AI to be a form of artificial intelligence that surpasses
human intelligence. Learn more about super AI in Chapter
13, “The Future of AI.”

Comparing Predictive and Generative AI

Predictive and generative AI are related in that they use similar learning
techniques, but they’re substantially different in terms of what they’re
designed to achieve. In essence, predictive AI anticipates future outcomes
while generative AI creates new outcomes. Predictive AI is smart but
generative AI is creative. Both have their place in our collective future.

Note

Generative AI is the newest, most powerful form of AI, and

what the majority of this book is about.

Not surprisingly, generative AI models require larger datasets—that is, the

bank of information that AI needs to draw from—than predictive models.
Generative AI also requires a correspondingly larger amount of computing
power. Table 1.1 details the differences between predictive AI and
generative AI.

TABLE 1-1 Comparing Predictive AI and Generative AI

Predictive AI Generative AI

Objective Predict future Generate new data or

outcomes or trends content (in addition to
predicting outcomes and
trends)
 Predictive AI Generative AI

Data Historical and Large quantities of high-

requirements current data within quality data
a specific topic area

Output Predictions based Newly generated text,

on input data images, videos, and audio

Applications Finance, healthcare, All the applications of

marketing, predictive AI plus art and
chatbots, web design, writing, music
search production, travel and
transportation, caregiving,
training simulations, and
more

Tech Moderate to high Much higher computational

requirements levels of power and large data
computational storage capacity
power and data
storage capacity
Dispelling Some Common Misconceptions About AI

You’ve no doubt heard about AI and some of what it can do, but you may
possess some misconceptions about the technology—what it does and how
it does it. So let’s take a moment to look at and dispel some of the most
common misconceptions:

AI is the same as human intelligence. Some people have the

misconception that AI possesses human-like intelligence and thinks like
humans do. Nothing could be further from the truth. Today’s AI systems
are specialized and possess narrow capabilities. AI systems lack human
understanding, consciousness, and emotions. AI is not self-aware; it is
not human.
AI will achieve consciousness. The idea that AI will somehow achieve
sentience and become human is a concept rooted in science fiction.
There’s no evidence that AI will develop self-awareness, at least any
time soon. Today, AI is just a (very complex) computer program.
AI is always right. While AI systems are much more accurate than
humans for some tasks, they can make mistakes—especially if the data
they’re working with contains inaccuracies. In fact, AI is known for its
“hallucinations,” where it spits out wildly incorrect information.
Although AI does learn from its mistakes, it’s not 100 percent reliable. In
fact, you should verify everything that comes out of AI.
AI is a recent invention. You may just now be hearing about AI, but it’s
been around for decades. It’s only now becoming feasible on a mass
 scale.
AI will replace all human jobs. While AI can and probably will
automate many types of tasks, especially repetitive ones, it can’t replace
everything humans do. That said, AI stands to have a huge impact on the
job market—and not always in a positive fashion. (Learn more about the
risks of AI in Chapter 2, “The Benefits and Risks of AI” and AI’s impact
on how you work in Chapter 9, “Using AI at Work.”)
AI will solve all the world’s problems. If AI is so smart, it’s just a
matter of time before it figures out the answer to life, the universe, and
everything, right? (And, no, the answer isn’t “42”—and if you’re not
sure what that means, you can ask an AI tool to explain it to you!) Well,
you might think so or hope so, but as promising as AI is, it has its
limitations. It’s just a tool that we humans can use to tackle all those
problems, nothing more and nothing less.
AI will cause the extinction of the human race. This is the big fear:
that AI achieves sentience, gets smarter than us, and decides that it
doesn’t need us puny humans. While some say it’s a possibility, others
say it’s highly unlikely. (Turn to Chapter 13 to learn more about this and
other possible future scenarios for AI technology.)

You may have more thoughts and opinions about AI that may or may not be
accurate. I hope you discover the real truth as you read this book.
A Brief History of Artificial Intelligence

Before we get into the nitty gritty details of how AI works, let’s take a brief
look at where it came from and how it developed. (If you’re not a history
buff, feel free to skip this section!)

AI technology isn’t really new, although recent developments in the

technology have brought it further into our everyday lives. Just when and
where did AI first come about—and how has it developed over the years?

Artificial Intelligence Is an Ancient Concept

The idea of some type of artificial intelligence predates modern times by

thousands of years. Perhaps the first instance of inanimate objects
exhibiting intelligence is from the myth of the Greek god Hephaestus
creating automatons endowed with artificial intelligence. The concept of
artificially intelligent automatons continued through the ages, through Hero
of Alexandria’s mechanical theater to Leonardo da Vinci’s designs for a
mechanical knight to Wolfgang von Kemplelen’s chess-playing automaton,
The Turk. None of these fictional or real-world designs used true artificial
intelligence, however.

The Birth of Modern AI

AI took a step closer to reality in the mid-twentieth century with the

development of the modern computer, which was necessary to power AI
technology. That enabled Alan Turing, in 1950, to devise what became
known as the Turing Test to determine whether a computer possesses
intelligence.

Essentially, the Turing Test evaluates a machine’s ability to exhibit

intelligent behavior by having an evaluator blindly pose a series of
questions to both a computer and a human being; if the evaluator can’t tell
the difference between the two responses, the machine is said to possess
intelligence. The Turing Test is still being used today to evaluate AI
systems.

The modern era of AI is generally agreed to have started in 1956 during a

conference of AI experts at Dartmouth College. This conference jump
started the AI movement, and government and industry attention and
funding soon followed—as did a plethora of AI research.

One byproduct of this research was a natural language processing program

called ELIZA that laid the foundation for today’s chatbots. ELIZA could
simulate human conversation by using pattern-matching algorithms that
gave the impression that users were conversing with a real human being.
FIGURE 1.1

A typical conversation with ELIZA, the prototype for today’s AI-driven chatbots. (Public domain
image courtesy of Wikimedia Commons)

The First AI Boom—and the First AI Bust

The period from 1980 to 1987 saw what many called the first AI boom.
During these years, AI research showed rapid growth with many important
breakthroughs, all supported by ample government funding. This AI boom
was followed by what became known as the AI winter, from 1987 to 1993,
which saw a dramatic decline in interest in and funding of AI technology.
Private investors and the government, seeing a relatively low immediate
return on their previous investments, pulled back on their AI funding.
The AI winter began to thaw in the 1990s. Rapid increases in computing
power, paired with an explosion of data spawning from the development of
the public Internet, sparked new breakthroughs in everything AI—natural
language processing, machine learning, deep learning, robotics, and the
like. It was during this period that IBM’s Deep Blue computer program first
defeated Russian chess grandmaster Garry Kasparov, a major victory for AI
technology.

AI in the Twenty-First Century

AI bounced back significantly at and after the turn of the century, again
fueled by advancements in computing power and data inventory. AI
developments previously limited to research laboratories were transformed
into actual products for businesses and computers. This modern AI boom
continues to this day and keeps accelerating at its pace.

Consider the following AI-related developments since the turn of the

century:

Google search engine, using predictive AI to anticipate user queries,

launched in 2000.
Netflix’s recommendation system was implemented in 2000.
Amazon’s recommendation engine was introduced in 2001.
Roomba, the first AI-driven robotic vacuum was introduced in 2002.
 NASA’s AI-powered rovers Spirit and Opportunity landed on Mars and
traverse the red planet without human guidance in 2004.
Waymo, Google’s self-driving car project was launched in 2009.
IBM’s Watson natural language computer system launched in 2011.
Apple’s Siri digital assistant launched and, in 2011, competed on the
Jeopardy! quiz show.
Amazon’s Alexa voice assistant launched in 2014.
Google’s open-source deep learning framework, TensorFlow, launched
in 2015.
Facebook, Twitter, and other social media companies started using AI to
power the algorithms that introduce new content to users in the 2010s.
OpenAI research laboratory, founded in 2015, launched the GPT-3 AI
language model in 2020 and DALL-E image generator in 2021.
Vehicles from Tesla and other car manufacturers that incorporate self-
driving features reached the marketplace in the 2020s.

The 2020s also saw the advent of generative AI, the successor to the earlier
and simpler predictive AI. As we’ve discussed, generative AI lets users
generate new content by entering a simple prompt—a major advancement
in the use of AI technology.

In short, AI has a long history with accelerated development in recent years.

A lot has happened—and there’s a lot more to go.
AI in Fiction

While AI was being developed in real-world research laboratories, the

concept of artificial intelligence also took hold in fiction. Some of the more
notable fictional imaginings of AI include

Star Trek: The Next Generation, the television series that ran from 1987
to 1994 (and had subsequent movies and sequels) that featured the
android Lieutenant Data as an AI-powered almost-human member of the
ship’s crew.
The Terminator, a 1984 film by director James Cameron that starred
Arnold Schwarzenegger as an android from the future fighting a war
between humans and the AI network known as Skynet.
Westworld, a 1973 film (and later, in 2016, a television series) that
revolved around a wild west-themed amusement park populated by
androids that eventually gain sentience.
2001: A Space Odyssey, a landmark 1968 film by director Stanley
Kubrick (and written by Arthur C. Clarke) that featured a sentient
computer system named HAL 9000 that eventually runs somewhat
amuck.
“I Sing the Body Electric,” a 1962 episode of The Twilight Zone, written
by Ray Bradbury, all about an intelligent android grandmother. (This
episode became the basis for Bradbury’s 1969 short story of the same
name.)
 I, Robot, a 1950 collection of short stories by Isaac Asimov that
introduced Asimov’s Three Laws of Robotics, which have had a major
impact on both science fiction and real-world artificial intelligence.

There’s a lot more than that, of course, but this gives a sense of how the
concept of artificial intelligence has intrigued society over time. I wonder
what kind of AI fiction will be written now that artificial intelligence is
finally becoming a reality?

How Does Generative AI Work?

Now that you know what AI is and where it came from, it’s time to learn a
little bit about how AI works.

In a nutshell, the AI process starts by collecting large amounts of data from

a variety of sources. The AI system needs this data as source material to
learn how the world works, just as we humans need real-world experiences
along with books and other media to feed us information.

With all this data in hand, AI now starts looking at the data. It uses a variety
of algorithms and technologies to extract key information from the
assembled data, identify patterns, and make connections between different
pieces of data.
Making these connections helps the AI model learn. It uses what it learns to
make predictions and then evaluates its results and learns from them. It
adapts based on these outcomes, essentially teaching itself from its
experiences. Again, it’s very much like the way we humans learn by
experiencing and adapting, just all done by machine.

Understanding the AI Process

The AI process is as simple and as complex as what was just described. To

do all that, however, requires the use of some very sophisticated
technologies and lots of computing power. It’s a five-step process that keeps
reiterating and learning, as shown in Figure 1.2. Again, if this type of
technical detail makes your eyes glaze over, skip ahead!
FIGURE 1.2

The five steps of the AI process

Step 1: Data Collection

The data collection process assembles data of various types from multiple
sources. Data can be in the form of text, speech, images, videos—you name
it. It can come from existing databases, social media feeds, websites on the
Internet, and more.

More targeted AI requires data that is applicable to only a specific purpose.

For example, AI designed for use in healthcare would collect reams of data
from medical sources but wouldn’t need data about interplanetary
mechanics or seventeenth-century architecture. In contrast, general AI
models require a vast amount of data across a wide variety of topics—a lot
of everything. To plan a dinner menu for a small party, AI would need lots
of recipes and examples of menus of small dinner parties.

Whatever the model, the more (applicable) data, the better. An AI model
operating on a limited amount of data won’t have enough information to
learn from and make decisions. The more data available, the faster and
more appropriate the model’s learning.

Step 2: Data Processing

Data processing is actually a big step that involves a lot of different

substeps and technologies.

First, the data collected must be cleaned and preprocessed. That means
rooting out inaccurate or incomplete data and standardizing or formatting
the data to make it easier to process. This often involves the use of natural
language processing (NLP) to understand the meaning behind the words in
text and speech.

Next, AI uses various algorithms to analyze and learn from the data. As
previously discussed, an algorithm is a set of rules to be followed in a
process or problem-solving operation. Think of an algorithm working like
“if this happens, then do that.”
To use these algorithms, AI employs machine learning (ML) techniques
that train algorithms to find patterns and identify underlying structure.
Machine learning enables the model to learn automatically without being
programmed to do so.

AI models also employ deep learning, a type of machine learning that uses
artificial neural networks to recognize complex patterns at multiple levels.
Deep learning mimics human neural networks to process data, find
connections between data, and make inferences based on that data.

Note

A neural network is a type of machine learning algorithm

with multiple interconnected nodes that perform specific
functions, such as receiving data, processing data, and
generating results.

Some AI engines also employ large language models (LLMs) to more

quickly process large amounts of text. An LLM is a kind of deep learning
model that is pre-trained on vast amounts of data. Think of it as a
prepackaged data starter kit that enables AI to have a bit of a head start
rather than beginning from scratch.
When working with visual data, AI models often employ computer vision.
This technology uses pattern recognition and deep learning to interpret the
content of an image.

The goal of this data processing is to interpret the data collected, make
predictions based on that data, and then ultimately act on those predictions.
It’s the prediction part of the process that leads to the next step, generating
outcomes.

Continuing with the dinner party example from Step 1, in this step, AI
would read all those recipes and menus to learn what was in them.

Step 3: Outcomes

After data is processed and patterns within that data identified, AI models
use those patterns to predict various outcomes. For example, a marketing-
focused AI model might use data patterns to predict future market trends.

In the outcome step, AI must determine whether specific data matches

previous patterns. This sets up a pass/fail situation that helps the AI model
determine outcomes that can be used for future decision-making.

Using the dinner party example, in this step, AI would try putting together
different dishes in a sample menu and maybe even hold a “test” dinner to
see how people respond to the suggested menu.
Step 4: Assessments

The outcomes of the previous steps are now assessed to gain further insight.
This assessment process involves analysis of the outcome, discovery of
what triggered the outcome, and feedback that can be incorporated into the
algorithm going forward—which is what happens in the next step.

Continuing the example of a dinner menu, AI would evaluate the results of

your test menu to determine what worked and what didn’t.

Step 5: Adjustments

If data passes the outcomes test, it verifies the previously identified pattern.
If data fails the test—that is, if it doesn’t fit into those previous patterns—
then the AI model must make adjustments. It might adjust the input data,
algorithm rules, or target outcomes.

These adjustments are then incorporated into the AI model in a kind of

feedback loop. That is, the adjustments are fed back into the data collection,
data processing, and outcome steps to fine-tune the entire model. Thus the
model learns over time as it processes data.

In terms of the dinner party example, this is where AI would take what it
has learned and make any necessary adjustments to create the final menu.
(AI would also collect feedback from your real dinner to determine how
you might change things for the next dinner you host.)
What’s Necessary to Make Generative AI Work

The AI process is simple enough to understand but requires a large amount

of resources to make it work—especially the more advanced generative AI.
An accurately trained generative AI model must have processed millions of
pieces of data, and that takes a lot of computing power.

The first thing you need to make AI work is reams of data from a variety of
sources. Such quantities of data were not readily available until the very
recent past, when the birth of the Internet provided easy access to virtually
all the data in the world.

Processing all that data requires enormous amounts of computational

power. All those computations and reiterations can be accomplished only by
computers running fast, high-powered CPUs and GPUs. Such computing
power simply didn’t exist, at least affordably, until very recently.

Note

A CPU (central processing unit) is the “brain” that powers

all the necessary functions of a computer. A GPU (graphics
processing unit) is another specialized electronic circuit that
accelerates the rendering of images and videos. AI needs
 CPUs to sift through and process all the data it is fed and
GPUs to create AI-generated images.

Finally, all those computers need electricity to run—and AI requires vast

amounts of electricity. Running an AI model is a very large-scale operation,
one that requires large amounts of capital to finance. (Microsoft just
recently signed a deal to reopen the Three Mile Island nuclear power plant
to power its AI efforts.) Again, until recently, only a very few large
companies could afford to create and maintain their own AI models.

Who Are the AI Power Players?

Given AI’s huge need for resources of all kinds, it’s not surprising that some
of the biggest players in AI are also some of the biggest tech titans today—
companies that have the financial and technical resources to create large-
scale AI engines. These power players include

Alphabet (Google), which is heavily invested in using AI in their search

and advertising services
Apple, which is interested in incorporating AI in its operating system and
applications
Amazon, which, through its Amazon Web Services (AWS) division,
hosts many of the large AI models
 Anthropic, founded by former members of OpenAI and backed by
Amazon, with the goal of researching the safety and reliability of AI
systems
Baidu, a large Chinese tech company with its fingers in all sorts of
technologies, including AI
IBM, which has decades of AI experience and a huge stake in computer
hardware
Lenovo, a major provider of enterprise storage solutions, necessary to
store all the data necessary for today’s AI models
Meta (Facebook and Instagram), which possesses an enormous amount
of computer and graphics processing power
Microsoft, which, like Apple, is interested in incorporating AI in its
operating system and applications, including all the Office apps
Nvidia, which manufactures some of the most popular GPUs used in AI
processing
OpenAI, an AI development company partially funded by and partnered
with Microsoft
Oracle, whose cloud infrastructure is in high demand by other AI
companies

Because of the high costs involved, these big tech players have a huge
advantage over smaller players. A smaller company may be able to develop
a unique AI model, but it still has to rely on these larger companies to host
and power that model. In today’s world of AI, the big tech titans have a
substantial advantage.

How to Use Today’s AI Generators

An AI generator is a tool—an app or a website—that uses generative AI to

create new content based on user input. These generators are typically
specialized in terms of the types of content they generate: some generate
text, others images, still others music.

I get more into all-purpose AI generators in Chapter 3, but looking at these

services is a good way to understand how AI works. They’ll probably be the
primary way most people interact with AI for the near future.

To give you an idea of how to use an AI generator, I’ll use ChatGPT as an

example. It’s the AI engine from OpenAI, one of the largest AI research
companies. (Other AI tools such as Anthropic’s Claude, Google Gemini,
and Microsoft Copilot work in a similar way.) ChatGPT generates text
responses to user prompts and functions kind of like an online chatbot. You
can use ChatGPT to answer questions, provide information, and write
things for you—letters, reports, articles, even poetry.

As you can see in Figure 1.3, the main ChatGPT page looks a little like a
search engine query page. When you register (for free) and sign in, your
previous prompts are listed in the left column. The prompt box is at the
bottom of the page.

FIGURE 1.3

The ChatGPT AI generator

To use ChatGPT, all you have to do is enter a prompt—that is, a query or

suggestion—into the Message ChatGPT box and press Enter. As you can
see in Figure 1.4, ChatGPT then does what you ask it to do, generating its
results above the prompt box. You can continue the conversation by
entering additional prompts; your prompts and ChatGPT’s responses are
displayed one after another, just like a conversation.
FIGURE 1.4

A typical ChatGPT prompt and response

It’s really as simple as that. The more detailed your prompt, the more
precise ChatGPT’s response. You can specify the length of what you want,
the style, even specific details. (Figure 1.5 shows a very detailed prompt
and ChatGPT’s response.)
FIGURE 1.5

Asking ChatGPT to write a newspaper article

Other AI generators work in a similar fashion—although customized for

different types of output. For example, an AI image generator, such as
DeepAI’s AI Image Generator (shown in Figure 1.6), typically offers
options for different types of images and styles (photographic, illustrative,
and so on).
FIGURE 1.6

The prompt panel of DeepAI’s AI Image Generator

Note

Find out more about using AI generators and other popular

AI tools—many of which are free—in Chapter 3, “Getting
Started with All-Purpose AI Tools.”
How You Can Use Generative AI In Your Daily Life

As you can see, generative AI is going to be an important technology for

lots of people and businesses going forward. But what does it matter to you,
personally?

The reality is that generative AI has applications across a wide variety of

activities and industries. It’s likely to impact your home life and the
activities in which you participate. It is also likely to impact your job and
the jobs of people around you. It’s going to be virtually everywhere, in one
form or another.

Let’s take a quick look at how AI will impact various industries and
activities.

Warning

You can’t use AI-generated content for all purposes. Chapter 2, “The
Benefits and Risks of AI,” goes into when you can use AI-generated content
and when it’s not a good idea. For instance, you wouldn’t claim AI-
generated content as your own.
AI in Art

Amateurs and professionals alike are already using AI to create new

artwork. AI can create photorealistic images to accompany newspaper and
media stories based merely on text instructions. It can be used to create art
in comic books and magazines. It can be used by fans to create artwork of
their favorite stars and fictional characters.

Unfortunately, generative AI can also be used to create deepfakes and

manipulated images meant to fool and confuse the public. We’re already
seeing AI-generated deepfake pornographic images of celebrities and
deepfake images of politicians in campaign ads. As generative AI keeps
getting better, it will become more difficult to distinguish what images are
real and which are AI-generated.

Note

Learn more about AI art in Chapter 7, “Using AI to Create

Art and Images.”
AI in Business

Many businesses are using generative AI to make their operations more

efficient and productive. AI is already being incorporated into all manner of
business processes and used to design new and refine existing processes.
Customer support operations are using AI by providing 24/7 AI chatbots to
answer customer questions. AI can assist in business analysis and provide
more informed input on decision-making. AI is already being and will
continue to be used to make businesses more competitive.

For workers, AI is making it easier to do a lot of different jobs. Repetitive

tasks are being automated and even higher-level tasks are being made easier
with AI assistance. For many tasks, workers are using AI to get a head start
and take care of some of the drudgework involved. It’s changing the nature
of work for many workers.

Note

Learn more about AI in business in Chapter 9, “Using AI at

Work.”
AI in Finance and Banking

Consumers are using generative AI to get personalized financial advice.

They use it to monitor their personal financial transactions. They can also
use AI to maximize the return on their investments and improve their credit
ratings.

Generative AI also has many applications for banks and other financial
institutions. Some financial institutions are using AI to evaluate your loan
risk and identify or prevent fraudulent transactions. They’re starting to use
AI to automate backend operations and reduce human error. Like other
businesses, they’re using generative AI to provide highly detailed customer
service via 24/7 chatbots. They’re also using AI to ensure they comply with
complex governmental and industry financial regulations.

AI in Gaming

Generative AI is changing the world of computer games. Generative AI can

create and already is creating realistic gaming environments for today’s
most advanced computer games. It can create complex characters and
animations. It can develop detailed gameplay and create immersive
environments.

Generative AI is capable of creating more sophisticated graphics and visual

effects than are currently available. It can also develop plotlines and
enhance interactions between human players and in-game characters—all of
which greatly improves gameplay and benefits gamers at all levels.

AI in Healthcare

AI is already having a significant impact on our health and wellness—both

directly to individuals and through improvements in the healthcare industry.

For individuals, generative AI can help you better understand your

conditions and your doctor’s instructions. It can help you create
personalized health and exercise plans. It can anticipate the onset of illness
or disease. It can also provide answers to your health-related questions.

For providers and healthcare facilities, generative AI can help more

accurately diagnose illnesses and diseases and improve patient outcomes. It
can help physicians better prescribe and manage medications. It can help
clinics and hospitals better serve their patients in terms of scheduling,
billing, and support. And AI can help the pharmaceutical industry more
quickly develop more and more effective drugs.

Note
 Learn more about AI in healthcare in Chapter 11, “Using AI
for Health and Wellness.”

AI in Writing

If you do any writing at all, even if it’s just the occasional text message,
generative AI can help you write better. With the right prompts, AI can
write letters, blog posts, social media posts, even essays and reports. If
you’re not a writer by nature, AI can make you sound like one.

Note

Learn more about AI and writing in Chapter 4, “Using AI to

Find the Right Words.”

Generative AI can write stories and articles based on available information.

It can determine what readers want to read and provide customized news
feeds. It can perform in-depth research for human journalists. It can
automate formerly manual tasks and free up reporters to do more reporting.
It can help reporters meet their deadlines. It can provide news coverage in
areas that were previously economically unfeasible.
Unfortunately, generative AI also can be and is being used to replace human
journalists at many media outlets. This can lower the quality and
availability of news to many users. And with AI hallucinations and limited
data, you need to verify all AI-generated content before sharing it either
publicly or privately.

AI for Learning

Teachers today are starting to use generative AI to develop their lesson

plans. Trainers are using AI to create realistic training simulations. AI can
also help automate grading and assessing student process. It’s helping
teachers develop personalized learning for individual students—and helping
students better learn difficult material.

Unfortunately, generative AI can also help students cheat. It can write

papers for students that read and sound just like something the student
wrote—or even better. It can anticipate and answer test questions. It can
help students get around the system by doing a minimum amount of
original work—all of which is something educators have to figure out how
to deal with.

AI in Marketing and Advertising

Generative AI is accelerating the mass personalization that is driving

today’s marketing industry. Thanks to generative AI, marketers are able to
better target potential customers and develop personalized advertising and
marketing campaigns. This should, at least in theory, improve lead
generation and sales as well as increase customer satisfaction.

Marketers can also use generative AI to create brochures and

advertisements, and script and even produce commercials. Generative AI
can handle a marketing campaign from inception to graphic design to
copywriting to print, web, and video creation, with a minimal amount of
human input.

AI in Music

Commercial customers are already using generative AI to create

background music, music for commercials, music you hear when on hold,
and other similar uses. Unfortunately, unscrupulous producers and
streaming music services are also using AI to create songs and
performances that closely mimic those of popular artists—without paying
any royalties to the original artists.

On the plus side, some innovative musicians are using generative AI to help
them create new music. Some producers and record labels are also using
generative AI to “clean” older recordings to make them sound better for
modern audiences. And AI is not just for professional musicians; some fans
are using generative AI to create music that sounds like and honors the
musicians they love.
 Note

In 2023, producer Giles Martin, son of famed producer

George Martin, worked with Paul McCartney and Ringo
Starr to use AI to clean up and enhance some old, previously
abandoned tracks by the late John Lennon and George
Harrison. The result was the last “new” Beatles song, “Now
and Then.”

AI in Programming and Software Development

Generative AI can write programming code much faster and more

accurately than human developers. This is enabling the release of more
sophisticated computer programs and apps on a faster development cycle.
Generative AI can also write HTML code to develop better and more fully
featured websites. It’s not an understatement to say that AI is shaking up the
developer industry.

AI in Transportation and Travel

The automotive industry is using AI to develop the first true self-driving

car, which promises to revolutionize the trucking and ridesharing industries,
as well as how you drive yourself and your family. AI is also helping
municipalities better manage traffic flow and predict when street and
equipment maintenance is due.

Just as important, AI is helping map apps be more accurate and provide

better driving directions. Real-time AI analysis can help you avoid traffic
slowdowns and find the fastest route to your destination.

Note

Learn more about AI in transportation in Chapter 10, “Using

AI to Manage Your Travel and Transportation.”

AI Everywhere

Bottom line, generative AI is going to be everywhere, whether you

recognize it or not. AI will affect your life in ways you cannot even imagine
—both for better and for worse. That’s why you need to know how AI
works, so you can better use it proactively in your own life.
 Note

Learn more about how AI will affect your life in Chapter 2.

Summary

In this chapter, you learned what AI is (the simulation of human intelligence

in machines) and what it isn’t (simple rules-based processes). You learned
the difference between the older predictive AI and the newer, more creative
generative AI. You also learned the history of AI, dating all the way back to
the days of the ancient Greeks through the birth of modern AI in the 1950s
and today’s explosion of AI development.

In addition, this chapter showed you how AI works by examining the five
steps of the AI process: data collection, data processing, outcomes,
assessments, and adjustments. More important, you learned how AI is
already affecting a variety of industries and activities.

However it works, the reality is that AI is here today and here to stay. It’s
only going to get smarter, more powerful, and more impactful in the months
and years to come.
OceanofPDF.com
2

The Benefits and Risks of AI

IN THIS CHAPTER

Understanding the potential benefits of AI

Understanding the potential risks of AI
How to spot AI in the wild
Debating the ethics of AI-created content
AI and copyright law

AI proponents predict that this developing technology will result in a

bounty of benefits. According to its most enthusiastic supporters, AI will do
everything from improving productivity and enhancing creativity to curing
diseases and halting or even reversing climate change. On the other hand,
naysayers warn of the negative impacts of AI—lost jobs, deepfakes and
media manipulation, and technology that surpasses and even replaces the
human race.

Which of these AI-influenced futures is most likely? Is AI all rainbows and

unicorns, or do we need to prepare for war against amoral AI-powered
terminators?

The reality is that AI promises some very real benefits but also brings with
it some very real risks and issues. We need to know enough about what AI
is and how it works to embrace its likely benefits while guarding against its
possible downsides.

Understanding the Potential Benefits of AI

You’ve no doubt seen some of the effects of AI in your everyday life. Some
of what AI does will benefit you. Some might negatively impact you. Some
won’t affect you at all. It all depends on what directions AI takes and, to a
large degree, the things you do from day to day—the type of work you do,
the entertainment you choose, and more.

Let’s start by looking at some of the potential benefits of AI, of which there
are likely to be a plethora.

Automating Boring Manual Processes

One thing that AI does very well is figure out how to do repetitive work and
do it quickly, without complaining or needing a break. If you do this type of
work, you may find AI stepping in to do it in your place—which may or
may not be a good thing.

AI can do many boring, repetitive tasks much better than humans; AI is

faster and more accurate than any human being could hope to be. AI also
doesn’t get bored or need to take bathroom breaks. In many ways, AI is the
perfect worker for activities like proofreading documents, filing contracts,
and comparing invoices. AI-based systems can also pack boxes, assemble
products, and perform other manual tasks at high speed and with a low
incidence of errors. Businesses will love it.

Improving Business Productivity

Speaking of businesses, since AI can do repetitive tasks faster and more

accurately than human beings, it is improving efficiency and productivity
across a wide range of industries. AI, once it gets going, should also be able
to figure out new and better ways to do many of these tasks. It’s not just a
rule follower but a reinventor of rules as it learns from what it does.

The reality is that AI is already benefiting businesses in a number of

important ways. Here are some examples:

Streamlining existing processes

Devising and implementing new processes
Eliminating human errors
Improving workplace safety
Reducing the need for human labor, especially for repetitive tasks
Producing more targeted marketing campaigns
Sorting, filtering, and organizing data
Analyzing data and providing detailed insights
Improving decision making
Designing new products and services
All these functions increase productivity and reduce costs, which makes AI
highly attractive to the business community.

Note

Learn more about using AI in business in Chapter 9, “Using

AI at Work.”

Reducing Risk

Consider all the activities that place people at risk—working with

hazardous materials, operating heavy machinery, working in harsh
environments, and more. By merging AI and robotics technologies, humans
may no longer need to go deep underground or to unsafe heights, defuse
bombs, or change lightbulbs on tall cellular towers. Autonomous robots,
powered by AI, can do the job safely and keep humans out of harm’s way.

Making More Things Available 24/7

Unlike human beings, AI systems don’t need to eat, sleep, or take breaks.
That means they’re available 24/7, even on holidays. Even better, AI
systems remain at peak capability and capacity over that entire period; they
don’t experience “peak productivity” because they’re always working at
their max.

That also means that AI-powered systems can deliver service at all hours of
the day and night. Instead of closing a call center after a certain hour, an AI-
powered call center is available to take customer calls every minute of
every hour. This means businesses effectively never need to close, plus they
can now serve customers around the globe, no matter when that might be in
local time.

Personalizing the User Experience

AI will further enable the trend of mass personalization. Imagine an AI

system analyzing your past browsing patterns and using that to create in
real time a personal home page on your favorite shopping site, with deals
created in the moment just for you. AI will know what colors you prefer,
how big the type needs to be, what kind of products you’re interested in,
and more, and use that information to create a totally unique and
personalized shopping experience.

Enabling Better Recommendations

Adjacent to the creation of AI-powered user experiences, expect AI systems

at your favorite streaming video and audio sites to make better and more
personalized recommendations for you. You’re used to Amazon, Disney+,
Netflix, or recommending shows based on your past viewing habits; in most
instances, those recommendations are based on shared actors or genres.
With AI, providers can go beyond that by analyzing a whole range of
behaviors to determine seemingly unrelated shows or playlists you’re apt to
like. Thanks to AI, your favorite streaming service will know what you like
and dislike—maybe better than you do!

Improving Digital Assistants

The use of digital assistants has been increasing in recent years, both in the
home (such as Amazon’s Alexa, Apple’s Siri, and the Google Assistant) and
in businesses. To date, those assistants have not been very “smart,” typically
capable only of answering basic questions and performing basic tasks on
command.

Integrating advanced AI technology into these systems will make them

significantly more useful. Imagine a not-too-distant version of Alexa that
knows when you come home, can sense what kind of mood you’re in, and
selects the proper music for your mood. Or it knows that you’re interested
in a particular news story or topic and feeds you new stories as they
develop.

Note
 As I write this, Apple just announced that they’re
incorporating OpenAI into their Siri digital assistant and
Amazon is rumored to soon announce a subscription-only
version of Alexa that is AI-powered. The merging of AI with
now-decades-old digital assistant technology is happening.

On the business front, envision a chat assistant in the bottom corner of a

website that doesn’t just provide canned answers to common questions but
can actually respond to questions in real time, just like a real human support
person. Customers won’t be able to tell whether they’re conversing with a
human or a chatbot, which benefits both the business and its customers.

Managing Messages

If an AI-powered chatbot can interact with users in real time, why not
employ personal chatbots to answer your phone calls, email messages, and
texts? This is particularly appealing if you’re inundated with messages
either at home or at work; let an AI assistant manage all your inboxes,
respond to those messages it can, and route to you only those messages to
which you need to personally respond. It will be like having your own
personal assistant on your computer or phone.
Improving Healthcare

The medical field is employing AI in a number of different ways, all

beneficial to patients.

Physicians benefit from AI systems that ingest relevant patient information,

compare it to existing data (including patient records, lab results, and data
from clinical trials), and make instant and accurate diagnoses. The AI
models can also devise extremely detailed and personalized treatment plans
that have more positive health outcomes for patients.

Physicians can also use AI to perform delicate procedures where even the
slightest error could be life-threatening. When programmed correctly and
allowed to learn over time, an AI-powered robotic surgical system can be
more precise than a surgeon and virtually error-free.

(And in case you think this latter application is a futuristic dream, know that
the Smart Tissue Autonomous Robot—STAR—has already performed
laparoscopic surgery without any human guidance whatsoever.)

For individuals, AI technology can help people better manage their health
and wellness, as well as understand physician advice and instructions. AI
can also catch errors that might develop when treatment spans multiple
physicians or clinics, such as multiple prescriptions that shouldn’t be taken
with one another. That sort of coordination is difficult with traditional
manual systems but is a snap for AI-based systems.
 Note

Learn more about using AI in healthcare in Chapter 11,

“Using AI for Health and Wellness.”

Enhancing Learning

AI is helping students learn more and learn faster by developing learning

plans personalized to each individual. Consider the vast amounts of data
available on student performance and learning styles and how AI can
analyze and extract actionable insights to improve learning efficiency.
Thanks to AI-based learning systems, students can learn at their own pace
and focus on those areas where they need more work.

Augmenting Creativity

Today’s AI models are already capable of creating short stories, poetry,

images, and songs. Many media sites are using AI to write news stories and
blog posts. Fans and some artists are using AI to create new artwork.
Companies are using AI to create pleasant-sounding background music for
phone systems and commercials.
Going forward, AI-created art will get better—more human-like and more
creative. Writers, artists, and musicians are learning how to use AI to assist
them in their work, generating new ideas, exploring new styles, and
developing a new type of creative process that merges human creativity
with machine output.

Note

Learn more about using AI for writing in Chapter 4, “Using

AI to Find the Right Words.” Learn more about using AI for
art in Chapter 7, “Using AI to Create Art and Images.”

Making Your Life Easier

Here’s the bottom line about all these current and potential AI benefits: AI
will make all our lives easier. We’ll become more productive, more
effective, and less bothered by menial tasks. Our lives will be more
enjoyable and safer. AI is already changing our world for the better and will
continue to do so, in ways we cannot fully imagine. But risks abound.
Understanding the Potential Risks of AI

Just as AI promises numerous potential benefits, it also comes with risk.

Will AI really make your life better—or will it negatively impact your life
and livelihood? Again, it all depends on what you do and how AI develops.

AI Can Spread Misinformation

AI is just a tool. It does what users ask it to do. If someone prompts an AI

system to produce a picture of a giraffe flying a helicopter, AI will do it.
The initiators of that prompt can then use that image in whatever way they
want.

Thus, we come to a real-world issue concerning the use of AI content:

mischievous or malicious individuals can use AI to create blatant
falsehoods, either in words or visuals, and then spread those falsehoods
over social media and other channels. AI-generated text, images, and videos
can be particularly convincing, especially as AI models continue to
improve. If someone wants to convince people that a giraffe can fly a
helicopter, a photorealistic image of that scenario can be very persuasive.

Realistic but false text, images, audio, and videos are called deepfakes. In
the past, people have used image editing programs such as Adobe
Photoshop to manually manipulate images and create deepfakes. Today, AI
image generators can do the job better and faster with just a few simple
prompts.
Say, for example, you wanted to conduct a smear campaign against a
neighbor you don’t particularly like. You can feed an AI image generator a
picture of your neighbor and prompt it to create a photorealistic image of
that neighbor burning trash on their lawn. You could then take this very
real-looking picture to your neighborhood association and try to get your
neighbor in trouble.

In this same fashion, celebrity deepfakes are popular. Take a popular

actress, tell the AI image generator to create a picture of said actress sans
clothing, and—voilà!—you have a ready-made pornographic image ready
for distribution on the Internet.

This sort of AI-powered manipulation can also be used for political and
propaganda purposes. We’ve already seen deepfake photos purporting to
show people doing things they didn’t really do, deepfake videos purporting
to show events that didn’t really happen, and deepfake phone calls
impersonating politicians saying things they didn’t really say. AI can make
these deepfakes extremely convincing—so convincing that voters could be
swayed to change their vote from one candidate to another.

Note
 AI-created deepfakes make it difficult for people to
determine what’s real and what’s not. To learn how to spot
faked AI content, read ahead to the “How to Spot AI in the
Wild” section later in this chapter.

In addition, AI can be used to spread false information over social media.

Most social media platforms use AI algorithms to decide what shows up in
users’ feeds. Manipulate that algorithm just a bit, and you can fill peoples’
feeds with political falsehoods and biased viewpoints. That could be
dangerous for any functioning democracy.

Even though the Federal Communications Commission (FCC) recently

outlawed AI-generated political robocalls and major tech companies signed
an accord to prevent AI from being used to disrupt elections, AI-powered
misinformation remains a major short-term and long-term threat to elections
in the United States and abroad. FBI Director Christopher Wray recently
warned about foreign adversaries using AI technology to influence U.S.
elections, saying that AI makes it “easier for both more and less-
sophisticated foreign adversaries to engage in malign influence.”

If nothing else, the threat of AI-generated deepfakes could cause people to

question even legitimate stories and images. If you can’t tell what’s real and
what’s fake, what can you believe?
AI Can Be Biased

AI results are based on the data fed into large language models. The more
data, the better the results.

Equally important to the quantity of data available is the quality of that

data. In AI, as in most things, it’s a garbage in, garbage out type of
situation. Bad data will result in unreliable AI models.

Because AI relies on the data it’s fed, biased data can be a significant
problem. Remember, most AI models get data by scraping content from the
Internet. Unfortunately, there’s a lot of flawed or biased content on the
Internet, and those characteristics can be absorbed into an AI model.

Bias can infiltrate AI systems in a number of ways. AI can ingest training

data that reflects historical or social prejudices. It can include training data
that includes biased human decisions and use data that either over- or
underrepresents specific groups, thus reinforcing existing biases. It can even
treat opinions or obvious jokes in the training data the same as it does hard
facts.

In addition, AI can exhibit the biases of the people who develop its
algorithms. As AI researcher Olga Russakovsky notes, “AI researchers are
primarily people who are male, who come from certain racial
demographics, who grew up in high socioeconomic areas, primarily people
without disabilities.” That creates a very specific worldview that is, to some
extent, exhibited in AI output.

For all these reasons, AI content today often exhibits the same biases that
exist in our society at large. Without conscious upfront programming, AI is
likely to perpetuate those biases in the decisions it makes.

Consider, as an example, a company’s use of AI to vet job candidates. If an

AI model reflects a society’s bias against specific ethnic or racial groups, a
company will continue to hire fewer people from those groups.

Similarly, AI-powered speech recognition software can fail to understand

certain accents and dialects because the generally white male researchers
don’t speak that way. This can cause problems for an AI chatbot trying to
understand or respond to questions from customers of certain backgrounds.

As another example, consider AI-generated images. Given the gender and

race bias present in today’s AI systems, if you ask an AI tool to create an
image of a businessperson, what are the odds that it will show a white male
and not, perhaps, a female of color? (Pretty good; see Figure 2.1, the result
of a single such prompt with DeepAI’s AI Image Generator.)
FIGURE 2.1

The stereotypical image generated by the prompt, photo of a businessperson. (Image generated by
DeepAI’s AI Image Generator.)

AI models can also reinforce society’s historical age bias, especially in

employment. This is particularly concerning given the use of AI-powered
recruitment systems; if the algorithms used by these systems are biased
toward younger candidates, older job candidates may be unfairly excluded
from consideration for some jobs.

Guarding against all forms of bias is essential in creating trustworthy AI

content. Otherwise, AI will increasingly exhibit those biases, both good and
bad, that exist in our society today.
AI Can Invade Your Privacy

Here’s a major concern of AI critics today: AI is a major threat to our

personal privacy, and they will only grow.

Think back to Chapter 1, “Artificial Intelligence: What It Is and How It

Works.” Do you remember where and how AI models get all the data they
use to train and learn?

The answer is that AI gets its data from you and me and everyone around
us. Most of the data that’s been fed into AI large language models has been
scraped off the public Internet. That means not only website content but
also social media posts, online messages, and other communications
between unsuspecting individuals.

That’s right, your friendly neighborhood AI model is based at least in part

on your own thoughts and words, as well as those of your friends and
neighbors. And the models have obtained your information without asking
you permission. If it’s out there on the Internet, the thinking goes, it’s free
for the taking. It might be different if the content is behind a paywall or on a
private site that requires registration or permission. But anything that’s out
there publicly, the AI companies say, is there for the taking.

That includes content that isn’t actually on the Internet but has been
supplied online—conversations you’ve had with chatbots, chats you’ve had
with friends and family, and questions you’ve asked on support sites.
To be fair, AI large language models don’t target your individual data per
se; instead, they incorporate it and data from millions of other people into
their models. It’s not using your data against you to do harm, as a scammer
would, but it’s still using your data without your permission.

Note

Separate from the ingestion of large amounts of personal

data for generative AI, predictive AI often uses your
individual data to make personal recommendations. That’s a
different issue, and one to which you probably consented
(via terms of service) when you signed up for a given
streaming service or online store.

Despite numerous data privacy laws on the books, few if any such
regulations explicitly protect your data privacy from AI. Some regulations
have been proposed, but neither the United States nor the EU have enacted
laws that cover AI’s use of personal data. That leaves your data out there for
the taking by any AI large language model that wants to use it.

That’s not a good thing.

AI Will Replace Some Jobs

With every new technological age comes some degree of change and
displacement. The industrial age eliminated many formerly manual factory
jobs. The automotive age displaced workers in the horse and buggy
industry. The rise of the Internet resulted in job losses in traditional media
and communications companies. This sort of change is inevitable.

Don’t be surprised, then, if the biggest near-term impact of artificial

intelligence technology is a significant loss of jobs. Companies large and
small are looking to AI to help them improve their productivity, which
means replacing expensive human workers with cheaper, less-demanding,
more efficient AI systems. For these companies, “improved productivity”
means fewer employees, which means layoffs—sometimes for even the
most seasoned workers.

Now, many employers will couch this scenario as letting AI take over
repetitive jobs so they can “repurpose” employees to higher-value tasks.
While that is a possible scenario, it’s equally possible and perhaps more
probable that many employees displaced by AI either won’t have the
necessary skills for those higher-value jobs or that those jobs won’t exist at
all. While the impact of AI will differ from company to company (and
industry to industry), it’s likely to have a net negative impact on the human
workforce.
What industries will be most impacted by the AI revolution? AI is likely to
have an effect across the board, but in particular, anticipate job losses in the
following sectors:

Agriculture, with AI-powered robots automating many manual tasks,

such as planting and harvesting, especially on larger farms
Finance and banking, with AI automating both customer-facing and
back-office jobs
Healthcare, with AI assisting or replacing many scheduling and back-
office functions
Legal services, with AI taking over contract generation and management
Manufacturing, with AI-powered robots replacing factory workers
News media, with newspapers and websites using AI to generate articles
and posts
Transportation, with self-driving vehicles eliminating human drivers to
transform the trucking and rideshare industries

How big will this AI-powered job disruption be? Goldman Sachs estimates
that generative AI could eventually replace up to 300 million jobs
worldwide, with many occupations experiencing a 25 to 50 percent job loss.
This would be a huge disruption to the job market—and to the way of life
for hundreds of millions of workers.

Unlike the industrial revolution, which primarily impacted manual or blue-

collar workers, the AI revolution is likely to also affect higher-paid white
collar workers. That will be a major difference from previous technology-
based changes and a big concern for skilled workers everywhere.

It’s not all doom and gloom, however. On the plus side, Goldman Sachs
predicts that artists, computer system analysts, HR managers, legal
professionals, mental health professionals, surgeons, teachers, writers, and
those in leadership roles are less likely to be replaced by AI because of the
need for human judgment and creativity in those roles. In addition, there is
already a huge demand for jobs programming and training AI systems.

Will your job be one of those lost to AI? Perhaps, and even if not, many of
your coworkers will be impacted. Prepare to be disrupted.

Note

I take issue with Goldman Sachs’ prediction that AI will not

affect writers. I used to write blog posts for a number of
websites, up to a dozen or so posts a month. In recent
months, that work has dwindled to next to nothing. It seems
that many of the companies for which I used to write wanted
lower-cost content and AI filled the bill. Instead of paying
for my expert writing, they opted to go with AI’s free
content, which is apparently good enough for their purposes.
 My side writing income has pretty much evaporated, and AI
is to blame. (Thank you, AI!)

AI Will Make Mistakes

AI constantly makes mistakes—or in AI parlance, hallucinates. We all must

verify its output. If we try to rely too much on AI, especially for mission-
critical tasks, we will be disappointed when things go wrong, which they
will. Witness AI-powered self-driving cars that get into accidents because of
faulty or less-intelligent AI systems, or AI image generators that give
people six fingers and a missing ear.

Today’s AI systems will make mistakes. Putting all our trust in said
systems, at least at this point in time, is ill advised. If you rely completely
on AI and AI isn’t perfect, the decisions you make based on that AI may be
flawed. Likewise, if you’re using AI to manage operations or systems, you
may experience system interruptions if AI gets some of the data wrong.

AI will get better and more reliable, but it’s not there yet, wishful thinking
aside.

AI Uses Significant Resources

AI is a resource- and power-hungry technology. Today’s increasingly larger

AI models require vast amounts of power, both in terms of electricity and
computing power. AI models need many fast and powerful CPUs and
GPUs, vast amounts of data storage capacity, fast and reliable Internet
connections, and lots and lots of electricity to run it all.

Unfortunately, none of these items are cheap or limitless. AI is an expensive

technology, which is why so many large models are the provenance of
today’s large tech titans, such as Amazon and Microsoft, that have the
financial and other means to pull it off.

Looking just at AI’s electricity needs, one expert calculated that by 2027,
the AI sector will consume between 85 and 135 terawatt hours per year. To
put that in perspective, that’s about half a percent of all global electricity
consumption. That’s massive—and increasing daily.

At some point there may not be enough available resources to power all the
AI systems currently being developed. What do AI companies do if there’s
a chip shortage or a lack of storage or not enough electricity to go around?
Or, equally likely, if the costs of these resources rise to unaffordable levels?
The ability of the AI industry to grow may be constrained by resource
availability and pricing.

In addition, all the resources that power AI have a major impact on the
environment. The energy usage alone contributes significantly to fossil fuel
usage and the resulting climate change. AI is not in the least bit
environmentally friendly.
 Note

These are just the known risks of today’s AI. Experts have
additional reservations about AI’s future impact, which you
can read more about in Chapter 13, “The Future of AI.”

Who Is Legally Responsible for AI?

Here’s an interesting question. In the event of harm caused by AI—a self-

driving car that runs amok and kills a pedestrian, for example—who is
legally responsible? Is it even possible to hold AI systems responsible for
their actions?

This is a complex question where the answer is far from clear. Do we hold
responsible the company that developed the AI model? Or the company that
sold that particular AI? Or the one that used it in their application (such as
the automotive manufacturer that built and sold the self-driving car)? Or do
we hold the AI’s programmer responsible? Or is it the fault of the individual
using that AI?

Establishing the legal liabilities of AI systems is imperative but challenging.

When something goes wrong with such a system, some person or some
entity needs to be held responsible—doesn’t it? Can we hold a computer
algorithm legally responsible for its actions? Who pays the price when
something goes wrong?

Legal experts have been wrangling about this one for some time now with
no apparent conclusions. While I have no idea how this one will shake out, I
do predict that a lot of legal hours will be billed over this issue.

How to Spot AI in the Wild

I talked earlier in this chapter about how AI can be and is being used today
to write text, create images, and more. Unfortunately, that means that AI can
be used to write stories that spread lies and propaganda or create images
that represent scenes that never took place. The AI-generated content looks
so real and reads so naturally that it’s difficult to tell it from the real thing.

How, then, can you keep from being fooled by fake content? How can you
identify AI in the wild?

How to Identify AI-Written Text

Many, many companies are using AI to create content for their blogs, social
media feeds, and websites. AI content is increasingly popular because it’s
free or low cost; companies don’t have to pay a real human writer to write
it.
AI-written content can sometimes be good. It can appear well-written. It
can be informative.

On the other hand, AI-written content can sometimes be poorly written; it

can read just a little “off.” And, worst-case scenario, it can contain incorrect
information or even deliberate misinformation.

When examining written content, you find on the web or elsewhere, look
for these signs that it may be AI-generated:

Incorrect information: AI systems are often trained on limited data sets

that don’t contain precise information. In this scenario, AI might “guess”
at an answer and guess incorrectly. In addition, misinformation
deliberately spreads falsehoods that are often easily identified.
Controversial opinions presented as facts: While human propagandists
can also couch falsehoods as facts, AI systems do it better. Malicious
actors are using AI to spread misinformation, so if you see something
especially controversial, question its authorship.
Outdated information: AI systems are often trained on data sets that
contain older information. References in a given article that are several
years old could signal that it may be AI-generated from information that
isn’t current.
Repetitive words and phrases: Many AI models have a limited
vocabulary and continually reuse those words and phrases they know.
 Tone of voice: AI writing tends to be drier and, dare I say, more robotic.
Humans tend to write more conversationally and informally.

All that said, today’s most advanced AI models can write text that is
difficult to distinguish from human-written text. Because of these
advancements, it’s becoming more and more difficult to determine which
text is written by a human and which by AI. If you can’t tell the difference
between AI and human text, you’re not alone.

Then how can you identify text that was generated by AI? The best piece of
advice is to trust your instincts. If a piece of information doesn’t seem right,
you should look for another source—ideally one you can trust. Always go
with a trusted source over a random article from an unknown website.

Note

Some companies, including OpenAI, are working on

embedding watermarks in their AI-generated text, typically
in the form of specific word patterns. This will help experts
better identify text that was written by AI systems.
AI Text Detectors

Want to find out if a given article or piece of text is AI-generated? Several

websites offer AI text detection functionality. They examine various
features of a piece of text and determine, with a fair amount of accuracy,
whether that text was generated by a human being or an AI engine. The
most popular of these AI text detectors include

AI Text Classified (https://freeaitextclassifier.com)

Copyleaks (https://copyleaks.com/ai-content-detector)
QuillBot (https://quillbot.com/ai-content-detector)
Scribbr (www.scribbr.com/ai-detector)
ZeroGPT (www.zerogpt.com)

As I explain later in this chapter, many educators are using these tools to
help detect AI-generated homework and papers from their less-than-
completely-honest students. You can use them the same way.

How to Identify AI-Generated Photos and Images

AI-generated images are often more dangerous than AI-generated text. It’s
the old adage of a picture being worth a thousand words; we tend to believe
things we can see with our own eyes—even if those images are
manipulated.
Some individuals create AI-generated images for their own amusement.
Some create AI-generated images for profit. And some malicious actors
create AI-generated images to try to fool people like you and me.

Fake AI-generated images can be used to try to influence public opinion.

They can be used to try to influence voters in an election. They can be used
just to confuse people and stir up trouble.

Whatever the goal, we collectively need to be aware of these fake images

and do our best to identify and avoid them. Given the highly realistic
quality of some of these images, however, that can be difficult.

How, then, can you tell an AI-generated image from a real one? Here are
some tips:

Too many—or not enough—fingers: For some reason, today’s AI

image generators have problems with human hands. Some AI generated
images show people with four fingers, some with six, and some, like the
one in Figure 2.2, with some bizarre arrangement of fingers that bear no
resemblance to reality. Always check the hands and fingers (and other
limbs) in a photo; if something’s obviously not right, it’s probably AI
generated.
FIGURE 2.2

A not-very-human looking human hand generated by AI. (Image generated by Microsoft Image
Creator from the prompt “human hands.”)
Unnatural body proportions and parts: Similar to the hands and
fingers thing, AI doesn’t always get body proportions right. Ears might
be slightly misplaced, larger, or smaller than they should be. Faces might
be oddly asymmetrical. Legs and arms might be thicker or thinner than is
physically possible. If it doesn’t look right, it’s probably been poorly
constructed or manipulated.
Bad hair: I’m not talking about messy hair or a bad part. AI often has
trouble with all the details in human hair. Hair might appear blurred or
have unusual changes in texture. It might look wrong on a head, like a
bad wig. Even little issues, like looking too thick or too thin, can be a
sign of AI manipulation.
Overly rendered appearance: AI will sometimes “over render” the
details on a face or other object, making it look unnaturally sharp,
especially compared to other elements or the image background. It’s not
a natural look.
Weird or missing details: AI doesn’t always get the details right. You
can sometimes find surprising errors when you zoom in to a high-
resolution image. Look for things in the background that shouldn’t be
there, distinct elements blending weirdly into one another, items that
don’t quite match up, and other things that don’t make sense. AI’s
generated reality doesn’t always reflect our real world.
Unusual backgrounds: AI-generated images sometimes get foreground
elements right but do odd things to the background. Look for unusual
textures, unnaturally repetitive patterns, or glossy effects. Another
telltale sign is a blurry or airbrushed background.
Poor architecture, furniture, and accessories: AI can make little
mistakes that really stand out, especially with buildings and rooms. Look
for oddly curved walls, sloping ceilings, misaligned steps, and such. A
chair might be missing a leg or a coffee table might have too many legs.
Consider minor elements in the image, such as coffee mugs, purses,
jewelry, and the like; AI often gets the size wrong or has them hanging in
mid-air, defying the law of gravity. It’s almost as if these details are an
afterthought for the AI models.
Nonsense text: Look for any items with text within the image, such as
newspapers, books, and posters. More often than not, AI generates
nonsense words and letters for these elements, as shown in Figure 2.3.

FIGURE 2.3

Look closely at the newspaper; those aren’t real words. Also, the woman has three legs and a
very odd-looking left hand. (Image generated by PIXLR AI Image Generator from the prompt
“woman reading a newspaper”)

Stereotyped images: AI isn’t very creative. Ask it to create a picture of

a doctor, and you’ll get a guy who looks like someone out of central
casting. These stereotypes can also play into racial, gender, and other
 biases. If a picture looks like a stock photo, it may be AI generated. (Or
it could be a stock photo, too.)

Another way to spot AI-generated fakes is to search for similar images

online. A real image is likely to be replicated on other sites, sometimes from
a slightly different angle taken by a different photographer. Especially look
for similar photos on legitimate news sites. If nobody you trust is displaying
a similar photo, it’s probably a fake.

You also need to consider whether the image is believable. If you’re

presented an image of your city’s mayor striking a small child with a cane,
it’s probably a fake. (Unless you happen to be living in England during the
Victorian era, that is.) If it shows a popular TV celebrity being handcuffed
and led to jail, it’s probably a fake. (Well, for most celebrities, anyway.)
And if it shows a well-known person doing something criminal or utterly
embarrassing, it’s probably a fake, too. (Again, depending on the
individual.)

The point is, people and groups are using AI to create images of things that
haven’t really happened and are unlikely to happen. Always consider the
context of an image; if it’s something that doesn’t seem likely, it probably is
fake.
AI Image Detectors

As AI-generated images continue to improve, it’s becoming increasingly

difficult to tell what is AI generated and what is real. For that reason, you
may want to use an AI image detector to do the detecting for you.

AI image detectors examine a variety of factors to determine whether an

image is created by a human or by AI. These sites have a pretty good track
record; if they say it’s likely an image is AI generated, that should be
enough for you to be suspicious, at the very least.

The most popular AI image detectors today include

AI or Not (www.aiornot.com)
Hive Moderation (https://hivemoderation.com/ai-generated-content-
detection)
Illuminarty (https://illuminarty.ai)
Winston AI (https://gowinston.ai)

Some of these image detectors are free; others charge for use.

How to Identify AI-Generated Videos

AI-generated videos suffer from many of the same flaws as AI-generated

images. In addition to the issues discussed in the previous section, be on the
lookout for
Implausible content: If the content of a video seems implausible (a
politician espousing a particularly offensive or unusual position, for
example), it may be an AI-generated fake. Trust your instincts, above all.
Strange shadows or light flickers: AI doesn’t always get the light
source right. It also has trouble keeping the lighting consistent when
there’s movement in the frame, such as a person talking.
Unnatural body language: People don’t always stand or walk right in
AI-generated videos. Look especially for jerky movements instead of
smooth, natural ones.
Changes in facial features: Does a person’s face suddenly change in the
middle of the video? Do they gain or lose moles or skin creases? AI
often has trouble maintaining those kinds of details from shot to shot or
even frame to frame.
Slow or no blinking: For whatever reason, artificially created people do
not always blink right in AI-generated videos. They often blink slowly,
and sometimes they don’t blink at all. It’s a dead giveaway.
Robotic facial movements: Watch the person’s eyes—do they move
naturally or robotically? Do their eyebrows track the way they should?
Does their nose crinkle when they smile? These are all telltale signs that
you’re watching a computer-generated person.
Poor synchronization: Trying to make lips move with spoken words is
a difficult task. Sometimes the speaker’s mouth will get out of sync with
the audio track. Other times the audio gets chopped up to match lip
movement. AI definitely has trouble with this.
 Odd background noises: AI doesn’t always get the right background
sounds for what it’s supposed to be showing. Do the background noises
match the room onscreen or what’s supposed to be happening? A street
scene with no car noises would be a certain fake.

As with AI-generated images, you should also search the Internet for other
instances of this video. If you only see it in your TikTok feed but nowhere
else, it might be suspect.

How to Identify AI-Generated Music

You may not have heard much of it yet, but some companies are using AI to
generate music. This is typically background music, the kind of tunes you
hear in an elevator or when you’re waiting on hold, but it’s still very real—
and often sounds just like other similar music. You may even find AI-
generated selections appearing in your streaming media playlists, designed
to sound similar to the other music you’ve selected.

How, then, can you pick the AI-generated songs out of the batch? Here are a
few things to listen for:

It’s too perfect. Human musicians often make mistakes, no matter how
slight. They bend notes, they rush or drag the tempo, they sound like
human beings. AI-generated music doesn’t include this type of human
touch. If it’s too perfect, it may be AI generated.
It’s too repetitive. AI models create music based on learned patterns. If
a song is too repetitive (and it’s hard to tell, as much music is repetitive
by nature), it could be AI generated.
It’s too formulaic. Similarly, AI-generated music doesn’t (and can’t)
take the creative leaps that human songwriters can. Human music often
takes unexpected directions using unconventional chord progressions,
unusual melodic jumps, and atypical song structure. If it’s too formulaic,
it may be AI generated.
The lyrics don’t sound right. Not all AI-generated songs have lyrics but
those that do often lack coherence or emotion. The rhyme schemes might
be correct, but the words won’t convey the same meaning as those from
a human songwriter.
It doesn’t connect emotionally. Humans know how to use words and
music to convey deep emotions. AI models don’t. As a result, AI-
generated music is often soulless and superficial.
It lacks energy. Computer-generated music might sound professional
but lack the energy that live musicians provide. This is a tough one
because many contemporary musicians today use computer programs to
create music in their home studios. But even those that do find ways to
pump up a performance. AI is kind of low-energy all the time.
It sounds almost but not quite like the original artist. Some
companies are using AI to generate “soundalike” recordings that sound
like a popular performer but aren’t really from that performer. Be
 particularly wary of new music from long-dead artists that just wouldn’t
be possible.

Musicians can generally tell whether a song is from another musician or

was generated by AI. It may be more challenging for casual listeners. At the
end of the day, you have to trust your ears.

How to Identify AI-Generated Propaganda

Finally, let’s examine how you can identify and avoid AI-generated
propaganda in your social media and news feeds.

The key to avoiding this type of AI-generated misinformation is to think

logically. Yes, you can work through all the tips presented in the previous
sections for identifying AI-generated text, images, and videos, but it’s often
easier just to trust your nose. If something smells fishy, it probably is.

In particular, follow this advice:

Consider why someone might be spreading disinformation. There’s

typically a purpose behind the lie—and remember that propaganda,
especially the AI-generated kind, tends to spike around election time and
major news events.
If it sounds too outrageous, it probably is. If that rumor about
foreigners abducting children at the mall was real, it’d be all over the
real news; you wouldn’t have to read about it in a friend’s social feed.
 Check the source. If the so-called news comes from a friend of a
friend’s second cousin, don’t assume it’s real. Same if it comes from an
official-sounding news source that you haven’t heard of before. Check
the major media to see if the news is there.
Consult a fact-checking site. Snopes (www.snopes.com) or
FactCheck.org (www.factcheck.org) can help to debunk misinformation,
rumors, and conspiracy theories.

Summing all this up, there’s a lot of fake or misleading content out there.
That kind of content existed before AI and will only get worse (that is, more
believable) with the help of AI technology. It behooves us all to be highly
skeptical of claims we see online, especially those that come from dubious
sources or are spread via social media. No matter how real something might
look or sound, there’s a good chance it isn’t.

Most important, don’t share anything you think might be misinformation.

It’s easy to graze through a provocative post and quickly click the share
button. Take your time, determine whether it’s true, and don’t share it until
you’re sure. You want to be part of the solution, not part of the problem.

Debating the Ethics of AI-Created Content

The use of AI-generated content brings up a number of ethical concerns.

When is it okay to share AI-generated content? When is it okay to use it for
your own personal use? Can you use it for schoolwork? If you create it
yourself (with an AI generator), do you own it? If you found it online, who
owns it?

The rest of this chapter examines the ethics of AI-generated content. The
issues aren’t always clear.

Sharing AI-Generated Content

How ethical is it to share AI-generated content? It depends on how it’s

presented.

In no instance should you present AI-generated content as anything other

than what it is. Do not present AI content as something you created
yourself. Do not present it as “official” content from another source. Do not
try to pass it off as anything other than AI-generated content.

For example, it is unethical to use AI to generate artwork and tell people

you created it yourself. You would be lying about the creation and
ownership of that artwork. If you share said artwork, you need to clearly
label it as AI generated.

Similarly, you should not claim authorship of any text that AI writes for
you. You didn’t write it so you shouldn’t label yourself as the author. The
content—whether an article, blog post, white paper, or book—should be
clearly labeled as AI generated.
Bottom line: You need to be transparent about how you use AI. Claiming
AI-generated content as your own creation is highly unethical. (It’s also
unfair to the original creators of the content used to train the AI engine.)

Using AI for Schoolwork

Schoolwork is a special use case for AI-generated content. If you’re a

student, consider how tempting it may be to have an AI answer homework
questions or even write your term paper for you. What could be easier than
that—just type a detailed prompt and ensure yourself a decent grade!

The problem is, using AI to do your schoolwork is extremely unethical.

And, if caught, you’ll probably have that grade thrown out—and you may
be thrown out of school along with it. Educators view this unauthorized use
of AI as cheating.

By the way, if you think your teachers won’t be able to tell if you’re using
AI to do your schoolwork, think again. As I mentioned earlier in this
chapter, teachers can use numerous tools to identify the original source of
any given passage and determine if it came from an AI engine. Figure 2.4
shows an example. So don’t bank on flying under the radar; if you’re using
AI to cheat on your schoolwork, there’s a good chance you will be caught.
FIGURE 2.4

Determining whether text was generated by AI or by a human with GPTZero. (Image courtesy
GPTZero.)

Just as you shouldn’t use Wikipedia as a source for your schoolwork, you
also shouldn’t use AI to do your schoolwork for you. You may be able to
use it to get ideas, much as you get ideas from searching the web with
Google or Bing, but you shouldn’t have it finish your actual work. You have
to answer those questions and write those papers yourself, without any
outside help. Doing otherwise is cheating—and could lead to serious
ramifications if you’re caught.
Handling Plagiarism

It’s clearly unethical to plagiarize another person’s work. But is using AI-
generated content plagiarism?

The quick answer is, not necessarily. AI engines generate their content not
from a single source but rather based on a wide variety of inputs, ideally
synthesizing content and ideas from all those inputs.

However, AI models can inadvertently copy existing content verbatim. In

this instance, the supposed AI-generated content is nothing more than that:
original content regurgitated. And that is plagiarism.

How can you protect yourself from plagiarizing based on AI-generated

content? The easy solution is not to rely exclusively on AI content. Use it
for ideas but do your own writing—or, at the very least, rewrite the content
that AI generates. And if you want to be sure that you’re using content that
isn’t plagiarized, feed the text into one of the AI text detection tools
discussed previously to find out.

Using AI Content for Malicious Purposes

As you’ve no doubt realized, AI can be used for good or for evil. To be

ethical, you should strive to use AI purely for good purposes and avoid
doing bad things with it.
Unfortunately, there are those among us who have no compunction about
using AI to lie, mislead, cheat, and even steal. The use of AI in this fashion
is not just unethical; it’s immoral and often illegal.

You shouldn’t do that.

AI should not be used to spread misinformation, perpetuate sexist or racist

biases, incite social unrest or violence, or otherwise try to trick people, or
incite social unrest or violence. You should not use AI to generate or spread
harmful content.

No universal “code of AI conduct” prohibits AI engines from contributing

to such malicious behavior. AI does not possess a sense of ethics or
morality. AI has all the ethics of a cellphone signal; by itself, it’s a
completely neutral technology. It’s what people do with it that might be
unethical.

I can’t tell you how you should use AI engines and AI-generated content.
No doubt there are some that will choose to use it for malicious purposes.
All I can do is urge you not to. It’s the right thing to do.

Dealing with Built-In Bias

As you recall from earlier in this chapter, AI systems often have built-in
biases, derived either from the data they learned from or the people who
programmed them. It’s important to recognize that these biases exist and be
prepared to identify and deal with them.

In particular, you need to recognize the potential for bias in the content
generated by AI models. Don’t automatically trust what an AI model tells
you as being the absolute unbiased truth. Learn to examine AI output to
identify any built-in bias, and then edit that content to minimize the effects
of that bias.

AI companies can mitigate the amount of bias in their systems by choosing

datasets that are less biased or totally unbiased and include information
from diverse and representative sources. Future AI models might be trained
to identify bias and filter out that data.

Until then, however, it is on consumers of AI data to recognize the potential

for bias and deal with it when found. We should also carefully consider the
prompts we use to generate AI content and avoid any potential bias in the
instructions we provide. We can’t assume that AI models are color- and
gender-blind; in fact, we may have to assume that they are not.

Dealing with Inaccurate Content

What do you do if what you ask AI to generate is created from content that
is inaccurate or incomplete? AI output is only as good as the data input—
and not all that data is equally robust or accurate.
This means you shouldn’t automatically accept AI content as the final word
on any given subject. Treat it with the same skepticism you’d treat any
other data source and double-check it against other sources.

Here’s an example. I asked ChatGPT if Spenser, the fictional detective

character created by author Robert B. Parker, has a beard. Here’s the
response I received:

Yes, Spenser, the fictional private detective created by Robert B. Parker, is

often described as having a neatly trimmed beard and mustache in many of
the novels. This characteristic is one of the consistent visual details
associated with his appearance throughout the series. The description of his
beard is often mentioned along with other aspects of his appearance, such as
his muscular build and hawk-like features. However, it’s important to note
that specific details of characters’ appearances can sometimes vary slightly
across different books and interpretations.

Unfortunately, this answer is wrong. Spenser is never described as having a

beard; it’s his colleague Hawk that has “a neatly trimmed beard and
mustache.”

I happen to know this because I’ve read all the Spenser novels, start to
finish. ChatGPT obviously hasn’t, or at least not well enough to avoid
confusing two of the main characters. (The other tell: ChatGPT’s
mentioning of the character’s “hawk-like features.”)
So should you trust AI content to always be totally accurate? Of course not.
You should treat it like any other source of information you come across, as
something that needs to be checked and verified. It isn’t always right.

Protecting Privacy

The final ethical concern involving AI is that of data privacy. Right now AI
engines scrape data off the web to train their models. If your private
information—name, address, email, and the like—is online at all, chances
are it’s been vacuumed up by one or more AI engines and resides in a big
database somewhere.

That probably doesn’t make you happy. It may also violate one or more data
protection laws in this and other countries.

Obviously, this exposes your personal data to potentially malicious use.

Someone might be able to query an AI engine about you and receive in
return information you’d rather remain private. Even worse, criminal types
could use information surfaced from AI to create highly personalized spam
or phishing messages just for your benefit.

And if you think you haven’t put any personal or confidential information
on the web, think again. Personal information can come from anywhere,
including social media posts, answers you’ve provided to AI chatbots on
customer support sites, and even email messages you’ve sent or received.
As to confidential information, if you’ve ever used AI to touch up a
business presentation, create a contract, or write an email message to staff,
you’ve provided AI engines with more grist for the mill. All that
information could be available to AI models and anyone using them.

All this means is that AI has some very serious issues to address concerning
data privacy. This will be an issue going forward.

AI and Copyright Law

Beyond ethical issues, AI also faces some tricky legal issues. One particular
such issue is the question of how AI conforms to existing copyright law.
Copyright is designed to protect the creators of certain types of content,
including books, movies, music, and audio. Does AI violate any copyright
laws by using content to train its large language models? Does AI-created
content qualify for copyright? Does AI content created to mimic existing
works violate the copyright for those works? Those are the primary
questions.

Does Using Copyrighted Material to Train AI Violate Copyright Laws?

As you learned in the previous chapter, many AI models get content by

scraping data off the public Internet. Some of what these models scrape is
public and obvious fair game. But some of the information being fed into
the AI models is copyrighted—books, reports, news articles, songs, movies,
and the like. Isn’t AI prohibited from using these copyrighted items without
permission?
AI companies argue that using copyrighted material to train their models is
perfectly permissible because of so-called fair use laws. These laws permit,
under certain conditions, the use of copyrighted material without the
consent of the copyright owner.

What are those permissible fair use conditions? There are four factors to
consider:

The purpose and character of the use, especially whether it is intended

for commercial resale or for nonprofit educational purposes
The nature of the copyrighted work
The amount of the copyrighted work used as a percentage of the whole
work
The effect of said use on the potential market for or value of the
copyrighted work

The AI industry as a whole argues that using a work to train an AI model is

not a commercial purpose but rather an educational one. If that is true, as
the AI industry claims, then the amount of the work used is irrelevant, as is
the type of work used. The industry also argues that ingesting a work to
train an AI model doesn’t impact the marketability of the original work one
iota. Thus copyrighted works used in this manner fall under the banner of
fair use.
Some people argue with this interpretation. First, they argue that if AI
companies actually make money from their AI models, then using a
copyrighted work to train said models is a commercial use. They also argue
that by training on a given work, this enables the AI model to generate
content based on or similar to that work, thus diminishing the original
work’s value.

Not surprisingly, there have been several lawsuits filed by copyright holders
with more sure to come. Plaintiffs to date include the Authors Guild, Getty
Images, the New York Times, and authors Michael Chabon, Paul Tremblay,
and Sarah Silverman; so far none of these cases have come to trial, which
leaves the question open.

Time will tell how this issue is resolved. If ingesting copyrighted material
for AI training purposes is found to violate copyright law, however, expect
it to dramatically shake up the AI industry and change how AI models
identify content for training.

Is AI-Created Content Copyrightable?

On the flip side, content created by AI engines, without significant human

authorship, definitely does not receive copyright protection. That is, AI
companies cannot copyright the written, visual, or audio content that AI
generates.
That’s because United States Copyright Office (USCO) has explicitly stated
that works generated entirely by AI are not copyrightable. USCO considers
AI-generated content to be “derivative authorship;” it’s not created, but
rather adapted from existing material. It is USCO’s longstanding practice to
“require human authorship as a condition of copyrightability”—and AI-
generated content is not authored by humans.

So, no, you can’t copyright anything you “create” with an AI engine. It’s
not something you actually came up with on your own, so it’s not
copyrightable.

Can AI Violate Copyright Laws with Content It Creates?

Another issue arises when AI generates content that too closely resembles
existing works. Does it infringe on a work’s copyright to somehow copy
some or all aspects of that work?

Whether copying a work is a copyright violation depends on how similar

the AI-generated work is to the original. The way U.S. copyright law reads,
infringement may occur if the AI program both had access to the original
work (that is, if it ingested the original into its large language model) and
generated “substantially similar” output. What does that mean in practical
terms?

The difference seems to come down to the difference between mimicking a

style and copying content. So, for example, if you ask an AI engine to write
an article in the style of famous author Stephen King, that is not a copyright
violation because imitating a creator’s style isn’t theft. However, if you ask
an AI engine to write a book called The Stand about a post-apocalyptic
world ravaged by a biological virus and to do so in the style of Stephen
King, it’s likely that the engine will generate a work that very closely
resembles the source material—and that would be a copyright violation.

In other words, you can mimic or imitate something, but you can’t outright
copy it whether you’re a human being or an AI model.

Best Practices for Ethical AI Content Creation

Knowing all that you now know, what are the best practices for using AI to
ethically create content? Here are some of the main points to consider:

Don’t use AI to copy others’ content or style.

Do use AI to generate ideas and give you a starting point for your own
content.
Don’t assume everything you get from AI is 100 percent accurate.
Do fact check the content that AI generates.
Don’t present AI-generated content as your own.
Do be transparent about where AI-generated content came from.
Don’t use AI for malicious purposes.
Do be aware of bias built into some AI models.
Don’t use AI to cheat on your schoolwork
Bottom line: Use AI as a tool, not your only tool and not an unimportant
one, but also not the sole source of your content. Use it responsibly and
with proper oversight. And always, always, trust your instincts and follow
your own moral and ethical compass.

Summary

In this chapter, you learned about some of the best and the worst impacts of
AI technology. AI is already providing benefits to people in all walks of life
(and businesses, too) but also comes with its share of risks. We need to be
aware of these risks even as we strive to take advantage of AI’s benefits.

One of the more significant risks of AI is that the technology will be used to
mislead the public through artificially generated deepfakes and
misinformation. While it’s becoming more difficult to spot this malicious
AI-generated content, there are some telltale signs of AI involvement, of
which we should all be aware. (For example, count the fingers in all AI-
generated images!)

AI also presents some unique ethical and legal challenges. You should not,
for example, pass off AI-generated content as your own or use AI content to
complete your schoolwork. Where AI fits within our current copyright laws
is also in debate, including the legality of using copyrighted material to
train AI engines and whether AI-generated content violates copyright laws
when it hews too closely to the original material.
Bottom line: You need to learn how to use AI safely and responsibly, which
includes learning how to recognize and not get fooled by AI-generated
content. In the right hands and with the right intent, AI can be a powerful
tool. In the wrong hands, it can do major damage. You need to do
everything in your power to make sure that AI is used for good, not for evil.

OceanofPDF.com
3

Getting Started with All-Purpose AI Tools

IN THIS CHAPTER

What is an all-purpose AI tool—and what can you do with it?

How to use a generative AI tool
Getting to know publicly available AI tools
Using the most popular AI generators
Which generative AI tool is the best for you?
How to construct a great prompt
Using AI-generated results—with caution

As we get into the how-to section of this book, I want to introduce you to
some all-purpose AI tools that are designed not just to predict behavior but
also generate new content. Some of these tools (sometimes called general-
purpose AI tools or AI generators) are standalone tools; others are
incorporated into apps or websites you already use.

These all-purpose AI tools are fairly straightforward and intuitive: You tell
the tool what you want to create, via a simple prompt (whether in English or
another language), and the AI tool creates it.

What generative AI tools are available today? Which are best for your
specific needs? Are they free? And—the most fun part—how do you use
them to generate the best results?

The answers to all those questions, and more, are in this chapter.

What Is an All-Purpose AI Tool—and What Can You Do With

It?

All-purpose AI tools—such as ChatGPT, Gemini Chat, Meta AI, and

Microsoft Copilot—are designed to perform a wide variety of tasks,
whereas task-specific tools have one area of expertise. All-purpose tools are
typically trained on large general-purpose datasets and are capable of
generating output in a variety of formats.

What can you do with an all-purpose AI tool? Here are just a few of the
things they can do:

Generating ideas
Answering questions about anything you can think of
Researching topics
Writing letters and emails
Writing text messages and social media posts
Writing blog posts and articles
Writing fiction, nonfiction, poetry, scripts, and other creative writing
Summarizing articles, meeting notes, websites, and more
Making travel plans
Carrying on conversations
Read on to learn how all-purpose AI tools work and how to get what you
want from them. We’ll go into more detail about specific applications in
future chapters.

How to Use a Generative AI Tool

Most all-purpose tools work the same way. You enter a prompt that
describes what you want and then you press Enter. The AI tool does what
you asked it to do—answer your question, write a letter, whatever.

Your prompt can be in the form of a question, a statement, or just a snippet

of text. Grammar and punctuation don’t matter that much. (That means you
don’t have to put a question mark at the end of a question—although you
can if you want.)

For example, if you need to write a letter to your homeowner’s association

complaining about a recent increase in monthly dues, you might enter the
following prompt:

Write a one-page letter to my HOA complaining about dues going up $20

per month.

You can specify the type of output—style, length, and so on—as well. For
example, if you want to give a humorous speech at a friend’s anniversary
celebration, you might enter the following prompt:
Create a three-minute humorous speech for Ron and Barb’s 20th wedding
anniversary celebration. Talk about how we met while I was taking out the
trash on the wrong day when we first moved into the neighborhood.

The more specific you are in what you want, the more accurately the AI tool
will do your bidding. If you don’t provide enough details, it will start
making assumptions and may or may not give you want you wanted.

As an example of how this works, consider Google Gemini, a popular all-

purpose AI tool. You enter your prompt into the Enter a prompt here box
at the bottom of the page, as shown in Figure 3.1.

FIGURE 3.1
Enter your prompt into the Enter a prompt here box.

Gemini’s response is displayed in the space above the prompt box, just
below your prompt, as shown in Figure 3.2. You can scroll down to view
more of the response. You can also highlight the response with your cursor
to select the response and press Ctrl+X on your computer keyboard to copy
it. You can then paste the response (Ctrl+V) into a word processing
document or other file, if desired.

FIGURE 3.2

AI’s response to your prompt.

Other all-purpose generative AI tools work in a similar way: Enter your
prompt into the appropriate prompt box, press Enter, and see the AI
response fill the screen.

Note

Learn more about creating effective AI prompts in the “How

to Construct the Perfect Prompt” section, later in this
chapter.

Getting to Know Publicly Available AI Tools

There are two primary types of all-purpose AI tools available today:

freestanding tools and tools that are embedded in other applications. Both
types of AI generators work in pretty much the same fashion. The type of
tool you choose depends on your particular needs.

Freestanding tools reside on their own websites or mobile apps. They can
answer your questions and generate the content you desire; they possess
broad knowledge about a wide variety of topics. These websites and apps
typically look a little like web search engines, with a big box into which
you enter your prompt. Output is often in the form of a free-flowing text
conversation.

Embedded AI tools are built into an existing site or app to help better serve
its purpose. You use the AI generator within the app and the results appear
in the app; in some instances, you may not even realize that the website or
app is using AI to deliver its services. For example, an AI generator
embedded into a word processing program might let you generate content,
such as facts or other details, that gets inserted directly into the letter you’re
writing.

This section helps you decide which type of AI generator is best for you and
when. As you’ll see, it depends on what you’re doing. If you want a wider
range of options and results, a freestanding AI tool is more appropriate. If
you’re working within an app or program and need help generating ideas or
other content, use the embedded AI generator .

Freestanding AI Tools

The most popular AI generators today are freestanding tools. They’re kind
of jack-of-all-trades tools, in that you can use them to create just about any
type of text-based content. You can also use them to generate new ideas,
answer questions, conduct research, and carry on conversations.

Some of these freestanding generative AI tools are free and available to the
general public to use. Others require some sort of subscription or other
payment. In general, the paid tools tend to be more advanced and targeted at
business and professional users; the free tools targeted at a general audience
often use older, less advanced versions of that site’s AI model and
sometimes put ads on the page.

The most popular freestanding all-purpose AI tools today are detailed in

Table 3.1. (Pricing is at time of publication.)

TABLE 3-1 Freestanding All-Purpose AI Tools

Free
AI Tool URL Paid Version
Version?

ChatGPT www.chatgpt.com Yes $20/month

Claude www.claude.ai Yes $20/month

Google gemini.google.com Yes $19.99/month

Gemini

Meta AI ai.meta.com Yes None

available

Microsoft copilot.microsoft.com Yes $20/month

 Free
AI Tool URL Paid Version
Version?

Copilot

Perplexity www.perplexity.ai Yes $20/month

Pi www.pi.ai Yes None

available

Poe www.poe.com Yes $19.99/month

Note

Google Gemini was previously known as Google Bard.

Microsoft Copilot was formerly known as Bing AI.

Embedded AI Tools

Many computer applications and mobile apps are beginning to incorporate

generative AI. These tools go beyond the simple voice recognition of a
digital assistant or predictive text in a word processor, both functions
powered by the simpler predictive AI. We’re talking about full-featured
generative AI that can perform all manner of tasks, just like freestanding
tools. You can use these tools to ask questions or generate content to use in
the app itself.

For example, the latest version of Microsoft Windows incorporates

Microsoft Copilot AI as part of the operating system. (Microsoft calls it
Copilot in Windows.) It appears as an icon on the right side of the Windows
taskbar. Click this icon to display the Copilot pane, shown in Figure 3.3.
Enter your prompt into the Ask me anything box, press Enter, and Copilot
displays its response, as shown in Figure 3.4.
FIGURE 3.3

The Copilot pane in Microsoft Windows.

FIGURE 3.4

Copilot in Windows’ response to the prompt how do you make a meatloaf with no onions for 12?

Another example is when Gmail suggests potential responses to an email

you have received, such as “Tuesday works for us,” in response to a
message about planning a meeting, or “Hooray!” in response to an email
with good news. Other companies are incorporating AI into their apps in
similar ways. For example, Google has integrated AI into its Google
Workspace apps (such as Gmail, Google Docs, and Google Sheets) via a
tool called Gemini for Google Workspace. Unfortunately, this tool, like
Google Workspace itself, is only available for subscribing businesses.
Using the Most Popular AI Generators

Enough talk. Let’s get down to business with the most popular generative
AI tools today. They all work similarly but have their own distinct
personalities and quirks.

Note

The tools I cover are available for public use at the time of
writing, but there are others in development—so keep
looking for what’s new!

Using ChatGPT

ChatGPT is the consumer-focused generative AI tool from AI research

company OpenAI. It may be the most popular generative AI tool today,
with more than 180 million users. OpenAI says that more than 80 percent of
Fortune 500 companies have integrated ChatGPT into their operations.

Note
 GPT stands for Generative Pre-trained Transformer, which is
another name for a large language model. (I discussed large
language models in Chapter 1, “Artificial Intelligence: What
It Is and How It Works.”)

OpenAI introduced its first version of the GPT model in 2018 and released
ChatGPT in 2020. The basic version of ChatGPT is free, although OpenAI
also offers a more powerful Plus plan for $20. The Plus plan includes access
to the most recent version of OpenAI’s GPT model as well as other AI
tools, such as the DALL-E image generator, which I talk about in Chapter
7, “Using AI to Create Art and Images.”

You access ChatGPT by pointing your web browser to www.chatgpt.com.

While you don’t have to sign up to use ChatGPT (it’s free, whether you sign
up or not), creating an account lets you save and return to previous chats.

Note

ChatGPT describes all prompts and responses as “chats.”

If you’ve created a free ChatGPT account, your past chats are displayed in
the left pane, in chronological order, as shown in Figure 3.5. Click the name
of any chat to see that entire chat in the main pane.

FIGURE 3.5

The ChatGPT home page.

The main pane displays the contents of the current chat and the Message
ChatGPT box at the bottom. Enter your prompt into the Message
ChatGPT box and press Enter on your keyboard. ChatGPT returns its
results in the main pane above the prompt box, as shown in Figure 3.6.
FIGURE 3.6

A ChatGPT chat.

To share the contents of a chat with another user, click the up arrow in the
top right corner of the ChatGPT window. This displays a Share Public
Link to Chat dialog box, as shown in Figure 3.7. Click the Create Link
button; then, when the next screen appears, click the Copy Link button.
You can now paste that link into an email or text message and share it with
others. When they click the link, they see the entire contents of the chat in
question.
FIGURE 3.7

Sharing a ChatGPT chat.

That’s how ChatGPT works on the web. OpenAI also offers a ChatGPT
smartphone app, available for free download from the Apple App Store and
Google Play Store, that lets you always have your AI assistant on the go.
(Make sure you download the genuine app; imposters have already made
their way into the app stores.) The app version works much like the web
version; enter your prompt into the Message box at the bottom of the screen
and ChatGPT provides its response in the main part of the screen, as shown
in Figure 3.8.
FIGURE 3.8

Using the ChatGPT smartphone app.

Even better, you can use the ChatGPT app to talk to AI using normal
speech, which may be a more natural form of interaction than typing
repeated text prompts. When you’re in the app, tap the headphone icon next
to the Message box and start talking. ChatGPT listens to what you say and
then responds in its own voice. You can carry on conversations with AI in
this fashion, which feels revolutionary.
Using Claude

Claude is the generative AI tool from Anthropic, an AI research and

engineering firm backed by Amazon.

Claude has a free plan that imposes daily usage limits and a Pro plan that
raises those usage limits and uses a more powerful AI model. The Claude
Pro plan costs $20 per month.

When you first access Claude at www.claude.ai, you’re asked to either sign
in with your Google account or create a new Claude account. Claude then
asks you your name so it can address you personally.

Claude’s main screen, shown in Figure 3.9, has a big front-and-center

prompt box. Enter your prompt into the Start your first message with
Claude box and either press Enter or click the Start Chat button.
FIGURE 3.9

Claude’s home page.

Claude’s response appears above the prompt box, which moves to the
bottom of the page, as shown in Figure 3.10. You can reply to Claude’s
response by typing into the Reply to Claude box; to start a different chat,
click the Start New Chat (+) icon in the top-left corner of the screen.

FIGURE 3.10

Claude’s response to a prompt.

Underneath each of Claude’s responses is a series of icons, shown in Figure

3.11, that let you Copy the response to your clipboard to paste into other
applications, Retry the request to generate a different response, or give the
response a thumbs up or thumbs down.
FIGURE 3.11

Responding to a Claude response.

Claude can also summarize and analyze content you upload. Click the
Upload Docs or Images (paperclip) icon to upload documents,
spreadsheets, PDF files, or images. For example, I uploaded an initial
version of this chapter and asked Claude to summarize it; that result is
shown in Figure 3.12.

FIGURE 3.12

Claude’s summary of an uploaded Word document.

This ability to work with uploaded documents makes Claude unique in
today’s generative AI landscape. Unfortunately, the usage caps of Claude’s
free version put limitations on how you can use the tool; if you exceed the
limits, you’ll probably need to subscribe to Claude Pro.

Using Google Gemini

Search giant Google created its own generative AI tool and launched
Google Bard in December 2023. The company changed the tool’s name to
Gemini two months later, in February 2024.

Like ChatGPT, Gemini is free to use by the public. Google also offers a
Gemini Advanced version that uses Google’s most advanced AI model; it’s
included with the $19.99 per month Google One AI Premium bundle, which
also includes 2TB of online storage.

Note

In addition to the web-based version, Google offers Gemini

apps for Android and Apple iOS phones and tablets. Look
for the Gemini app in your device’s app store.
When you first access Google Gemini (point your web browser to
https://gemini.google.com) you’ll be prompted to sign into your Google
account, if you’re not currently signed in. Signing in lets you keep a history
of past chats, which you can recall in the future.

By default, the side pane in Gemini is not expanded. You probably want to
expand it to see what’s there, so click the Expand Menu (three-line) icon.
You now see a list of your past chats as well as other controls.

At the bottom of the main pane is the Enter a prompt here box, as shown
in Figure 3.13. To generate results, do as the box says and enter your
prompt there; then press Enter on your keyboard.

FIGURE 3.13

The Google Gemini home page.

SOURCE: gemini.google.com

The response to your prompt is displayed in the main pane above the
prompt box, as shown in Figure 3.14. As you enter more prompts, the
previous response scrolls up.

FIGURE 3.14

The results of a Gemini prompt.

Note

Google Gemini can also summarize web-based articles and

reports as well as complete websites. Just use the prompt
 summarize [URL] and include the URL of the web page
you want summarized.

Unlike ChatGPT and most other all-purpose AI generators, Gemini lets you
generate images as well as text results. Just prompt it to create an image of
and then describe what you want. The results are shown in the main pane,
often with several options, as shown in Figure 3.15. Click a thumbnail to
display a larger image.

FIGURE 3.15

Creating images with Google Gemini.

Underneath all Gemini responses are a series of icons, shown in Figure
3.16, that let you react to that response. You can upvote (thumb up) or
downvote (thumb down) the response, depending on whether you like it.
You can click the Share icon to share it via a public link, export it to
Google Docs, or add it to an email draft in Gmail. You can also click the
Google “G” logo to double-check the response or click the More (three-
dot) icon to copy the response (to paste into another application) or report a
legal issue with the response.

FIGURE 3.16

Reacting to a Google Gemini response.

By the way, Google is now adding Gemini AI-generated summaries at the

top of many web search results pages. The AI Overviews, as Google calls
them, appear before the normal page listings, as shown in Figure 3.17.
These overviews could theoretically answer many user questions, saving
users the trouble of clicking through Google’s normal search results.
FIGURE 3.17

An AI Overview at the top of a Google search results page, powered by Google Gemini.

Warning

Google’s Gemini-powered AI Overviews aren’t always up to date or fully

accurate. To be safe, you’ll want to verify these overviews through
Google’s traditional search results.
Using Meta AI

Meta, the parent company of Facebook, Instagram, and WhatsApp, also

owns the Meta AI research lab, which is tasked with developing practical
applications for AI within Meta’s other products and services. As part of
that research, the company has released the Meta AI freestanding generative
AI tool.

The basic version of Meta AI is free. The company also plans to offer a
subscription-based Meta AI Plus plan, which will provide access to more
advanced AI models and let you enter more and longer prompts. As I write
this, Meta has not yet set pricing for Meta AI Plus.

To use Meta AI, point your web browser to www.meta.ai. If you’re already
a Facebook member, you can log in with your Facebook account.

Meta AI’s main page looks quite a bit like all the other general AI tools.
There’s a left-hand pane you can expand by clicking the right-arrow; when
expanded, you see prior conversations. (What ChatGPT calls “chats,” Meta
AI calls “conversations.”) The prompt box, labeled Ask Meta AI anything,
appears at the bottom of the main pane, as shown in Figure 3.18.
FIGURE 3.18

The Meta AI home page.

To use Meta AI, enter your prompt into the Ask Meta AI anything box and
press Enter. Meta AI’s response appears in the main pane, above the prompt
box, as shown in Figure 3.19. As you enter more prompts and Meta AI
issues more responses, the conversation scrolls.
FIGURE 3.19

A Meta AI prompt and response.

Underneath each response are three buttons, shown in Figure 3.20. There
are buttons for Good response (thumb up), Bad response (thumb down),
and Copy content to clipboard. Use that last button to copy the current
response so you can paste it into another app, such as word processor
document or email message.
FIGURE 3.20

Responding to a Meta AI response.

Note

Meta AI is also available in Ray-Ban’s Meta Smart Glasses

(www.ray-ban.com/usa/ray-ban-meta-smart-glasses), which
can be a useful option for people who are visually impaired.

Using Microsoft Copilot

Not surprisingly, Microsoft has developed its own generative AI tool, called
Copilot. Because Microsoft is a big investor in OpenAI, the two companies
have a strong working relationship, and Copilot uses the same GPT engine
from OpenAI that ChatGPT uses. This means its results are often similar
(and sometimes identical) to those from ChatGPT. Often but not always.
Copilot appears to use some slightly different algorithms than ChatGPT,
which sometimes results in different results, especially on prompts for more
creative output.

You access Copilot at copilot.microsoft.com. You can sign in with your

Microsoft account. The basic version of Copilot is free; the more fully
featured Copilot Pro costs $20 per month.

Note

Like ChatGPT and Google Gemini, Microsoft Copilot is also

available as a mobile app for Android and Apple iOS
devices. Search for the Copilot app in your device’s app
store.

As you can see in Figure 3.21, the Copilot home page is the mirror opposite
of most other AI generators, in that the pane with recent chats is on the right
side, not the left. In that pane you’ll also find tabs for other Microsoft AI
tools, including

Designer, for generating images (we’ll discuss this in Chapter 7)

Vacation Planner, for creating itineraries and booking travel (I discuss
this in Chapter 10, “Using AI to Manage Your Travel and
Transportation”)
Cooking Assistant, for finding recipes and generating meal plans
Fitness Trainer, for designing exercise and nutrition programs (I discuss
this in Chapter 11, “Using AI for Health and Wellness”)
FIGURE 3.21

Microsoft Copilot’s home page.

To use Copilot, make sure Copilot is selected in the right-hand pane and
then enter your prompt into the Ask me anything box at the bottom of the
page. As you can see in Figure 3.22, Copilot’s response is displayed in the
main pane, above the prompt box. Additional prompts and responses are
displayed in a scrolling fashion.
FIGURE 3.22

Viewing Microsoft Copilot results.

Underneath the response is something unique to Copilot—a Learn More

section with search results from Microsoft’s Bing search engine. Hover over
one of these links, shown in Figure 3.23, to view a snippet from that page.
Click a link to go to that page.

FIGURE 3.23

All the stuff underneath a Copilot response.

 Note

For some responses, Copilot also displays related

advertisements. That’s something you won’t find with other
AI tools.

Beneath the Learn More section are ways to respond to Copilot’s response
—Like, Dislike, Copy, Export, Share, and (uniquely) Read Aloud. Click the
icon to do the thing.

Finally, at the very bottom of the response are other related queries in which
you may be interested. Click a query to have Copilot generate a response.

Note

Some people will find the additional links underneath

Copilot results useful. Others will find them unnecessarily
intrusive—especially the ads, when they appear. It certainly
looks like Microsoft is attempting to make Copilot look like
 its Bing search engine and, in some instances, monetize
those results with ads.

Copilot in Other Microsoft Applications

I previously mentioned how Microsoft has incorporated Copilot into

Microsoft Windows. Think of that as a shortcut to Copilot that doesn’t
require you to visit the Copilot page in your web browser.

Microsoft is also incorporating Copilot into its Microsoft 365 (formerly

Office 365) applications—but only for businesses that have a Copilot for
Microsoft 365 or Copilot Pro license. These business users can use Copilot
to

Draft a new document or rewrite an existing one in Microsoft Word

Analyze trends and create data visualizations in Microsoft Excel
Create presentations in Microsoft PowerPoint
Organize your inbox in Microsoft Outlook

At this point in time, these AI features are only available for Microsoft’s
business subscribers and professional users. It’s possible that Microsoft will
migrate these features to other users in the future.
Using Perplexity

Perplexity is a generative AI tool that blends the features of a search engine

with AI powered by OpenAI’s GPT-3.5 model. It’s unique in that it cites the
sources of its responses, which makes it a good choice for research or
educational purposes.

The standard version of Perplexity is free. The Professional version, which

lets you choose your AI model (from GPT-4, Claude-3, or Sonar Large),
create images, and upload and analyze files, runs $20 per month.

You access Perplexity at www.perplexity.ai. You can create a new account

or sign in with your Apple or Google account.

As you can see in Figure 3.24, the left-hand pane displays a Library of past
threads. (What ChatGPT calls “chats” and what Meta AI calls
“conversations,” Perplexity calls “threads.”) You use the Ask anything box
in the middle of the main pane to enter your prompt.
FIGURE 3.24

Perplexity’s home page.

The results, shown in Figure 3.25, look quite a bit different from those from
other AI tools. First off, it’s not really a conversation, but rather a display of
information. Second, the sources for that information are displayed above
the response; click a source to go directly to that web page. Third,
Perplexity displays images related to your prompt on the right side of the
page. Fourth, you can also search related videos by clicking the Search
Videos link on the right. And fifth, if you subscribe to Perplexity Pro, you
can generate new images based on your prompt by clicking the Generate
Image link.
FIGURE 3.25

A research-oriented response from Perplexity.

That’s a lot that’s different from the other AI tools, but the differences don’t
end there.

If you want to drill down further, you can enter a related prompt into the
Ask follow-up box. To start a new thread, however, you have to click the
New Thread button in the left-hand pane. This displays a new Ask
anything box, like the one in Figure 3.26, into which you can enter your
new prompt.
FIGURE 3.26

Beginning a new thread in Perplexity.

Underneath each Perplexity response are a series of response icons, as

shown in Figure 3.27. From left to right, you can Share the thread, Rewrite
the response, Copy the response, Edit the Query to fine-tune the response,
or click the More (three-dot) icon to view sources or report this particular
response.

FIGURE 3.27

Responding to a thread in Perplexity.

Using Pi

Pi is an AI tool that works more like a chatbot. When you first log into Pi,
you’re asked to choose a voice so that Pi can respond to you verbally, as
well as via text. Pi is free to use, and you can log into it with your Apple,
Facebook, or Google account, or you can create a new Pi-specific account.
You access Pi at www.pi.ai. As you can see in Figure 3.28, its interface is
quite spartan. The panel on the left really doesn’t serve much purpose, and
all you see is the Talk with Pi box at the bottom. You can turn on or off Pi’s
voice responses by clicking the speaker icon at the top-right corner of the
screen.

FIGURE 3.28

Pi’s home page.

As you can see in Figure 3.29, a Pi conversation looks a lot like a typical
text message thread. Your messages appear in a darker shaded box,
followed by Pi’s responses.
FIGURE 3.29

A conversational thread with Pi AI.

That’s not to say you can’t use Pi for more heavy lifting. Like the other
generative AI tools, Pi can compose letters, write stories, provide
information, and do pretty much everything the other generative AI tools
can do. It just does it in the form of a conversation (or what Pi calls a
“thread”), which some people might find more friendly than some of the
other AI tools.

Using Poe

The final AI tool I’m sharing is called Poe. Poe is actually an AI aggregator
that provides access to a variety of AI tools in a single hub. Through Poe,
you can get responses from OpenAI’s GPT-4 and DALL-E, Google’s
Gemini, Meta’s Llama, Anthropic’s Claude, Stability AI’s StableDiffusion,
and more. Not all of these AI tools (what Poe calls “bots”) are available
with the free version of Poe, however; to get full access to all available
tools, you need to subscribe for $19.99 per month.

All these options make Poe a little less user-friendly than some of the other
AI tools, which you can see from its interface. When you access Poe’s
home page, at www.poe.com, and log in with your Apple, Google, or Poe
account, you see the screen shown in Figure 3.30. A lot of options appear in
the left panel, of which you may use some or none, as well as a Start a new
chat box in the main pane.

FIGURE 3.30

Poe’s home page—lots of options!

To start, select the AI tool you want to use for your query. You can choose
from Assistant (Poe’s own AI tool and the default), Gemini-1.0-Pro (from
Google), or ChatGTP. Click the More button to choose from additional
tools (again, not all available in the free version).

Once you’ve selected your AI tool, enter your prompt into the Start a new
chat box and press Enter. As you can see in Figure 3.31, the response looks
like what you’d find with other AI tools—until you scroll down to the
bottom of the response. There, as shown in Figure 3.32, you have the option
of comparing this initial response with responses from other AI tools. Click
a tool to see its response.

FIGURE 3.31

The response to a prompt in Poe.

FIGURE 3.32

Comparing Poe’s initial response to another generative AI tool.

This cross-tool functionality makes Poe a great choice if you’re not sure
which AI generator is best for you or if you just want to compare how the
different AI tools respond to a given prompt. It has a slightly longer
learning curve, which may not make it the ideal daily driver for AI needs.
Still, it’s an interesting approach to generative AI today.

Which Generative AI Tool Is the Best for You?

With multiple all-purpose AI tools available (and more coming online

seemingly daily), how do you choose which one to use?

Answering the question of which is the best AI tool is an impossible task.

Each tool is trained on a slightly different data set, and each employs
slightly different algorithms. That means you won’t get the exact same
results from each tool.

As a comparison, I asked some questions of three of the most popular

generative AI tools, each using its own unique AI model: ChatGPT, Google
Gemini, and Meta AI. Let’s see how they compare.

First, I asked each tool to tell me the best way to reheat leftover pizza.
ChatGPT got right to it and recommended a single approach (the oven
method), as shown in Figure 3.33. Gemini presented several different
methods in a conversational style, as shown in Figure 3.34. Meta AI also
presented several different methods but offered step-by-step instructions, as
you can see in Figure 3.35.
FIGURE 3.33

ChatGPT’s pizza reheating advice.

FIGURE 3.34

Google Gemini’s pizza reheating advice.

FIGURE 3.35

Meta AI’s pizza reheating advice.

Next, I asked each engine to write a short story about a lost girl in the
woods. (I like to tell my grandkids stories.) Here the three tools took wildly
divergent paths, as you can see in Figures 3.36 (ChatGPT), 3.37 (Gemini),
and 3.38 (Meta AI).
FIGURE 3.36

ChatGPT’s short story.

FIGURE 3.37

Google Gemini’s short story.

FIGURE 3.38

Meta AI’s short story.

Finally, to see which AI tool had the most poetic soul, I asked all three to
compose a haiku about snow. The results, all equally poetic, are in Figures
3.39 (ChatGPT), 3.40 (Gemini), and 3.41 (Meta AI).
FIGURE 3.39

A haiku from ChatGPT.

FIGURE 3.40

A haiku from Google Gemini.

FIGURE 3.41

A haiku from Meta AI.

Is any one of these three tools significantly better than the others? I don’t
think so; they just offer slightly different results, just as you’d get different
pieces of advice if you asked the same question from three of your friends.

That said, you may find that you like the results (or how they’re presented)
from one tool better than the others. If so, that is probably the right tool for
you.

The best thing to do is give them all a try. Prompt each one with a similar
request and see what results you get. You can then choose the one that feels
right to you—or use multiple tools as you prefer.

Note

You may find that the AI tool that appears better today might
not be better tomorrow. As AI models continue to evolve, the
different models used will frequently leapfrog each other in
terms of results and features. So be prepared to switch AI
 tools as the models develop over time; you may like a totally
different tool in the future.

How to Construct the Perfect Prompt

To get the most relevant, appropriate, and accurate results from a generative
AI tool you need to construct the right prompt. A well-crafted prompt tells
the AI tool exactly what you want and includes context, intent, and the
information you want included. The better and more detailed prompt you
enter, the more satisfying and more accurate results you’ll receive.

What that in mind, here are some tips for constructing a great generative AI
prompt:

Be clear about what you want. Clearly articulate what you want
created or answered. Include as much detail as you can. Provide any
necessary background information that might help AI provide a more
appropriate response. List things that you want included and those you
want excluded. Don’t make the AI tool guess what you want.
Specify the desired length. To avoid AI providing responses that are too
long or too short, request a specific length. If you want a three-paragraph
letter, tell it you want a three-paragraph letter. If you want something
short, say you want a short answer. Don’t expect the AI tool to know
how long a response you want.
 Specify the desired format. You should also specify what type of output
you want—letter, research paper, whatever. This helps guide AI to create
the most appropriate response.
Specify the desired style or voice. If you want an informal letter to
friends, tell AI you want informal output. If you want something written
in a specific genre or writing style, tell it. If you want content to use in a
technical report, tell the AI tool that you want technical output. AI tools
can output the same content a lot of different ways.
Use plain language. Write your prompt in conversational language and
avoid jargon-heavy language or phrases that only an industry insider
would understand.
Ask direct questions. If you want a specific answer, you need to ask a
specific question. Don’t make open-ended requests. You need to point AI
in the right direction to properly address your query.
If you don’t like what you get, try again. You don’t have to settle with
the first response generated by an AI tool. Evaluate the response and
then fine-tune your prompt to generate another, hopefully more accurate,
response. (And keep going until you get the response you desire.)

The more detail you include, the better-tailored the response, so don’t be
afraid of writing long, detailed prompts.

Here are some examples of well-crafted AI prompts:

 How do I change the air filter in a 2020 Honda Civic? Please include the
tools I will need and how long it will take.
What is a good menu for dinner with six people, with no red meat? One
person is allergic to eggs.
Recommend five exercises I can do in the house. I am a 45-year-old
woman with no health problems.
What is the oldest state capitol in the United States?
Create a one-page form letter for signing up for the neighborhood block
party on August 8. Include spaces for the participant’s name, address,
and contact information.
Recommend the five best things to do on a one-week trip to Paris in
April. We will not have a car, we don’t like museums, and we are on a
budget.
Generate a short story for a middle school audience about two friends
first meeting at summer camp.
Write a 500-word summary of War and Peace in simple terms so I can
explain it to someone.
What is the most soothing color to paint a newborn baby’s room?

Remember, the more detailed your prompt, the better the response. For
example, if you enter the prompt how do birds fly? into Meta AI, you get a
very general response, like the one in Figure 3.42. If, on the other hand, you
enter the prompt how does a hummingbird fly?, you get a more specific
response, like the one in Figure 3.43.
FIGURE 3.42

A general answer to the general question, how do birds fly?

FIGURE 3.43

A more specific answer to the more specific question, how do hummingbirds fly?

You can also fine-tune AI responses by including a specific length or style

in your prompt. For example, I started out by prompting Google Gemini to
write 200 words about bird migration. Just that, nothing more. The rather
straightforward response is shown in Figure 3.44
FIGURE 3.44

All about bird migration, in a general style.

Next, I asked Gemini to write about bird migration in the style of a standup
comic, bad jokes and all. See that response in Figure 3.45.
FIGURE 3.45

Writing about bird migration in the style of a standup comic.

Not satisfied with that, I then asked Gemini to write 200 words about bird
migration in the style of a hard-boiled detective novel. That response is
shown in Figure 3.46.

FIGURE 3.46
Writing about bird migration in the style of a hardboiled detective novel.

You get the point. AI can generate content in whatever form and style you
ask for. All you have to do is ask for it, in as much detail as possible.

Generative AI can even do very odd juxtapositions of content and style. For
example, when I asked Google Gemini to write about bird migration in
iambic pentameter, I got the results shown in Figure 3.47. When I asked it
to write a limerick about bird migration, it returned the results in Figure
3.48. AI does what you tell it to do, to the best of its abilities.

FIGURE 3.47
All about bird migration in iambic pentameter.

FIGURE 3.48

Bird migration in a limerick.

Using AI-Generated Results—with Caution

As I explain in Chapter 2, “The Risks and Benefits of AI,” AI results aren’t

always reliable. In fact, many of the AI tools warn you right up front that
their results may be less than accurate. (Meta AI puts it like this: “Messages
are generated by AI and may be inaccurate or inappropriate.”)

Why are these tools not always reliable? There are a number of reasons,
including

The data used to train the AI engine might not be up to date. In other
words, there is a cut-off date for what each AI knows.
 The training data might not include enough information about the topic
you asked about.
The AI tool might not understand what you asked it to do. (AI doesn’t
always understand context or all the subtleties of human language.)

In other words, AI can be just as inaccurate as a human being—which is

why it is important to always verify results generated by AI tools.

To verify the accuracy of AI results, start by cross-checking the AI’s results

with other reliable sources. And by “reliable sources,” I don’t mean another
AI engine, which could be using the same flawed training data. Instead,
compare the AI’s results with information from credible websites—or just
use your search engine to search for the same information and check the
sources to see if the results match.

In addition, you can ask the AI tool follow-up questions to clarify the
response or seek additional information. If you’re not clear what the results
mean, ask the AI tool to explain further.

Note

You may want to check the grammar, punctuation, and

spelling in AI results before you use them elsewhere.
 Sometimes AI-generated results can sound stilted and lack
the natural flow of human language. Look especially for
repetitive content, use of similar sentence structures, and
awkward phrasing.

Finally, also as noted in Chapter 2, you should not try to pass AI-generated
content off as your own creation. That means you shouldn’t submit an AI-
generated poem to a literary journal or use AI-generated content to
complete a schoolwork assignment. You can use AI-generated content for
your own use but not much beyond that.

How AI Tools Use Your Information

In Chapter 2, I discuss some of the privacy concerns with artificial

intelligence, specifically where AI gets its training data and how it uses it.
Not surprisingly, there are also privacy concerns with the prompts that users
enter and the results generated by those prompts.

First, know that everything you enter into an AI tool—like everything else
you do online—will be monitored, collected, and, in many cases,
monetized. If you’re not comfortable with AI companies using your inputs
and responses in this fashion, then you shouldn’t be using an AI tool.
Companies utilize user information (prompts and follow-up replies) to help
further train their AI models; it’s one way that AI learns. When you sign up
to use an AI tool, you’re implicitly giving your permission for the company
behind AI to use anything you enter to help them improve their models.

AI learns from the responses it provides to your queries and how you
respond to those responses. If you give a particular response a thumbs up,
the AI engine will know it did something right; if you give it a thumbs
down, it will learn from its mistake. So your user interactions also become
the property of the AI company.

This also means that any personal information you enter into an AI tool is
likely collected and used by the company offering the tool. It’s the same
with the personal information you might write in a Facebook or Instagram
post; whatever you enter, it’s now in the company’s possession.

AI companies, like social media companies (and they’re often the same
entities), might say that they “minimize” user data by keeping only that
information they need to provide and improve their services. They also
might say that they incorporate robust security measures to keep your data
safe. But we all know that even the most secure companies can suffer data
breaches and when was the last time you trusted a big company to do
anything good with your personal data, anyway?

The point is this: The more personal your interactions with an AI tool, the
more that AI knows about you and the more the company behind AI can use
that information for its own purposes. Avoid sharing too much personal
information about yourself or others, such as a Social Security number,
bank account and credit card numbers, or any personal information that a
hacker could use to guess your passwords. (So don’t enter your pet’s name
or your children’s birthdates.)

Above all, you need to stay vigilant and regularly monitor your personal
information, bank accounts, credit reports, and the like for unusual activity.
Treat your friendly neighborhood AI tool the same way you’d treat any
website or online service and provide only the minimal amount of
information necessary to do what you need to do.

Summary

This chapter covered all-purpose AI tools, also known as AI generators—

what they are, how they work, and how to use them. This chapter also
showed you how to create great prompts for generating the results you
want. You learned how to make your prompts more detailed by describing
the type, length, and style of output you want. You learned how changing
just a few words in a prompt can generate wildly different results.

Finally, this chapter showed you how to verify the results generated by AI
tools. AI, like the human-supplied information used to train it, isn’t perfect
—which means the results it generates, no matter how impressive, need to
be double-checked and verified.
OceanofPDF.com
4

Using AI to Find the Right Words

IN THIS CHAPTER

When and how you should (and shouldn’t) use AI for writing
Using AI for different types of writing
Using AI to improve your writing
Examining popular AI tools for writing and editing

As you learned in Chapter 3, “Getting Started with All-Purpose AI Tools,”

generative AI tools like ChatGPT, Google Gemini, Meta AI, and Microsoft
Copilot can write all sorts of content for you, from short and simple texts,
social media posts, letters, and blog posts to long and complex articles and
research papers. AI can also help with creative writing: nonfiction, fiction,
scripts, and poetry.

Even if you don’t want AI to do all your writing for you, AI can help make
you a better writer. We’ll talk about some of the ways in this chapter. And
we’ll cover some of the cautions to watch out for.

When and How You Should (and Shouldn’t) Use AI for Writing

Before we get started, let’s consider when and how to use AI for writing.
Just because you can use AI for writing doesn’t mean you always should.
It’s one thing to use AI to help you generate ideas, write notes to your
friends, or even create form thank-you letters. Those uses are all fine. It’s
something else to have AI generate a research paper or novel that you then
claim as your own. There are ethical and transparency issues to address.

First, using AI to write personal notes, letters, and social media posts is
perfectly acceptable, and lots of people do it. We all have to do various
types of casual writing, and AI can help you find the right words for what
you want to say. For these types of everyday writing tasks, AI can help level
the field.

In addition, AI is quite useful for what you do before and after you write a
piece. Beforehand, AI is great for generating ideas for things you need to
write, which you can then elaborate on as you write the final version. In
other words, AI is good for generating first drafts, which you can then fine-
tune.

AI is also good for editing your writing after the fact to make you sound
better. And it can rewrite or rephrase content for a different reading level—
taking a technical piece and rewriting it to an eighth-grade level, for
example, or beefing up a simple piece for a more technical audience.

That said, you shouldn’t use AI for certain writing assignments. For
example, using AI to write papers for school without a teacher’s consent is
cheating. Teachers want to know how well you’ve learned the lesson at
hand, and having AI do your work for you doesn’t make that possible.

Equally important, when you’re writing professionally, you shouldn’t claim

AI-written material as your own. Let’s say you have AI write a short story
or article for a community newsletter. For transparency’s sake, you need to
disclose that the piece was AI generated; otherwise, you’re claiming credit
for something that you didn’t write. Crafting an AI prompt isn’t writing.

Similarly, you can’t claim copyright protection for AI-generated material

because you didn’t write it.

You also have to be careful about AI using copyrighted material in its

responses. Some of the content used to train AI tools is copyrighted and
some people—including some copyright owners—consider AI’s use of that
material a copyright violation, akin to theft. If AI uses someone else’s
words exactly, and you use that, you could be subject to a copyright claim
by that party.

Bottom line: Using AI to draft a personal note, generate ideas, or improve

your writing is perfectly acceptable, and lots of people are doing it. AI can
be a wonderful writing aid, especially for people who aren’t professional
writers by trade. It’s a good use of the technology and perfectly legal,
ethical, and acceptable.
Using AI to do more formal or official writing is something else altogether.
If you’re writing something for which you would otherwise claim credit—a
novel, a research paper, even a newspaper article—then you shouldn’t have
AI do the work for you.

Will AI Replace Real Writers?

While AI is a valuable tool for anyone writing casually, what it can’t do, at
least in my personal opinion, is replace professional writers. (Like me!)

As a professional writer I may have a vested interest in this, but I have yet
to see an AI tool that can write as well as a pro. An AI tool just doesn’t have
the insight, the experience, and the personality of a human writer. I can
always tell the difference.

So, in my humble opinion, if you want the best quality writing for
professional purposes, enlist a professional human writer. As good as AI
currently is and probably will become, it will never have the heart and soul
that a skilled human writer brings to the table.

(Just to be clear, I wrote that myself, with no help from AI whatsoever.)

Using AI for Different Types of Writing

Any all-purpose AI tool can generate social media posts, texts, email
messages, letters, and other forms of communication. As with all things AI,
of course, the better your prompt, the more relevant and accurate results
you’ll receive. That is especially true when it comes to writing. You need to
be very precise as to what you want written and how.

Let’s look at some examples.

Note

Learn more about all-purpose AI tools like ChatGPT, Google

Gemini, Meta AI, and Microsoft Copilot in Chapter 3.

Writing a Short Note to Family or Friends

AI can create personal messages that sound like you wrote them yourself,
perfect for sending to your friends and family that you either don’t have the
time or the talent to write yourself.

For example, if you want to let your mother know that you’re coming home
for Thanksgiving (and bringing your significant other with you), you might
enter the following prompt:
Create an email message to my mother letting her know that I’ll be
coming home for Thanksgiving and bringing my new girlfriend, Jane.
We’ll arrive the day before and stay through Sunday.

ChatGPT’s very personal email message is shown in Figure 4.1.

FIGURE 4.1

An email informing your mother of Thanksgiving plans, courtesy of ChatGPT.

 Note

Some people consider using AI to write personal notes and

letters to be somewhat impersonal and insincere. Personal
messages, many people believe, should be personal—not
generated by machine. Use your own best judgment in this
regard.

Writing a Longer Letter to Family and Friends

You can also use AI to write longer emails, letters, and other casual
communications. For example, you can let your friends know what you’ve
been up to recently in an AI-generated letter. Just throw some recent
activities into a prompt, like this:

Write a two-page letter to my friends catching up on what I’ve done

this summer. I’ve gone surfing at Pacific Beach, hiking near Estes Park,
took a two-week road trip through Northern California with my friend
Eleanor, and spent the rest of the time studying for law school
admission tests.

Your very personal letter, ready to cut and paste into a Word document, is
shown in Figure 4.2. Google Gemini did a pretty good job with this one.
FIGURE 4.2

A “personal” letter from Google Gemini getting your friends caught up on your summer activities.

You get the point. When you don’t know what to write or don’t want to
write, use AI to help get started.

Writing an Email Message

Email messages are typically short and to the point. When prompting AI to
write an email message, include the subject of the email and a brief
description of what you want written. You should also tell the AI tool about
how long the email should be and the desired tone.
For example, to write an email inviting your neighbors to a cookout, you
might use the following prompt:

Draft an email inviting my friends to a cookout on June 15th at my

house. Ask them to bring sides or desserts.

Figure 4.3 shows the resulting message from Google Gemini that you can
cut and paste into an email.

FIGURE 4.3

An email invitation drafted by Google Gemini.

Writing a Thank-You Note

To have AI write a thank-you note, it has to know who you’re thanking and
what you’re thanking them for. It will do the rest.

For example, you might craft a prompt like the following:

Write a thank-you note to my Aunt Judy for the sweater she gave me
for my birthday.

Figure 4.4 shows the thank-you note that ChatGPT wrote in response.

FIGURE 4.4

A thank-you note written by ChatGPT.

Writing a Business Letter

Business letters are by nature more formal than personal notes or letters.
You need to tell AI who you’re writing to, what the letter is about, and
whether the style should be casual or formal business. For example, you
might use a prompt like the following:

Write a formal business letter to Don in accounting asking him to

prepare an analysis of this year’s sales vs. last year’s sales across all
territories. Ask him to get it done by the end of the week.

When thus prompted, Microsoft Copilot wrote the surprisingly detailed

letter shown in Figure 4.5.

FIGURE 4.5

A formal business letter written by Microsoft Copilot.

Writing a Social Media Post

In contrast to business writing, writing for social media is much less formal.
To have AI write a social media post for you, you need to tell it which
social network the post is for and what you’re writing about. For example,
here’s a prompt for a post on Facebook:

Write a short post for Facebook about our new golden retriever,
Buster.

Figure 4.6 shows the post that MetaAI created. It even added hashtags to go
along with the message.

FIGURE 4.6

A Facebook post written by MetaAI.

Writing a Memoir

Many people would like to tell the story of their lives, or the lives of a dear
relative or friend. This is a bit of a challenge for AI in that the AI engine
probably doesn’t know that much about your life or the life of the other
person. You’ll need to feed the AI tool lots of detailed information about
your life—what you’ve done, where you were, when things happened, and
with whom you did them.

Warning

Remember, all information you feed into an AI prompt is entered into the
AI database for further training and use. Be careful about what personal
information you share with an AI tool. And no matter how inspirational and
exciting, when you show others the output, you’ll want a disclaimer that the
content is generated by AI.

Because your life is probably longer than a simple prompt, you may want to
write your memoir in stages. Enter as much information as you can about a
particular time or event, have AI write it up for you, and then move on to
the next major event in your memoir.

For example, you might craft a detailed prompt like the following:

Write the story of how I met my future husband when we were both
attending a Lynard Skynyrd concert in Long Beach and we bumped
into each other at the merch table. We both wanted to buy the same t-
shirt, so we started talking, and he ended up driving me home
afterward.
Figure 4.7 shows how ChatGPT interpreted that story. Makes you wonder
how close it got to what really happened back then.

FIGURE 4.7

A personal memoir recollected by ChatGPT.

Writing a Short Story

AI is surprisingly creative and can write some inventive fiction. Tell it you
want it to write a short story, give it a brief plot, maybe even the names of
characters and locations, and then let it rip.

Here’s a sample prompt for a short story you can have AI write for a child
or grandchild:
Write a short story for a young girl named Raven about how she finds
a whale trapped in a fishing net and how she rescues the whale so it can
return to its family in the ocean.

As you can see in Figure 4.8, Google Gemini handled this task with aplomb
—and quite a bit of imagination.

FIGURE 4.8

The start of a short story conceived by Google Gemini.

Writing a Poem

Let’s tackle one final writing task—writing poetry, for your own or a loved
one’s enjoyment. (Again, you wouldn’t want to present an AI-generated
poem to a poetry journal or other public venue.) You can simply ask an AI
tool to write a poem about a given subject. You can provide more or less
detail about the subject or the desired length of the poem. You can even ask
AI to write in the style of a specific poem or poet.

For example, consider this prompt:

Write a short poem about the sound of thunder on a summer night in

the style of Edgar Allen Poe

Figure 4.9 shows what MetaAI wrote. It is indeed very Poe-like.

FIGURE 4.9
A poem like Poe might write about summer thunder, written by MetaAI.

Note

A handful of AI tools are specifically geared toward writing

poetry. These tools include AI Poem Generator
(https://poemanalysis.com), AI Poem Writer
(www.poemanalysis.com/poem-generator/), and Hypotenuse
AI Poem Generator (www.hypotenuse.ai)

Using AI to Improve Your Writing

AI can be a useful tool for improving your writing. Think of AI as being a

really thorough and fast editor or (depending on how involved it is) a
diligent researcher. AI can take one look at a piece of writing and perform a
variety of useful tasks, as detailed in the following sections.

Note
 You can also find specific AI tools to detect plagiarism,
translate text from one language to another, and summarize
your text. AI is also good for putting text into the proper
style—for example, adding abstracts, citations, and the like
to scientific or scholarly papers.

Using AI to Suggest Topics

Maybe you have an assignment to write a paper for a class or an article for
work and you can’t decide on a good topic. Maybe you’ve been staring at a
blank screen or sheet of paper for hours and just don’t know how to get
started. AI can be quite useful when you’re just starting out on a piece.

Any all-purpose AI generator, like the ones discussed in Chapter 3, can

function as an idea generator. There are also several AI-powered topic
generators available, including these:

Ahrefs Content Idea Generator is a free tool found at

https://ahrefs.com/writing-tools/blog-post-ideas that lets you input your
niche or target audience and then generates content ideas for selected
output formats (articles, blog posts, research papers, video scripts, and so
on)
CopyAI Idea Generator (www.copy.ai/tools/content-idea-generator), a
free tool from CopyAI that lets you select a niche or topic and then
generate story ideas
 HubSpot Blog Ideas Generator (www.hubspot.com/blog-topic-
generator), a free AI-powered tool that helps you generate titles and
ideas for blog posts

Whether you’re using an all-purpose AI tool or a more targeted idea

generator, here are some tips to help you and AI brainstorm topic ideas:

Ask for ideas. To get an AI tool to generate ideas, you have to ask it to
generate ideas. Try a prompt like Generate 5 ideas for articles about
road safety.
Provide seed information. Tell the AI tool about your target audience,
niche, output format (article, blog post, and so on), and anything else that
can help guide AI to generate useful content ideas. For example, you
might construct a prompt like I need ideas for an article about car
repair for people on a budget.
Focus on current trends. When asking AI to generate topic ideas, ask it
for topics associated with related industry or societal trends. It may be as
simple as prompting Generate 3 article ideas for what’s new in home
security.
Focus on audience questions. The better you know your target
audience, the better AI can target appropriate topic ideas. Find out what
your audience is talking or asking questions about and provide that
information to the AI tool to suggest related topics.
Provide the right input, and you’ll get a list of suggestions like the one in
Figure 4.10, from ChatGPT.

FIGURE 4.10

Using ChatGPT to generate topic ideas.

Note

For the best ideas, don’t rely on AI alone. Combine AI with

your own brainstorming to generate even more ideas and
 filter them to a relevant list.

Using AI to Outline Content

You might have an idea for a piece but have difficulty organizing your
thoughts. What should you write first, what do you absolutely need to
include, what can you leave out?

The best way to organize your thoughts is to draft an outline. This is

something at which AI excels.

As with idea generation, all-purpose AI tools are pretty good at creating

outlines based on topic ideas. There are also several AI-powered outline
generators, including

Akkio AI Outline Generator (www.akkio.com/tool/ai-outline-generator),

a free tool for crafting structured outlines for different types of projects
GravityWrite Blog Outline Generator (https://gravitywrite.com), a free
tool for creating outlines for blog posts
Hypotenuse AI Outline Generator (www.hypotenuse.ai/outline-
generator), another free tool that lets you enter your text to generate an
outline

To generate an outline, you have to (1) ask AI to create an outline and (2)
give it the broad specs of what the piece is about. For example, you might
enter a prompt like the following: Create an outline for an article about
pickleball. The AI tool should return a fairly detailed outline like the one in
Figure 4.11, from Google Gemini.

FIGURE 4.11

Using Google Gemini to create an outline for a pickleball article.

Using AI to Write—and Rewrite—Content

I already talked about using all-purpose AI tools to create content in various

styles and formats. There are also AI tools dedicated to writing and
rewriting text in a variety of styles.
The key to using any AI tool to create written content is to be very specific
in the prompts you use. You have to tell AI what you want to write about,
what kind of content you want (a letter, a social media post, a research
report, and so on), how long you want the output to be, and—if you want to
get really particular—the reading level of the generated content. With the
right input, just about any AI tool, including those listed here and the all-
purpose tools discussed in Chapter 3, can output written content appropriate
to your needs.

Note

Turn to Chapter 3 to learn more about using all-purpose

generative AI tools—including ChatGPT, Claude, Google
Gemini, Meta AI, Microsoft Copilot, Perplexity, Pi, and Poe
—as writing assistants.

Some of today’s most popular AI-powered writing tools include the

following:

HyperWrite (www.hyperwriteai.com), a suite of writing tools that

includes AutoWrite, Magic Editor, Outline Generator, and Summarizer;
the basic version (with limited usage credits) is free, and the Premium
 version (with more credits and a more advanced AI model) runs $19.99
per month.
Sudowrite (www.sudowrite.com), an AI tool specifically for fiction
writing; the company offers a free trial and paid versions starting at $10
per month.

Using AI to Edit Content

Even professional writers like me need editors. A human editor proofreads a

piece for punctuation, spelling, and grammar and edits for clarity,
conciseness, style, and accuracy.

Don’t tell my human editors this, but AI tools available today can do just
about everything that human editors do. These tools let you upload a file
(typically in Word or PDF format), analyze that document, and then return
it to you marked up with all sorts of useful suggestions you can accept or
ignore.

Note

These dedicated grammar-and-editing tools offer more

specific analyses and recommendations than all-purpose AI
 tools. I’ve found they’re more effective and easier to use for
that purpose.

Some of the most popular AI-powered editing tools today include

Grammarly (www.grammarly.com) is a tool for editing and proofreading

documents. Today’s Grammarly uses AI technology to identify problems
in a piece and suggest changes. It checks for correct spelling and
punctuation, clarity, passive voice, reading level, and more. Grammarly
offers a basic free version and a more powerful Premium version that
flags more types of errors for $12 per month.
Hemingway Editor (https://hemingwayapp.com), like Grammarly,
analyzes writing, proofreads copy, and suggests changes. To take
advantage of Hemingway’s AI-powered capabilities, subscribe to
Hemingway Editor Plus, which starts at $10 per month.
ProWritingAid (https://prowritingaid.com) fixes common grammar and
punctuation mistakes, eliminates weak words, and makes your writing
more clear and powerful. The limited-use basic version is free; the
Premium version ($12 per month) removes all usage limits.
QuillBot (https://quillbot.com) is an AI-powered paraphrasing tool. Type
or paste a block of text and QuillBot rewrites it for you. You can choose
from various modes, including Natural, Format, Academic, Simple,
Creative, and Shorten. The basic version, with various limitations on use,
 is free. The Premium version, for $8.33 per month, removes all
limitations.
Wordtune (www.wordtune.com) is an AI-based writing and editing tool
that can create new content from scratch, provide suggestions, rewrite
words and long passages, and summarize documents. Wordtune’s basic
version is free but has usage limits; the Advanced version ($6.99 per
month) lets you do more; and the Unlimited version ($9.99 per month)
removes all limits.

Warning

AI-powered editing tools are far from perfect, if only because our language
is far from perfect. You might find an AI editor suggesting changes that
make your text more difficult to read or obscure the point you’re trying to
make. View the suggestions from an AI editing tool as just that,
suggestions, and then rely on your own writing skills to use or ignore the
suggestions as you deem fit.

Examining Popular AI Tools for Writing and Editing

Now that you know what’s out there, it’s time to take a closer look at how
some of the more popular AI writing and editing tools work.
Grammarly

Grammarly (www.grammarly.com) is a popular tool, now enhanced with AI

technology, for improving the grammar in written works. Grammarly lets
you paste text directly into its editor or upload complete files for editing. To
get started, click the New tile on the home page, shown in Figure 4.12.

FIGURE 4.12

Grammarly’s home page.

Figure 4.13 shows the results of an editing session with Grammarly’s free
plan. (The Premium version would make even more suggestions.) The
original article (from my Classic Song of the Day blog) is shown on the left,
with suggestions in the middle column. Click a suggestion to highlight in
the document; if you agree, click Accept to have Grammarly make that
change for you.

FIGURE 4.13

The results of a Grammarly editing session.

When you click the Generative AI button in the right column you get even
more options. You can choose to improve your text, identify any gaps, or
generate more ideas. If you click the Set Voice button, Grammarly lets you
choose a level of formality and tone, as shown in Figure 4.14, and then
makes the appropriate stylistic suggestions.
FIGURE 4.14

Choosing the formality level and tone for a document.

Hemingway Editor

Hemingway Editor (https://hemingwayapp.com) is a grammar and style

checker similar to Grammarly. You can type your text directly into the
home page (shown in Figure 4.15), paste text you’ve copied from
elsewhere, or click the File link at the top of the page to upload a Word or
HTML file for editing.
FIGURE 4.15

The Hemingway Editor’s home page.

As you can see in Figure 4.16, the Hemingway Editor is a tad more
aggressive in its suggestions than Grammarly. It highlights different types
of missteps and suggestions in different colors; adverbs are in blue, passive
voice in lime green, too-complex phrases in purple, and so forth.
FIGURE 4.16

A document analyzed and marked up by the Hemingway Editor.

Click a given suggestion to view more details; if you want Hemingway to

fix it, click the Fix It For Me button. You’re then presented with the
suggested change, as shown in Figure 4.17. (In this example, it wants to
lower the reading level of this passage; it thinks I write too fancy.) To
accept the suggestion, click the Use Suggestion button.
FIGURE 4.17

Viewing a suggested change and letting Hemingway fix it for you.

HyperWrite

HyperWrite (www.hyperwriteai.com) is a suite of text generating and

editing tools. As you can see in Figure 4.18, the HyperWrite’s home page
lets you choose from a variety of tools, including Rewrite Content, Flexible
AutoWrite, Expert Writing Assistant, Text Summarizer, and more. Click a
tile to start using that tool.
FIGURE 4.18

The many tools of HyperWrite.

For example, Figure 4.19 shows the result of HyperWrite’s Rewrite Content
tool. You paste a bit of content (a sentence, a paragraph, maybe more), tell it
how you want it rewritten (simpler, more engaging, and so on), and
HyperWrite rewrites it for you.
FIGURE 4.19

A passage rewritten by HyperWrite to be more engaging.

HyperWrite’s Text Summarizer, shown in Figure 4.20, does just that: It

summarizes a piece of text. Paste a text passage or link to a page on the web
and HyperWrite produces a concise summary of the text in question.
FIGURE 4.20

A HyperWrite summary of a web article.

ProWritingAid

ProWritingAid (www.prowritingaid.com) is a full-featured editing/writing

aid. As you can see in Figure 4.21, you can type text directly into the tool,
paste text from another application, or upload a complete document.
FIGURE 4.21

ProWritingAid’s home page.

Figure 4.22 shows what ProWritingAid’s initial analysis looks like. The
right column scores your text in various categories, such as
grammar/spelling, style, sentence length, readability, and the like. Click a
category header to view a more detailed report of that type.
FIGURE 4.22

ProWritingAid’s initial analysis.

As you can see in Figure 4.23, errors and suggestions are underlined in the
text and detailed in the left column. Hover over a highlight to see
ProWritingAid’s suggestions, and then click the suggested change to make
it or click Ignore to ignore it.
FIGURE 4.23

Accepting or ignoring ProWritingAid’s suggestions.

QuillBot

QuillBot (https://quillbot.com) is an AI tool that paraphrases existing text.

This is useful when your writing is in a bit of a rut and you need to shake
things up. It’s also useful if you’re taking content from elsewhere and don’t
want to plagiarize it; QuillBot can rewrite it just enough to make it an
inexact copy.

QuillBot is also useful for rewriting an existing piece in a different style.

You can choose from a number of different styles for your rewrite, from
natural to academic to creative.
 Note

Some styles are available only if you subscribe to QuillBot’s

Premium plan.

In addition, you can use the Synonyms control to determine whether you
want QuillBot to make more or fewer changes. Note, however, that the
more you change the original text, the less accurate the rewrite will be.

Note

QuillBot can also translate text from English to more than a

dozen languages including French, Spanish, and German.

Figure 4.24 shows QuillBot’s home page. You can either type in the text
you want to rewrite, paste it from another application, or click the Upload
Doc icon to upload an existing document file. Select the style for your
rewrite and then click the Paraphrase button.
FIGURE 4.24

Getting ready to paste or upload the original text into QuillBot.

As you can see in Figure 4.25, QuillBot displays the original text in the left-
hand pane and the rewrite on the right. You can hover over any underlined
text to rephrase it again. Click the Export (down arrow) button to save the
rewritten text as a Word document or click Copy Full Text to copy the
rewrite into another document or app.
FIGURE 4.25

The original text on the left and QuillBot’s rewritten text on the right.

Sudowrite

Sudowrite (www.sudowrite.com) is a set of useful tools for fiction writers.

You can use Sudowrite to do everything from generating story ideas to
rephrasing existing text to writing the whole story for you.

Figure 4.26 shows Sudowrite’s main writing screen. From here you can do
several things:

Write: Continue writing an existing work

Rewrite: Rewrite an existing sentence, paragraph, or longer passage
Describe: Enter a word or phrase to generate suggestions
Brainstorm: Have Sudowrite suggest character names, descriptive
details, dialog, and even plot points
 Plugins: Other writing tools, including First Draft, Shrink Ray, Twist,
Characters, Poem, Visualize, and Feedback

FIGURE 4.26

Examining Sudowrite’s fiction writing tools.

The Write tool is one many fiction writers will find useful. It works by
adding onto text you’ve already started. It analyzes your existing text and
continues the story or other work from there. You can see some typical
results in Figure 4.27.
FIGURE 4.27

Writing new fiction with Sudowrite; the original text is in the middle, and the continuation is on the
right.

Wordtune

The last AI writing tool I want to talk about is Wordtune

(www.wordtune.com), which includes several individual tools, including
Rewrite, Read and Summarize, and Grammar Checker. As you can see in
Figure 4.28, you can directly enter text by typing, paste text from another
app, or upload document files.
FIGURE 4.28

Getting ready to enter, paste, or upload text into Wordtune.

To rewrite a sentence, click within that sentence and then select one of the
options from the top of the screen: Rewrite, Casual, Formal, Shorten, or
Expand. Wordtune displays a list of options as shown in Figure 4.29; click
to select the replacement phrase you like best.
FIGURE 4.29

Rewriting a document, one sentence at a time.

Wordtune can also summarize files you upload. Figure 4.30 shows one such
summary; the original text is on the right and a bulleted summary of key
points is on the left.
FIGURE 4.30

How Wordtune summarizes a document.

Using AI Dictation and Transcription Tools

Artificial intelligence does a good job understanding the spoken word and
transcribing it into text files. If you’re more comfortable speaking than
writing, you can use an AI dictation tool to transcribe your words into a
Word document.

Some of the most popular AI dictation tools today include

AudioPen (https://audiopen.ai)
Fireflies.ai (https://fireflies.ai)
Otter.ai (https://otter.ai)
SpeakAI (https://speakai.co)
Many of these tools are targeted at a business audience that needs to
transcribe and summarize meetings. While some AI, such as Fireflies.ai,
have free or low-cost plans for individuals, others focus on higher-priced
business plans. (Learn more about business transcription services in
Chapter 9, “Using AI at Work.”)

Summary

This chapter was all about using AI to help you find the right words and
write easier and better. That means writing both casually (emails, notes, and
social media posts) and formally (business letters, official correspondence,
and so on).

You learned when you should and when you shouldn’t use AI for writing—
and how to make it transparent when you do use it. You discovered a bevy
of AI tools for generating ideas, crafting outlines, writing and rewriting text,
and editing grammar, punctuation, and writing style. And you learned the
best ways to prompt an AI tool to write specific items, such as texts and
emails, letters, personal memoirs, and poems.

Here’s the bottom line: You don’t have to be a pro to use AI writing and
editing tools. In fact, AI is at its best when it helps regular people write
regular things. If you think you can’t write at all, AI can do it for you. And
if you can write just a little, AI can make your writing read better. It’s a
boon for casual writers. But don’t claim AI-generated writing as your own.
The words are taken from a database and could very well be someone
else’s.

OceanofPDF.com
5

Using AI to Find Information

IN THIS CHAPTER

How AI differs from traditional web search

Why AI is a useful search and research tool
What to beware of when using AI for search and research
Evaluating popular AI tools for information and research
How to focus AI to get the information you want
Using AI to summarize and understand information

Generative AI isn’t just for creating content. You can also use generative AI
tools to find specific information and even conduct research for you, from
finding great summer vacations to determining which car is best for your
family to conducting an in-depth study of 17th-century architecture. It’s
kind of like a web search engine, except (in many cases) better.

When comparing traditional web search with AI, the best analogy is when
you’re looking for information at your local library. Web search is like
flipping through the card catalog that points you to individual books for you
to read. AI is like having the librarian read the books for you and
summarize what they say. But, as with all things AI, remember that AI can
be wrong, so you’ll want to verify the results of all AI output.
How AI Differs from Traditional Web Search

At first glance, you might think that using AI for research seems a lot like
using a web search engine. And you’d be at least partially right.

In terms of user interface, the typical AI tool does indeed closely resemble
the search page of Bing, Google, Yahoo, or other web search engines. The
difference is in the results.

Web search engines are designed to interpret your queries, identify

important keywords, and return a list of web pages that best match those
keywords. This is called keyword matching, These web search engines don’t
analyze or understand your search; they just look for specific word matches.
They don’t actually answer your questions; they merely point you to web
pages that might do that.

Note

Google and other web search engines operate on the

assumption that the more popular a page is, the better it
meets searchers’ information needs. The more visitors a page
has, the higher it ranks in web search results—no matter the
quality (or lack of) of the page’s contents.
The result is a list of matching web pages with no real understanding of
whether they answer your question. The actual work of filtering and
interpreting the data is left to you. Figure 5.1 shows a Google results page
for the query tell me about Mars, full of links to other websites where the
real info resides.

FIGURE 5.1

Google search results for the query tell me about Mars.

Note
 For some queries, Google offers a “snippet” of basic
information, typically sourced from Wikipedia or a similar
site. That’s different from the full and thoughtful response
from scratch you get with AI.

In contrast, an AI generator attempts to understand the context of your

query to answer your question appropriately, typically in the form of a short
report or conversation. It sifts through the relevant sources and presents all
the information necessary to answer your query. Because it does the
research and analysis for you, you get what you were looking for in seconds
with no extra work required on your part (apart from verifying the results).
You don’t even have to use the right keywords. Just convey to the AI tool
what it is you want, and it will find it for you.

Figure 5.2 shows the first page of a lengthy ChatGPT result for the prompt
tell me about Mars.
FIGURE 5.2

ChatGPT response to the prompt tell me about Mars. (This is just the first page; it goes on with a
fairly lengthy response.)

The bottom line is that web search engines are relatively “dumb” and AI
tools are “smart.” The web search engine finds and gathers the sources of
information you ask for, whereas the AI engine works with that information
to generate immediately useful results in plain English. AI gives you what
you ask for and doesn’t make you figure it all out yourself.
AI in Google Search Results

Google has recently started to include AI-generated results at the top of

some search results pages before it offers its traditional web page listings.
These “AI Overviews,” as Google calls them, attempt to give you the exact
answers to the questions you’re asking so you don’t have to sift through
individual web pages as you would for its traditional results. Figure 5.3
shows one such AI Overview on a Google search results page.

FIGURE 5.3

An AI Overview at the top of a Google search results page.

There are several potential problems with these AI Overviews, however.

First, users have found that Google’s AI Overview doesn’t always provide
accurate results. If you accept AI Overview results at face value, you may
be opening yourself up to bad, misleading, or biased information.

Another potential problem arises for those websites that have historically
relied on Google search results to drive traffic to their pages. If users can
get the information they need from an AI Overview, they’ll be less likely to
click through to other pages listed in Google’s search results. That means
much less traffic for sites across the web, potentially resulting in sites
closing down due to lack of traffic—which could result in the slow and
painful death of the web itself.

A final problem with Google’s AI Overview is a challenge for Google itself.

Google derives virtually all its revenue from selling ad space on its search
results pages. If users no longer scroll through and click these “sponsored”
search results, Google could put itself out of business due to rapidly
declining advertising revenue, dramatically altering the way we use
websites and find information on the web. This may be an unintended
consequence of using artificial intelligence to try to answer all the world’s
questions.
Why AI Is a Useful Search and Research Tool

Let’s stop for a moment. Given that generative AI is a technology that can
create all sorts of new content, why might it also be useful for finding
information and conducting research?

There are several reasons why AI might be a superior tool for finding
information, answering questions, and researching specific topics. In
particular, AI excels at the following:

Fast and targeted searches: Yes, Bing, Google, and Yahoo are fast, but
you still have to sift through pages and pages of search results to find the
most relevant results. AI tools can find the specific information you want
in seconds, no manual review necessary on your part.
Understanding what you really mean: Have you ever been stymied in a
search because you just couldn’t come up with the right words for your
query? That’s not a problem with AI tools. They employ natural
language processing to understand the nuances of human language and
accurately ferret out your intent, even if you don’t use the ideal
keywords.
Summarizing and synthesizing information: AI doesn’t just find
information; it analyzes and understands it. AI tools can accurately
summarize complex papers and articles, identifying the main points and
important trends.
 Analyzing data: AI excels at ingesting, arranging, and analyzing data of
all types. You no longer need to spend hours or days poring over Excel
spreadsheets and databases. AI can not only do it faster than you but it
can find patterns and relationships that you may never notice.
Discovering new sources: Web search engines tend to rank highest the
most popular sites on a given subject. These sites might not be the best
sources of information however—just the most visited. AI looks beyond
the obvious to find new sources you may not have found otherwise.
Dive deeper into key topics: The conversational nature of AI encourages
deeper dives into whatever it is you’re researching. You may get an
initial answer from AI that inspires further questions; as you ask and AI
answers, you create a dialog that is increasingly informational.

For many people, the biggest selling point of using AI for search is that AI
tools generate results in complete sentences and paragraphs using the
appropriate level of language for the task at hand. They can even generate
complete results in the form of scholarly articles, papers, and the like,
which can reduce your ultimate workload. That’s significantly more user-
friendly than being forced to poke around through pages and pages of
traditional search results.
Things to Be Careful of When Using AI for Search and
Research

As promising as AI is for search and research, you should be wary of the

results that AI generates. That’s because AI is only as good as the
information on which it’s been trained. If the initial dataset contains false or
misleading information or is biased, AI might quote that bad information as
the truth. Know, however, that these tools are evolving regularly, so it’s
possible that the answer an AI tool gives today could be different tomorrow
—and the accuracy could be different as well.

Note

When an AI tool returns false or misleading information, this

is called an AI hallucination.

For example, I asked ChatGPT how many countries in Africa started with
the letter K. On the first day I asked, AI said there were three: Kenya
(correct), Kiribat (debatable; it’s an island nation off the coast of Africa),
and Comoros (an African nation, yes, but one that most definitely does not
start with the letter K). As you can see, that answer was not wholly correct.
(When I asked a few days later, however, ChatGPT gave the correct
answer.)

Another user asked Google’s AI Overview how many U.S. presidents had
graduated from the University of Wisconsin. AI returned the answer 13,
including Andrew Jackson, who “graduated in 2005”—despite the fact that
this particular former president died in 1845 and didn’t go to college. The
actual number of University of Wisconsin presidential alumni is zero; AI
apparently found graduates who had the same names as U.S. presidents and
counted them, often several times. (According to the AI, John F. Kennedy
graduated from UW in 1930, 1948, 1962, 1971, 1992, and 1993—quite an
accomplishment!)

Note that when I asked the same question of ChatGPT, it got the answer
correct. You may conduct the same search today and get an answer that is
different from the one I got—another example of why it is essential to
check information AI tools provide against proven sources.

AI also has trouble distinguishing satire from fact. For example, when a
user searched Google for “cheese not sticking to pizza,” the AI
recommended adding glue to the pizza sauce. (Don’t try that at home,
folks!) That answer is straight out of an older satirical thread on the Reddit
website, but Google’s AI apparently took that user-generated thread and
treated it as a fact. AI doesn’t always know when others are joking, which
can lead to problems.
AI can also present conspiracy theories as fact. One significant example
also comes from Google’s AI Overview. When asked how many Muslim
presidents the United States has had, it confidently responded that “The
United States has had one Muslim president, Barack Hussein Obama.” This
is, of course, not true (former president Obama was and is a Christian), but
the lie has been propagated on various conspiracy theory sites over the
years—and was obviously ingested as part of the AI’s training set.

Note

In all fairness to the various artificial intelligences involved,

their parent companies immediately corrected these mistakes
when they were pointed out. (This also means you probably
won’t be able replicate the wrong answers I’ve provided in
these examples.)

Finally, let’s not forget the fact that AI can sometimes return answers that
reflect age, race, or gender bias. That bias may be subtle or overt, but it is
often there merely because AI models train on data generated by flawed and
sometimes biased human beings.
These examples demonstrate the need to verify all the information you find
online, especially that proffered by AI tools. To avoid accepting falsehoods
as fact, you need to research the information provided by AI—which you
can do by clicking through the pages in traditional web search results. If an
AI result sounds fishy, look it up with a traditional search engine, then click
through on a reliable, verifiable source.

If you’re doing serious research, you need to do this due diligence anyway.
Don’t accept an AI result as the final answer or definitive source. Always
verify sources and include those sources in your research. AI tools are not
sources to cite in your research papers; always go to the original source and
cite that.

Evaluating Popular AI Tools for Information and Research

Now to the meat of the matter: Which AI tools are best for searching for
information and conducting research? There are two ways to go: all-purpose
AI tools or tools specifically fine-tuned for research.

Using All-Purpose AI Tools for Search and Research

To begin with, I direct you to the all-purpose generative AI tools discussed

in Chapter 3, “Getting Started with All-Purpose AI Tools.” All the tools
presented there—ChatGPT, Claude, Google Gemini, Meta AI, Microsoft
Copilot, Perplexity, Pi, and Poe—can be used to find information and
conduct research. All you have to do is enter a prompt similar to a search
engine query and the AI tool will present the information it finds that
answers your question.

For example, if I want to find out how many voters were in Minnesota, I
might ask a tool like ChatGPT how many voters are there in the state of
Minnesota? ChatGPT would generate a list of results like the one in Figure
5.4.

FIGURE 5.4

Using ChatGPT to find information about Minnesota voters.

You can also ask some AI models to graph certain data. For example, I
asked ChatGPT to graph the number of people in each major age group in
the United States. The result is shown in Figure 5.5.
 Note

Not all AI models currently have graphing capability.

FIGURE 5.5

An informational graph created by ChatGPT.

Similarly, if I want to find out how to convert kilometers to miles, I might

ask Google Gemini how do you convert kilometers to miles? Google
Gemini would then present the answer shown in Figure 5.6.
FIGURE 5.6

Using Google Gemini to convert kilometers to miles.

If you want to conduct more comprehensive research, an all-purpose AI tool

can still do the job, but you have to create more specific prompts. For
example, if you want to research the incidence of left-handedness among
the general population, you might enter the following prompt into Meta AI:
write a 2,500-word research paper detailing the incidence of left-
handedness among the U.S. and world populations. The AI tool would
then write such a comprehensive paper, complete with footnotes and
citations, the beginning of which is shown in Figure 5.5.
FIGURE 5.7

Using Meta AI to write a research paper about the incidence of left-handedness.

Note

Perplexity (www.perplexity.ai), which I also discuss in

Chapter 3, is a good choice for conducting basic research.
Unlike other all-purpose AI tools, Perplexity displays the
 sources for its results at the top of the response page. To go
directly to a source, just click the link.

Using Research-Specific AI Tools

While all-purpose AI tools might serve most people’s search and research
needs, there are also some AI tools that focus specifically on scholarly and
scientific research and the needs of professional researchers.

Table 5.1 presents key information about some of the most-popular

research-specific AI tools. A more in-depth overview of these tools follows.

TABLE 5.1 Popular AI Research Tools

Paid
Plans
Free Key
AI Tool URL (cost
Plan? Features
per
month)

Consensus https://consensus.app Yes $8.99 Uses

ChatGPT 4
model,
 Paid
Plans
Free Key
AI Tool URL (cost
Plan? Features
per
month)

summarizes
results

Elicit https://elicit.com Yes $10 Uses

research
“notebooks,
summarizes
top papers,
links to
original
sources

Scholarcy www.scholarcy.com Yes $9.99 Reads,

summarizes
extracts, and
organizes
key
information
 Paid
Plans
Free Key
AI Tool URL (cost
Plan? Features
per
month)

Scite https://scite.ai No $20 Analyzes

and
summarizes
scientific
articles,
highlights
where
papers are
cited

Consensus

Consensus (www.consensus.app), shown in Figure 5.8, bills itself as an “AI

search engine for research.” This is exactly what it is: an AI tool that
gathers scientific information from peer-reviewed articles and other
published sources.
FIGURE 5.8

The home page of the Consensus AI research engine.

The basic version of Consensus is free but offers limited usage. The
Premium Plan, priced at $8.99 per month, provides unlimited access to
OpenAI’s GPT-4 search model and other features.

To use Consensus, enter your query into the Ask a Research Question box
on the home page. Consensus then gathers relevant sources of information
and provides a summary of this information, as shown in Figure 5.9.
FIGURE 5.9

Summary and research results from Consensus.

Elicit

Elicit (https://elicit.com) is, as the site claims, an AI research assistant. This

AI tool can process and analyze data, revealing trends, patterns, and themes
in the data.

The Basic version is free but limits how much you can use it. The Plus plan
has higher usage caps, the ability to summarize up to eight papers, and the
ability to extract information from tables, for $10 per month.
To use Elicit, you first create a “notebook” for your research, as shown in
Figure 5.8. Then you enter a query into the Ask a Research Question box
and press Enter.

Elicit then provides a summary of the top papers on the subject, as shown in
Figure 5.11. This is helpful for researchers who need a quick overview of
key research. Elicit also lists relevant papers below the summary, complete
with links to the original sources.

FIGURE 5.10

Using Elicit to conduct research.

FIGURE 5.11

Summarizing relevant research papers with Elicit.

Scholarcy

Scholarcy (www.scholarcy.com) is an AI tool that reads, summarizes,

extracts, and organizes key information from scholarly articles. The basic
plan, which lets you generate just three summaries per day, is free.
Scholarly Plus is more useful, with unlimited and enhanced summaries for
$9.99 per month.

To summarize an article, go to the Article Summarizer page, shown in

Figure 5.12. Click the Click to Import button to select a file to summarize.
You can upload specific files or enter a URL to summarize a web-based
document.
FIGURE 5.12

Importing a file to summarize with Scholarcy.

Scite (https://scite.ai) imports the document, analyzes it, and produces a

summary like the one shown in Figure 5.13. Scroll past the summary to
view key concepts, abstract, synopsis, highlights, a comparative analysis,
and more.
FIGURE 5.13

Viewing a Scite summary of a scholarly article.

Scite

Scite analyzes and summarizes scientific articles. It’s unique in that it

highlights how and where a given research paper is cited, which helps you
gauge the importance and reliability of each citation.

No free version is available, although Scite does offer a 7-day free trial. An
individual subscription runs $20 per month.

The easiest way to use Scite is with the Scite Assistant, shown in Figure
5.14. Enter your question into the Ask a Question box and press Enter.
FIGURE 5.14

Initiating research with the Scite Assistant.

As you can see in Figure 5.15, Scite displays a three-paned results page.
Your query is in the left pane and Scite’s summary of results is in the
middle pane. The right pane displays all the references used; you can click
any reference to read the full text or perform other operations.
FIGURE 5.15

Scite’s research results, including citations.

How to Focus AI to Get the Information You Want

Prompting AI to search for information requires some subtle adjustments to

the prompting strategies discussed in Chapter 3. Read on to learn how to
best focus AI tools to retrieve the information for which you’re looking.

Crafting the Perfect Prompt for Research

The primary way to focus any AI tool for searching and researching is in
fine-tuning the prompts you enter. Here are some tips for crafting effective
prompts for finding and researching information:

Make your query as clear and specific as possible. The more accurately
you can describe what you’re looking for, the better AI can understand
your intent and deliver relevant results. For example, instead of
prompting AI to simply research restaurant options, ask it to research
dinner menus for a family of four with two kids under age five, offering
specific cuisine or dishes, and within a specific price range.
Use keywords. Although generative AI doesn’t use keyword matching,
including important keywords in your prompt helps focus AI on what
you’re looking for. As an example, if you want to know the effects of
climate change on farming, instead of prompting how is climate change
impacting farming, include related keywords as in the following
prompt: how is climate change impacting farming in terms of crop
yield, agricultural land use, and water scarcity. By including the
keywords “crop yield,” “agricultural land use,” and “water scarcity,”
you’ll get more targeted results.
Use the proper technical vocabulary. When researching technical
topics, it helps to know the lingo. AI models respond better to prompts
that use the proper technical or scientific terms. Using the right technical
language (if you know it) also helps inform the AI model about what
type and level of results you expect. For example, a layperson might ask
AI to explain how heart surgery works, whereas an expert might describe
the process as coronary artery bypass grafting.
Provide relevant context. If you’re writing a technical paper, tell the AI
tool. It will tailor its results for that kind of output. If you have a specific
length you need to hit or not exceed, include that in your prompt, as
well. For example, if you’re writing a scientific paper on insect
 migration and it needs to be 5,000 words long with an abstract and at
least a half dozen citations, say write a 5,000-word scientific paper on
insect migration with an abstract and at least 6 citations.
Tell AI exactly what you want. If you provide specific instructions, AI
will generate results in a general way that it thinks might be most
helpful. If you want something more specific, you need to ask for it. You
might want to tell AI to write at a specific grade level or present results
in list form or as an executive summary. If you don’t tell it what you
want, it won’t know to give it to you.
Leverage advanced search options. Some research-oriented AI tools
offer advanced search options, such as filtering by source type, date
range, or even specific domains. Take advantage of these tools to
generate a more relevant response.

It’s important that you don’t rely exclusively on AI to conduct your

research. As good as AI is (and will continue to improve in the future), it
shouldn’t totally replace traditional research methods. Consider AI just one
more tool in your kit, along with research databases and websites, for
conducting serious research.

Warning
Because most all-purpose AI tools generate content without citing sources,
it’s possible they could be working from bad or outdated content. When
doing scholarly or professional research, it’s imperative that your work be
both accurate and citable. That means diligently cross-checking AI-
generated results against verifiable results from reliable sources. And you
can ask the AI tool to cite sources; after it generates results, follow with the
prompt, what are your sources. You can then double-check them manually.

Examples of Using AI for Research

What does the ideal prompt look like for specific types of searches? Here
are some examples.

Researching a Major (or Minor) Purchase

You can use AI to help you decide what items to purchase. This is
especially useful when researching a major purchase, such as a house or a
car, but can be used to evaluate virtually any product or service. For
example,

Comparing two or more items: Compare the 2024 Honda CRV with the
2024 Nissan Rogue.
Listing the pros and cons of a given product: List the pros and cons of
this electric razor.
Evaluating the value of an item: How much is the property at 1234
Main Street worth?
 Asking for specific recommendations: What is the best 65" OLED TV
on the market today?

Researching How to Perform a Task

Ever get stuck with having to do something that you don’t know how to do?
You can ask AI for instructions—but make sure you’re as specific as
possible. For example,

Cooking: What’s the best recipe for beef stroganoff?

Repairing: How do I replace the condenser coil in my air conditioning
unit?
Creating: How do I create an origami penguin?
Driving: What’s the best route to drive from Chicago to Memphis?

Note

When searching for recipes, you can fine-tune your prompt

to include any dietary restrictions, number of servings needs,
and the like. For example, you could enter a prompt like find
a gluten-free recipe for shrimp etouffee that serves 6
people.
Researching a Current News Event

Finding information about a current news event is a bit different than

researching higher-level topics. What’s important here is finding the most
current information and ensuring it doesn’t come from a biased source. For
example,

Seeking official statements: Are there any official statements regarding

the recent rule changes in major league baseball?
Reporting the latest information: Who is running for office in my
district?
Seeking out diverse viewpoints: What are the differing opinions on the
upcoming ballot proposition?
Examining a person’s major accomplishments: What are the key events
that shaped Michael Jordan’s life and made him the person he is
today?

Note

Because AI engines train on large language models and it

takes time to assemble those vast amounts of data, AI models
may not be as up to date with current information as more
traditional web search engines. It’s not unusual for
 information garnered within the past few months to not yet
be available to today’s AI tools.

Researching Factual Information

AI is also great for researching specific facts. For example,

How far is it to the moon?

What is the square root of 12,345?
What is Quentin Tarantino’s birthday?
What is the adult population of Milan, Italy?

Warning

As mentioned in Chapter 2, AI results can sometimes exhibit biases built

into the training data or algorithms they use. Don’t assume that AI is always
neutral and unbiased; be on the lookout for institutional or societal age,
gender, race, and other biases in the results generated by any AI engine.

Researching a Scientific or Technical Topic

When researching a scientific or technical topic, it helps to be as specific as

possible in your prompts. For example,
 Generating a high-level overview or summary: Provide a high-level
overview of the field of quantum mechanics
Summarizing recent developments in a field: What are the current
challenges in nuclear fusion?
Identifying important topics: What are the most important topics
concerning electric vehicles today?
Citing credible sources: List the top five peer-reviewed articles on
astrobiology.

Researching Historical Events

You can take different approaches to doing historical research, depending

on what information you want. For example,

Examining the causes of an event: What caused the War of 1812?

Examining the impact of an event: What was the economic impact of
the Cold War?
Examining the key players of an event: Who were the key figures
during the reign of King Henry VIII?
Comparing two events: How did the American Revolution compare to
the French Revolution?

Fine-Tuning AI Results

When you use the right prompt, you might be perfectly satisfied with the
first results returned—or you might find the results are not totally related to
your question, not specific enough, or just plain wrong. Fortunately, when
AI’s first results aren’t quite right, you can work to improve your query and
fine-tune the next round of results.

Here are some things to try:

Ask for more or less detail

Provide more information and context
Specify the level of the results (general or technical audience,
grade/reading level, age level, and so on)
Specify the length (in words or pages) of the desired results
Specify the type of result you want (article, paper, blog post, and so on)
Specify from where you want the AI to pull information (specific
websites, media sources, journals, and so on)

The key is to iteratively refine your search. Evaluate AI’s first response and
then adjust your prompt and see what you get.

Using AI to Summarize and Understand Information

In addition to searching for and researching information, you can also use
AI to summarize and help you better understand that information. You can
use AI to summarize books, movies, technical articles, legal documents,
even meetings. The key is to either use the proper prompt for an all-purpose
AI tool or choose a tool that specializes in summarization.
When using an all-purpose AI tool, you need to include the word summarize
in your prompt. For example, you could ask Google Gemini to Summarize
the plot of the play Our Town, and you’d get something like what’s shown
in Figure 5.16.

FIGURE 5.16

Google Gemini’s summary of the plot of Thornton Wilder’s Our Town.

When it comes to summarizing specific documents, such as legal contracts,

you want an AI tool that lets you upload files for analysis. Of those tools
discussed in this chapter and in Chapter 3, Claude, Elicit, Perplexity, and
Scholarcy all let you upload or link to articles and then summarize them.
You could also choose an AI tool designed specifically to summarize
documents. Some of the most popular of these specialized AI tools include
AskYourPDF (https://askyourpdf.com), Sharly (https://sharly.ai), and
SummarizeBot (www.summarizebot.com).

Note

Learn more about AI tools for summarizing meetings in

Chapter 9, “Using AI at Work.”

Summary

Whether you’re a casual searcher or professional researcher, AI can be of

great value. In this chapter, you learned why AI is a useful tool for
searching and researching and how it differs from traditional web search.
You learned how to use all-purpose AI tools for research and discovered
some of today’s most popular AI tools for information and research,
including Consensus, Elicit, Scholarcy, and Scite. You also learned about
some AI tools designed specifically to summarize documents, including
AskYourPDF, Sharly, and SummarizeBot.
Finally, you learned how to focus AI tools to find the information you want
and need. You learned how to craft more effective prompts for research and
how to edit your prompts to fine-tune the results you receive. You even
learned the best ways to prompt AI tools for specific situations.

Using AI for searching and researching can be a real time-saver and game
changer. But as with all things AI, you need to ensure that the AI output is
accurate and free from falsehoods and bias. You shouldn’t blindly accept
everything AI tells you as fact—just as you wouldn’t accept everything the
neighborhood gossip tells you as the truth. You can trust AI, to an extent,
but you should verify its results. Trust but verify is the right approach.

OceanofPDF.com
6

Using AI to Connect with People and Pursue

Interests

IN THIS CHAPTER

Using AI to keep up with friends and family

Using AI to find and pursue hobbies and interests
Using AI to improve your social media communications
Conversing with AI by text or voice

AI can be a powerful tool to help you bridge personal connections and

explore interests and hobbies. Social media networks already use AI to
recommend compatible friends and fill your feed with posts you might like.
You can also use AI within and outside those networks to better connect
with friends and family and find new online communities of people who
share your interests. But don’t get too personal with AI. Any personal
information you supply feeds directly into the tool’s next training database
—and can possibly be used against you in the event of a data breach.

Using AI to Keep Up with Friends and Family

Let’s start with how you can use AI to better connect with family and
friends. AI can help you improve your communication, understand others’
viewpoints, and stay in touch with those you love.

Ways to Use AI to Connect with Your Friends and Family

There are many ways you can use generative AI to connect with your
friends and family, whether via email, text messaging, social media, or in
person. Here are just a few ideas, all doable with any all-purpose AI tool,
such as ChatGPT, Meta AI, or Microsoft Copilot:

Plan an event: When you want to get a group of people together and do
something fun, prompt AI with what are some suggestions for a game
night? or what are some suggestions for a themed party for a bunch
of 60 somethings? You can also specify the location, such as what are
some suggestions for a group event at a local park? or what are some
suggestions for a co-worker evening at my house?
Find a conversation starter: Just ask AI, what are some good
conversation starters? You can also be more specific about the
composition of the group, such as what are some good conversation
starters for a group of 30 somethings? or what are some good
conversation starters for an all-generation family gathering?
Play a game with AI: Your friendly neighborhood AI tool can play all
sorts of text-based games, from 20 questions to trivia games. You might
prompt AI this way: Let’s play a game of movie trivia or Play a
memory challenge game. To find out what games a given AI tool can
play, use the prompt: What games can you play? You can also specify
 whether you’re looking at multiplayer games, one player versus the
machine, and the like.
Get recommendations for activities: Use AI to generate
recommendations for TV shows, movies, music, or games based on your
own or a friend’s or relative’s preferences. For example, you might
prompt, what are some good shows to watch with a friend who likes
rom-coms?
Tell some jokes: Why so serious? Ask your AI tool for a list of jokes or
anecdotes to lighten the mood in any gathering. For example, you might
prompt your AI with generate a list of jokes about mothers-in-law.
(Don’t use that last one if your mother-in-law is actually around.)
Share your genealogy: Use an all-purpose AI tool or an AI-powered
genealogy tool like Ancestry (www.ancestry.com), FamilySearch
(www.familysearch.org), or MyHeritage (www.myheritage.com) to
explore your family’s history and share what you find with family
members.

Using AI to Share Memories

Another way to better connect with friends and family is to use AI to help
you share memories. Here are some ideas:

Upload a batch of photos to an AI image generator (see Chapter 7) to

create a collage, slideshow, or virtual scrapbook.
 Upload old photos to an AI image generator to touch up to modern
standards or apply a special effect, like sepia tone or turning photos into
cartoon.
Feed your memories into an all-purpose AI tool and ask it to turn those
memories into a compelling story.

Note

Learn more about AI image generators in Chapter 7, “Using

AI to Create Art and Images.”

Using AI to Better Understand People

One challenge with communicating with someone else is that you might not
completely understand each other. It could be that you’re from different
generations, cultures, or backgrounds with different influences. Or maybe
you live in different parts of the country. Or maybe you just don’t know
each other that well. (How well do you know your second cousin on your
mother’s side?)

Here’s where AI can help. You can start by describing the person of interest
and the situation you’re in and then asking AI for advice on how to proceed.
AI can also analyze text and conversations with that person to get a better
handle on their personalities, their likes and dislikes, and what the two of
you might have in common.

As a first example, you might use a prompt like the following:

I’m going to visit with my uncle who lives in Alaska. We haven’t seen each
other in five years and don’t have a lot in common. He likes outdoor
activities and works as a mechanic. I’m not too outdoorsy and have an
office job. What should I know about him that might help when we get
together?

Figure 6.1 shows how Claude AI recommends you proceed.

FIGURE 6.1

Some get-acquainted advice from Claude AI.

For more personalized analysis, try pasting or uploading one or more

letters, email messages, social media posts, or texts from this person into
the AI tool and ask it to provide a personality analysis. Precede these
samples with the following prompt:

Analyze these posts and tell me what that person is like.

 Warning

Protect your privacy by removing addresses, phone numbers, and other

personally identifiable information from any information and items you
upload to an AI tool.

Figure 6.2 shows a sample personality analysis from ChatGPT.

FIGURE 6.2

A personality analysis from ChatGPT.

You can also use AI to do a “sentiment analysis” of the same text samples to
judge the emotional tone of the messages. Use the prompt: Perform a
sentiment analysis of the following messages. ChatGPT’s analysis is in
Figure 6.3. (It’s a much different thing than a personality analysis.)

FIGURE 6.3

A sentiment analysis from ChatGPT.

You can go even further by then asking AI for advice on the best way to
respond to that person. Use the prompt how should I respond to this
person? The AI tool will write you a response that’s appropriate to the
person’s personality and sentiment.
Another approach is to ask AI for advice on how to deal with a given
person, especially if there’s some sort of conflict or communication issue
involved. Here’s an example of a situation that might be worthy of some AI
advice:

My cousin and I have never gotten along. He’s a blue collar guy and I’m an
information worker. He’s a Republican and I’m a Democrat. He’s a country
boy and I’m a city gal. How can we learn to get along?

Figure 6.4 shows Google Gemini’s advice.

FIGURE 6.4
Relationship advice from Google Gemini.

Note

As with all things, AI is only a tool you can use to improve

your personal connections. It’s not meant to replace those
connections. Take the effort to spend quality time with your
loved ones—and cherish that time together.

Use AI to Find and Pursue Hobbies and Interests

AI can be a great resource for pursuing your current hobbies and finding
new ones. It’s all about asking the right questions.

Finding New Hobbies

If you’re not sure what hobbies might interest you, AI can help.

One approach is to strike up a conversation with an AI chatbot about your

interests and let the AI recommend online communities that might include
some like-minded individuals.
 Warning

Before joining online communities, check them out as you would real-
world communities. Enter the community’s name into a search engine and
see what appears in the search results. See if the community actually exists,
has recent posts (if not, it may be a “dead” community), and what others
say about it.

Another approach is to ask an AI tool directly, based on key information

about you. Tell the AI how old you are, your gender, where you live, when
you have free time (and how much of it), whether you like indoor or
outdoor activities, and, if you like, how much money you have to spend on
a new hobby. You should also mention whether you have any specific skills
or equipment you want to take advantage of as well as any limitations that
might get in the way.

That’s a lot to remember and a lot to input. You might end up with a prompt
like this:

I’m a middle-aged professional with a fair amount of disposable income. I

live in the suburbs and have lots of free time on weekend afternoons. I like
outdoor activities and have money to spend on them. I do have asthma,
however, so that could limit things. Can you recommend some hobbies for
me?

For this particular person, Meta AI had some very specific

recommendations, that you can see in Figure 6.5.

FIGURE 6.5

Some specific hobby recommendations from Meta AI.

Here’s another example:

I’m looking for a new hobby. I’m a woman just out of college. I’ve always
had a passion for music, especially oldies. I don’t have a large budget
because I spend a lot on rent. I like to get away from the city on weekends
and do things with my friends. What would you recommend?

Figure 6.6 shows what ChatGPT recommends.

FIGURE 6.6

Some hobby recommendations for a cash strapped 20-something woman from ChatGPT.

Or this one:
I just retired and am looking for things to do where I can meet new people. I
try to stay in shape and like taking walks, exploring new places, and helping
out in the community.

Figure 6.7 shows what Google Gemini recommends. Note that it

automatically picked up my location and recommended real activities
nearby.

FIGURE 6.7

Some recreational recommendations for a retiree from Google Gemini.

Pursuing Existing Hobbies

AI can also help you pursue your existing hobbies. Use the same techniques
discussed here to find topic-focused groups online and off and to find more
information about your hobby.

For example, if you like to cook, you can use AI to find new recipes. Just
use a prompt like this one:

Find some easy recipes for noodle-based dishes.

Figure 6.8 shows what Claude AI came up with.

FIGURE 6.8

Some easy noodle recipes from Claude AI.

Or maybe you’re big into arts and crafts. I have a friend who likes to make
candles for herself and friends; she might use a prompt like this:

How can I combine aromatherapy and candle making?

Figure 6.9 shows what Meta AI recommends.

FIGURE 6.9

Candle making advice from Meta AI.

Gardening is another popular avocation for many. If you’re looking to
expand your home garden, you could use a prompt like this:

What native flowering plants are good to plant in Nebraska?

Figure 6.10 shows some suggestions from Microsoft Copilot.

FIGURE 6.10

Some gardening suggestions from Microsoft Copilot.

Although AI will probably never be as informed as an avid hobbyist, you

might be surprised what information it can turn up. Consider AI as another
tool to use in the pursuit of your favorite hobbies.
Finding Online Communities

You can also use AI to find others who share your interests and hobbies in
online communities.

The easiest and most obvious way to find specific online communities is to
ask an AI tool. Just enter a prompt that asks for recommendations based on
a specific interest or topic, like the following:

Can you recommend online communities for model train enthusiasts?

The AI tool should return a list of promising communities, like the one in
Figure 6.11.

FIGURE 6.11

Some recommendations for model train enthusiasts, from Microsoft Copilot.

 Note

Learning a new language or traveling? You can also use AI

to help converse with people who speak other languages.
Learn about AI-powered translation tools in Chapter 10,
“Using AI to Manage Your Travel and Transportation.”

Using AI to Improve Your Social Media Interactions

An all-purpose AI tool can be useful when you’re communicating on social

media, too. Whether it’s a short message on Instagram or X or a longer one
on Facebook or LinkedIn, AI can generate posts and responses for you,
taking some of the work and worry out of keeping up with your social
feeds. AI can detect the tone and context of what others are posting and
respond in kind.

Just tell the AI tool which social network you’re posting to and what you
want to post about. A typical prompt might look something like this: Create
a Facebook post asking for the best deep dish pizza in the
neighborhood.
And, if you don’t know how to respond to someone else’s social media
post, let AI respond for you. Just enter the prompt how should I respond to
this? followed by the original message pasted into the prompt box. The AI
tool will suggest an appropriate response.

In addition, many social media companies are beginning to incorporate AI

into their platforms. It won’t be long before you see some sort of “write
with AI” button on Facebook, X, and other social networks. When
available, it will help you be more social online.

Beware Deepfakes and Clickbait on Social Media

As you learned in Chapter 2, “The Risks and Benefits of AI,” it’s easy to
use AI to create “deepfake” audio, video, images, and text that look real but
aren’t. Malicious individuals can use these images to attract attention in
social media posts and possibly get you to take actions you might not
otherwise take or believe things that aren’t true.

Some of these deepfakes purport to be of celebrities, maybe less-than-fully

clothed or doing something unusual. Others are of children asking for help
or financial donations. Others are of attractive people who say they like
your profile and would like to be friends.

It’s also common for individuals and entities to use AI-generated deepfakes
as propaganda, especially in election years. Don’t be fooled by a picture of
a given candidate doing something questionable; it’s probably a deepfake
meant to confuse or negatively influence voters.

Even if the post doesn’t result in a scam, just clicking it racks up traffic for
the individual who posted it. That can earn the individual advertising
money, thus helping that person profit from their AI-generated deepfakes.

No matter how hard the social media platforms try to guard against these
scams (such as labeling content as AI-generated), the criminals always seem
to stay a step ahead. The best advice is not to take anything you see on
social media at face value, even if it’s a repost from someone you know.
Question and verify everything you see and hear; with AI technology
continuing to advance, you simply can’t believe your own eyes and ears.

Conversing with AI by Text or Voice

Talking with AI by text or voice is an efficient way to communicate and get

the information you need. It can provide you with useful information and
teach you about an array of topics.

You probably already know how to communicate with Alexa, Google

Assistant, and Siri voice assistants, by asking basic questions and getting
appropriate answers. But the personal conversations that you can have with
generative AI today go far beyond that.
Today’s AI tools can carry on conversations that are surprisingly lifelike,
responding not only to your words but also your tone and context. You may
even find that talking to an AI chatbot about some topics is more
comfortable than talking to a human being; AI won’t judge you or make
you feel uncomfortable.

Not surprisingly, some people even use AI to feel connected in today’s

hypersocial online world. For example, some caregivers are using AI to
“talk” with and provide companionship to the people for whom they
provide care.

Note

Learn more about AI and caregiving in Chapter 12, “Using

AI to Help Caregivers.”

There are many ways to engage in AI-powered conversations. You can talk
to most all-purpose AI tools via text, just as you would with a friend via text
messaging. Even better, some AI tools offer voice-based conversation,
which can feel just like talking to someone in the room.
Talking with AI via Text

Most all-purpose AI tools, such as Claude AI, Google Gemini, and Meta AI,
offer text-based conversations with their chatbots. This type of interaction
resembles a conversation with someone via text messaging or social
messaging posts.

With these text-based chatbots, you type your side of the conversation into a
prompt box and the ongoing message thread between you and AI is
displayed in a scrolling text message–like fashion. You type, AI responds,
you respond to the AI, AI responds to your response, and on and on. Figure
6.12 shows such a conversation with Pi AI.
FIGURE 6.12

Carrying on a conversation with a sympathetic Pi AI.

Warning

When chatting with an AI tool, never reveal intimate personal details,

sensitive business data, or any private information like credit card and
Social Security numbers.
Talking with AI via Voice

The latest thing in AI is a new generation of is AI tools that listen to your

spoken words and reply in synthesized human-sounding voices. This ability
to listen to the spoken word and respond accordingly is called multimodal
AI because AI can interact via multiple modes of input and output—text,
voice, images, and (in some cases) physical gestures.

You might think that multimodal AI is similar to voice-activated digital

assistants such as Amazon’s Alexa, Apple’s Siri, or the Google Assistant,
and you’d be partially correct. While the interaction might appear similar on
the surface, the original crop of digital assistants aren’t really very “smart.”
Current digital assistants use natural language processing to understand the
words you’re saying but they really don’t parse your input to understand the
meaning behind your words. They recognize certain phrases (such as “turn
on the living room lights” or “play my oldies playlist”) but they don’t and
can’t go much beyond that.

Multimodal AI, in contrast, goes beyond simple phrase recognition to

understand and respond to your voice in a more comprehensive manner.
When you speak to a multimodal AI tool, it doesn’t just process your words;
it takes into account your tone of voice, your past comments, even (in some
cases) your facial expressions. You end up with a richer, more intuitive,
more natural interaction that makes the original digital assistants feel like
crude robots. (Which, in reality, they are.)
One of the current leaders in voice-activated multimodal AI is ChatGPT,
with its ChatGPT-4o engine. ChatGPT-4o interacts with you via your
smartphone. You talk to it and it talks back to you. It even uses vocal
inflections and pauses to sound just like a real person.

When you first launch the ChatGPT app on your phone (available for free
from your phone’s app store), you’re asked to choose a voice for the AI. As
you can see in Figure 6.13, you can choose from four voices—two male
(Ember and Cove) and two female (Breeze and Juniper).

FIGURE 6.13

Choosing a voice for the ChatGPT AI.

When you’re ready to talk with ChatGPT, tap the headphone icon (not the
microphone icon!) to the right of the Message text box, shown in Figure
6.14. This opens the Voice Chat screen.

FIGURE 6.14

Tap the headphone icon to talk with ChatGPT AI.

You can now start talking with ChatGPT. As shown in Figure 6.15, it listens
to what you have to say and, when you pause, responds in the voice you
chose. Keep talking to continue the conversation.
FIGURE 6.15

Talking with ChatGPT AI.

Talking with ChatGPT and other multimodal AI tools will never replace
talking with a friend or loved one, but it does make it easier to get the
information you need in a way that feels natural. After a while, you may
find yourself forgetting that you’re talking to AI and feel like you’re talking
to another human being.
 Note

Not to be outdone, Apple has announced that it will

incorporate OpenAI’s ChatGPT into its Siri digital assistant
as part of its Apple Intelligence initiative to utilize AI in
many of its products and services. Amazon is also said to be
developing multimodal AI capabilities for its Alexa digital
assistant and Google is developing its own multimodal AI,
based on its Gemini AI model, called Google Astra. Expect
these and other new multimodal tools to go live as soon as
feasible.

Summary

In this chapter, you learned the many ways you can use AI to connect with
friends and family, improve relationships and communications, discover
new hobbies and pursue existing ones, and find new communities of like-
minded people. You also learned how to use AI in your social media
interactions and how to talk with multimodal AI tools that respond to your
voice and respond in their own voices.
It’s fascinating just how human-like some AI tools can appear—even if
they’ll never replace hanging out with a friend in person. But take caution;
always verify information you see and never reveal anything personal that
you don’t want to ultimately go into a huge public database.

OceanofPDF.com
7

Using AI to Create Art and Images

IN THIS CHAPTER

How AI generates images

How to use an AI image generator
What kinds of images can you generate with AI?
How to ethically use AI image generators
Comparing popular AI tools for creating images
Crafting the perfect prompt to generate perfect images

You’ve seen how AI can generate letters, social media posts, and other text-
based content, but some AI engines can create images, both realistic and
fantastic, based only on users’ prompts.

You can generate your own images with AI image generators available to
the general public, many for free. With them, you can create real-looking
pictures of what you might otherwise have only imagined in your mind—
and it’s easy to do.

So read on to learn how to use AI to have fun and expand your creativity, no
matter your artistic abilities!
How AI Generates Images

Let’s start with a quick look at the technology behind the phenomenon.

Like all artificial intelligence tools, AI image generators are “trained” on

large sets of data—datasets consisting of billions of images and descriptions
of those images. Once trained, AI can generate new images based solely on
text prompts.

The images you create with AI can be of anything or anybody that the AI
has learned about in its training. By training on billions of images of all
sorts and styles, it can create artwork that looks painted, drawn, or
photographed. It can mimic a variety of art styles and depict characters and
scenes that might or might not exist in reality. You want a picture of a baby
elephant hang gliding over an active volcano? AI knows what a baby
elephant looks like, what hang gliding looks like, and what a volcano looks
like. Thus, AI can create what looks like a photograph of just that.

Don’t believe me? Just look at Figure 7.1.

FIGURE 7.1

An AI-generated image of a baby elephant hang gliding over an active volcano, courtesy of Microsoft
Designer’s Image Creator.

How to Use an AI Image Generator

AI image generators work more or less like other AI tools. You enter a text
prompt that describes the image you want, and then AI creates that image.
The more detailed your prompt, the more accurate the image.

For example, if you’d like a painting of San Francisco’s Golden Gate

Bridge at sunset to put on a throw pillow, you might enter a prompt like
this: Create a watercolor painting of the Golden Gate Bridge at sunset.
(Figure 7.2 shows such a creation.)

FIGURE 7.2

An AI-generated watercolor painting of the Golden Gate Bridge at sunset, courtesy of Adobe Firefly.

Your prompts should describe the type of output you want (photo, painting,
drawing, and so on) and what you want illustrated. For example, Figure 7.3
shows the result of the prompt photo of couple walking through the
woods in the fall.
FIGURE 7.3

An AI-generated photograph of a couple walking through the woods, courtesy of Midjourney.

Note that with some image generators, you only have to describe the image.
With others that also can generate text-based content, you may have to
preface the prompt with create an image or generate an image.

Most AI image generators let you download or share the artwork they
create. In some cases, you may need to subscribe to a paid plan to generate
and download images in high resolution.
If you don’t like the results, tweak your prompt and try again. Some image
generators even include a “try again” button that automatically generates a
different image based on your initial prompt.

Recognizing AI-Generated Images

It’s likely you’ve scrolled through your social media feed and seen what
looks like a picture of a person or celebrity or fantasy figure that you’ve
never seen before. This picture may look slightly unusual, such as a
different actor wearing a Superman costume, a well-known female celebrity
wearing the briefest of swimsuits, or someone you recognize from TV
doing something crazy or stupid. The picture might appear, at first glance,
to be a photograph but, on closer examination, it looks slightly off—maybe
too realistic, with incorrect details, or just difficult to imagine. When you
see one of these realistic-but-questionable images, chances are it’s AI-
generated. The image may not even be questionable; it may appear to be an
action shot at a sports game or a news photo shot in the field but still be
generated by AI.

How can you tell whether an image is real or AI-generated? As AI image

generation gets better and better, you can’t always tell. I covered some ways
to recognized AI-generated images in Chapter 2, “The Benefits and Risks of
AI.” In general, look at the details in the picture, which AI doesn’t always
get quite right. Also question the feasibility of a picture; if it seems
improbable, it probably is.

What Kinds of Images Can You Generate with AI?

As you’ll learn later in this chapter, there are a number of AI image

generators you can use to create all sorts of images. To create an image, all
you have to do is enter a text prompt that describes what you want to see.
The more descriptive your prompt, the more satisfactory the image.

What kinds of images can you generate with AI—and how? Let’s look at a
few examples of different types of images and artwork you can create and
the prompts you might use to create them.

Cartoons and Comic Book Art

With AI, you can turn any real person into a cartoon. Some AI tools let you
upload a photo and have the tool make the person in the photo into a
cartoon or comic book character. When you’re talking about famous people,
living or dead, just tell the AI to draw the person as a cartoon character in a
particular situation. For example, Figure 7.4 shows how ChatGPT
responded to my prompt to draw Abraham Lincoln as a superhero flying
over Washington, D.C. Go, Super Abe!
FIGURE 7.4

Abraham Lincoln as a flying superhero, courtesy of ChatGPT and DALL-E.

AI tools that let you provide your own “guide” images are even more fun.
Upload one of your photos, choose a cartoony style or model, and see what
happens. Figure 7.5 shows what happened when I had OpenArt’s AAM XL
model turn me into a cartoon.
FIGURE 7.5

Your author as a cartoon, courtesy of OpenArt.

Collages

AI tools are great for making art collages. With most AI generators, all you
have to do is prompt them to make a collage of a particular subject. For
example, Figure 7.6 shows a collage generated by DreamStudio from the
prompt create a collage of vintage sports cars.
FIGURE 7.6

A collage of vintage sports cars, courtesy of DreamStudio.

Fantasy

An AI image generator is adept at creating new characters and worlds,

which makes it perfect for illustrating fantasy scenarios. Choose an
appropriately fantasy-oriented style or include the word “fantasy” in your
prompt, such as create an illustration of a fantasy world with dragons
flying over an ancient village. Figure 7.7 shows one possible outcome.
FIGURE 7.7

A dragon-filled fantasy world, courtesy of OpenArt.

Fine Art

Looking for some art to hang on your wall? You can use AI to create the
sorts of images you find in art galleries—and then enlarge it for your wall
or put that art on T-shirts, coffee mugs, or whatever you like. Just describe
the art you want, like this: create a photo of a sailboat on a calm ocean
with the sun setting on the horizon. Figure 7.8 shows how Midjourney
interpreted those instructions.
FIGURE 7.8

A fine art image from Midjourney.

Greeting Cards

You can also use AI image generators to produce more practical projects.
For example, if you want to create a unique holiday greeting card, just tell
the prompt something like create a holiday card with snow-covered trees
and the message “Happy Holidays.” Figure 7.9 shows what Google
Gemini created.
FIGURE 7.9

A holiday greeting card from Google Gemini.

Manufactured Reality

AI can generate an image of just about anything you imagine. It can do so

in various illustrative or cartoony styles or in photograph-like images that
look just like the real thing. Try imagining a different reality and have AI
take a “photo” of it, with a prompt like create a photorealistic image of
1930s New York City with a giant robot stomping down Broadway.
Figure 7.10 shows what that might look like.
FIGURE 7.10

A giant robot stomping down Broadway, generated by Microsoft Designer’s Image Creator.

Warning

Some AI image generators have built-in filters to keep you from generating
images that might infringe on copyright or other intellectual property rights,
such as images of celebrities or popular characters. (For example, some
tools won’t generate images of copyrighted and trademarked superheroes,
such as Superman or Batman.) Even if the AI image generator doesn’t
prevent the generation of such images, however, don’t assume the image is
fine to use. Always obtain written permission before using a copyrighted
image.

Portraits

You can use AI to generate portraits of any type. Just tell the AI tool as
much as possible about the type of portrait you want, such as create a
portrait of an African American woman sitting in a garden. You’ll get
something like the portrait in Figure 7.11.
FIGURE 7.11

An AI-generated portrait, courtesy of Adobe Firefly.

Not all portraits need be photographic. If you have a particular art style in
mind, you can tell the AI engine to create a charcoal sketch, create an oil
painting, create a black and white portrait, or even create a caricature.
Figure 7.12 shows an oil painting of an old fisherman.

FIGURE 7.12

An oil painted portrait, courtesy of DreamStudio.

Some AI generators let you upload your own photos that you can then
manipulate into a better portrait than what you started with. As an example,
I uploaded a photo of myself with messy hair and wearing an old sloppy
shirt to OpenArt and used the prompt, create a corporate headshot
wearing a blue suit and tie. The result is shown in Figure 7.13; it cleaned
me up pretty good, and even combed my hair!

FIGURE 7.13

A corporate-ready headshot, courtesy of OpenArt. (Would you hire this guy?)

 Warning

Any images you upload become part of that AI model and can be used to
further train the model and generate future images. Don’t upload anything
you don’t want shared or don’t have authorization or permission to upload!

How to Ethically Use AI Image Generators

Using AI image generators can be fun. You just have to be careful how you
use any images generated by AI. You can’t claim them as something you’ve
created yourself, you shouldn’t use them to fool people, and you have to be
careful about using any copyrighted images.

Here are some best practices to follow:

Don’t claim credit. The art generated by AI is not art that you
personally create. Some artists have gotten into trouble trying to pass off
AI-generated artwork as their own creations. It’s not.
Don’t claim copyright. You can’t copyright artwork that you didn’t
create yourself. Period.
Don’t pass off AI-generated art as real. AI can generate photorealistic
images. Don’t try to fool people into thinking they’re real images. As
before, be transparent and let people know the images were AI
generated.
Don’t use copyrighted images. When creating images with AI, be
careful not to infringe on any copyrights. That means don’t create
images of copyrighted characters and don’t use images that are clearly
based on copyrighted material.

Note

Better yet, use AI that was trained only on licensed

images. For example, Generative AI by Getty Images is
trained only on the Getty Images library; thus, the images
are licensed and there is indemnification from copyright
claims.

Don’t create images of real people. Along the same lines, you
shouldn’t use AI to create images of real people, either celebrities or
people you know. Many AI image generators won’t let you do this, at
least when it comes to well-known individuals. While creating an image
of Tom Cruise at your backyard barbeque might be fun, it’s an
inappropriate use of his image.
 Don’t create deepfakes. Similarly, using AI to create images of real
people in less-than-real situations is not only unethical; it may be illegal.
Don’t use AI as propaganda, to misinform, or to spread false
information. You should never use AI to try to fool people in any way,
shape, or form.
Don’t use AI for commercial work—without saying so. While it’s
okay to use AI for noncommercial purposes, pros shouldn’t rely on AI to
do their work for them. Professional artists should never use AI to create
artwork for commercial use, unless they clearly disclose so. Just as you
can use AI to generate story or message ideas and then write the final
version yourself, professional artists can use AI image generators to
generate a bunch of art ideas but then draw or paint the final version
themselves, based on that inspiration.
Be transparent. If you’re presenting AI-generated artwork, even if it’s
just on social media, let people know that it’s AI-generated. A simple
credit line or caption along the lines of “Generated by AI” or “AI-
generated art” lets people know what they’re actually seeing.

Bottom line, you need to be responsible when using AI image generators.

You can have lots of fun but don’t take that fun too far.

Warning
Many image generators block pornographic, violent, or types of offensive
images. Keep it clean out there, folks!

Comparing Popular AI Tools for Creating Images

There are a large number of AI image generators available today. While all
work in pretty much the same fashion, some have slightly different features
and, because they use different AI engines, can generate vastly different
results.

Table 7.1 provides details about the most popular AI image generators.
Information about each tool follows.

TABLE 7.1 Popular AI Image Generators

Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

Adobe https://firefly.adobe.com Yes $4

Firefly
 Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

DALL-E https://www.chatgpt.com No $2

DeepAI AI https://deepai.org/machine-learning- Yes $5

Image model/text2img
Generator
 Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

Deep Dream https://deepdreamgenerator.com Yes $9

Generator $9

DreamStudio https://dreamstudio.ai Yes $1

by Stability 10
AI cre

Google https://gemini.google.com Yes $1

Gemini

Hotpot AI https://hotpot.ai Yes $1

Art 10
 Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

Generator cre

Image https://designer.microsoft.com/image- Yes Fr

Creator from creator
Microsoft
Designer

Midjourney www.midjourney.com No $1
$1
 Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

NightCafe https://creator.nightcafe.studio Yes $5

$4

OpenArt https://openart.ai Yes $1

 Pa
Pl
Image Free
URL (co
Generator Plan?
pe
mo

That’s the basic information about a lot of sophisticated AI generators. As to

which one is best, that’s a matter of opinion, to some degree. The best
advice I can give is to give them a try and see which generates the images
you like best—and be prepared to use multiple image generators to get the
widest variety of results.

The following sections offer a more detailed look at each of these AI image
generators. Although they all do pretty much the same thing, they each do it
somewhat differently and some are better at some things than others.

Adobe Firefly

Adobe Firefly (https://firefly.adobe.com) is both a freestanding AI image

generator and an AI tool for use within Adobe’s Creative Cloud products,
such as Photoshop. The freestanding version of Firefly is free, with less-
restrictive paid versions also available; the embedded version is available to
all Creative Cloud subscribers.

As with most AI image generators, you describe what you want via a text
prompt and Firefly does the rest. Firefly can also punch up existing images
by adding style and textures and generating content to fill in portions of an
image. Since Firefly is trained on a massive dataset of stock photos from
Adobe Stock as well as public domain and freely available images from
Creative Commons, you get high-quality results that you can use both
personally and commercially.

Note

Creative Commons is a non-profit organization that enables

creators to upload their work and make it available to the
general public. Images protected by a Creative Commons
license can be used for free with certain restrictions.

To use Adobe Firefly, simply enter a descriptive prompt into the prompt box
and click the Generate button, as shown in Figure 7.14. Four results are
then generated, as shown in Figure 7.15. To download an image, hover over
it and then click Download. To download all four images, click Download
All.

FIGURE 7.14

Adobe Firefly’s home page.

FIGURE 7.15
Adobe Firefly’s image results, as well as further controls.

Controls on the left side of the page let you adjust the AI model used, the
image’s aspect ratio, content type (art or photo), and other options. You can
also upload your own images for editing via the AI tool.

DALL-E in ChatGPT

One of the best-known AI image generators today is OpenAI’s DALL-E.

DALL-E is available in the ChatGPT AI tool (https://chatgpt.com), along
with the regular text-based chatbot.

Note

Learn more about ChatGPT in Chapter 3, “Getting Started

with All-Purpose AI Tools.”

DALL-E works within ChatGPT. You specify that you want to generate an
image by adding generate an image of or create a photo of or create a
watercolor painting of or something similar to the front of your prompt.
For example, if you want to create a photo-like image of a pelican riding a
skateboard on the beach, you’d use the prompt: create a photo of a pelican
riding a skateboard on the beach. Just make sure you’re as descriptive as
possible about what you want, including the output type (photo, artwork,
and so on) and style.

As you can see in Figure 7.16, after you enter the prompt in ChatGPT,
DALL-E generates the image and displays it in the main pane. Hover over
the image to upvote or downvote the image; click the Download (down
arrow) icon to download the image.

FIGURE 7.16

A single image generated by DALL-E in ChatGPT.

At present, there is no free version of DALL-E available in ChatGPT. To
get access to DALL-E image generation, you need to subscribe to ChatGPT
Plus, which runs $20 per month.

Note

At the time of writing, ChatGPT was unusual in that it

enables downloading only in the WEBP file format, not in
the more popular JPG or PNG formats. This may change
sometime in the future.

DeepAI AI Image Generator

The AI Image Generator (https://deepai.org/machine-learning-

model/text2img) is one of several AI tools offered by DeepAI. (DeepAI also
offers a text-to-speech tool and an AI Video Generator.) There is a free
version available, with limited access. DeepAI Pro runs $4.99 per month
and offers up to 500 “calls” (image prompts) per month.

You generate an image from the main page, shown in Figure 7.17. Enter
your description into the Create an image from text prompt box, choose a
model (standard definition or high definition), and then select one of the
available art styles. The AI Image Generator displays five of the most
popular art styles by default; click View all styles to see more.

FIGURE 7.17

The user interface of DeepAI’s AI Image Generator.

When you click the Generate button, you see the results, as shown in
Figure 7.18. From here you have several options:

Download the image to your computer in JPEG format

Enhance the image by regenerating it
Remove Background to do just that and show only the foreground
image
 Animate to import the image into DeepAI’s AI Video Generator and
create a short video based on the image

FIGURE 7.18

An image generated by DeepAI’s AI Image Generator.

Deep Dream Generator

Deep Dream Generator (https://deepdreamgenerator.com) is a somewhat

basic AI image generator, at least in terms of extra features. While there is a
free trial available, it’s somewhat limited in the number of images you can
generate—about 10 images in total, depending on the complexity of the
image. Paid plans let you generate more images and run from $9 per month
for the Basic plan (36 images per day) all the way up to $99 per month for
the Ultra plan (360 images per day).
To use the Deep Dream Generator, shown in Figure 7.19, simply enter your
prompt into the Text Prompt box, choose the desired AI model (Artistic,
Photonic, etc.), and then scroll down to the bottom of the page and click the
Generate button. You can also upload image files (by clicking the Visual
Prompt button) to provide visual guidance to the image generator.

The Deep Dream Generator generates the desired image, as shown in

Figure 7.20. Click the Download button to download the image to your
computer in JPEG format.

FIGURE 7.19

The prompt entry page for Deep Dream Generator.

FIGURE 7.20

An image created by Deep Dream Generator.

DreamStudio by Stability AI

DreamStudio (https://dreamstudio.ai) works like most of the other tools

discussed here. It also lets you upload your own images to create variations
and different versions.

DreamStudio operates on a “credits” system. Each credit is good for

generating five images and, if you sign in with your Google account, you
get 25 credits for free. You can purchase an additional 1,000 credits for $10.

As you can see in Figure 7.21, the prompt panel is where you get the party
started. Select what style of image you want (for example, photographic,
anime, comic book), enter your description into the Prompt box, select how
many images you want to generate (the default is four), and then click the
Dream button. It’s pretty easy to use.

FIGURE 7.21

DreamStudio’s easy-to-use prompt panel.

Figure 7.22 shows the images that DreamStudio generates. Hover over an
image to generate additional variations, edit the image, download the
image, or delete the image.
FIGURE 7.22

Photorealistic images generated by DreamStudio.

Google Gemini

As noted in Chapter 3, Gemini (https://gemini.google.com) is Google’s AI

model. Unlike many other general-purpose tools, Gemini lets you generate
both text and image output.

Google Gemini’s basic version is free to use; Google Gemini Advanced

offers the next generation model (that theoretically can generate better
results) and the ability to use Gemini from within Google Docs and other
apps. Image generation is available in both versions.

To use Google Gemini for image generation, you need to type create an
image of in front of your normal prompt. You then elaborate on this basic
instruction with more details, such as create a photorealistic image of or
create a fantasy illustration of.

As you can see in Figure 7.23, Gemini generates four images in response to
your prompt. Mouse over any image and click the Download (down arrow)
icon to download that image.

FIGURE 7.23

Generating images with the Google Gemini AI tool.

Note
 Google also offers the ImageFX image generation tool,
located at aitestkitchen.withgoogle.com/tools/image-fx.
ImageFX, which uses the same AI model, is essentially
Gemini’s image generation but with a more user-friendly
graphical interface. It should generate similar images.

Hotpot AI Art Generator

The Hotpot AI Art Generator (https://hotpot.ai) is a web-based image

generation tool. It can also upload your photos to create corporate headshots
and illustrated avatars.

Hotpot offers a free plan, which puts watermarks on all your images, and a
paid subscription, which costs $10 for 1,000 credits. (Each generated image
costs 50 credits.)

To use the Hotpot AI Art Generator, just answer the questions on the AI
Generator panel, shown in Figure 7.24. Enter what to draw, what not to
draw, style, your own image (if you choose to upload one), how many
images to make, and aspect ratio. Click Create to generate the image(s).
FIGURE 7.24

Prompting the Hotpot AI Art Generator to create a new image.

You see the generated image, as shown in Figure 7.25. Click AI Resize to
resize the image; AI Edit to remove the background, add text, or make
other edits; Download Image (down arrow) to download the image in PNG
format; or Share Image to share the image on social media.
FIGURE 7.25

An image generated by the Hotpot AI Art Generator.

Note

Hotpot’s Corporate Headshots tool is useful when you’re job

hunting or just need a nice-looking headshot. Upload any
decent photo of yourself and Hotpot turns it into a
professional-looking headshot photo. It also puts you in a
variety of corporate-appropriate outfits. This feature is priced
separately from the main Hotpot AI Art Generator; it runs
$10 for 40 images.
Image Creator from Microsoft Designer

Image Creator from Microsoft Designer

(https://designer.microsoft.com/image-creator) is another AI image
generator based on OpenAI’s DALL-E model. (It was formerly known as
Bing Image Creator.)

Image Creator is a free tool; no subscription plans are offered. You can also
access it from Copilot in Windows, as discussed in Chapter 3.

Using Microsoft Designer’s Image Creator is as simple as entering your

instructions into the prompt box and clicking the Generate button. As you
can see in Figure 7.26, several results are displayed.

FIGURE 7.26

Generating images with Image Creator from Microsoft Designer.

Click any image to view it larger and access various options. On the next
page, as shown in Figure 7.27, you can Download the image in JPEG
format; Copy the image to paste into another application; click the Create
Design button to use the image in another design, using Microsoft
Designer; or Edit the image in Microsoft Designer by cropping it, removing
the background, and so forth.

FIGURE 7.27

Working with a generated image in Microsoft Designer’s Image Creator.

Midjourney

Midjourney is a powerful AI image generator that allows a comparatively

large amount of image customization. There is no free version available;
you need to sign up with a Google or Discord account and then choose a
subscription plan. Plans are priced from $10 per month for the Basic Plan to
the $120 per month Mega Plan that offers significantly faster processing
and can run more jobs concurrently.

You access Midjourney at www.midjourney.com. When you log in to your

account, you see the Explore page (shown in Figure 7.28) with samples of
generated images from other users. Click an image to see the prompt and
configuration options that generated that image.

FIGURE 7.28

Midjourney’s Explore page.

To generate your own images, click Create in the sidebar to display the
Create page, shown in Figure 7.29. You now see images you’ve previously
created as well as the Imagine Bar at the top of the page. This is where you
enter your prompts.

FIGURE 7.29

Midjourney’s Create page, complete with Imagine Bar.

What makes Midjourney unique is that it lets you configure its image
generation engine in a number of interesting ways. This lets you fine-tune
the images you create.

For example, to upload a reference image, click the Add Images to Your
Prompt icon on the left side of the Imagine Bar. You can upload images
that provide examples of characters or styles you want to reference.
For even more customization options, click the Settings icon on the right
side of the Imagine Bar. This opens the panel shown in Figure 7.30, with
the options detailed in Table 7.2.

FIGURE 7.30

Customizing Midjourney’s image generation settings.

TABLE 7.2 Midjourney Image Generation Settings

Setting Description

Image Size Sets the size and aspect ratio of the image. Select from
one of three preset ratios (Portrait 3:4, Square 1:1,
Landscape 4:3) or drag the slider left or right for more
extreme aspect ratios.

Stylization Low values produce images that closely match the

prompt but are less artistic; high values are more
artistic but less related to the prompt.

Weirdness The higher the value, the more quirky or offbeat the
image is likely to be.

Variety Lower values provide more reliable and repeatable

results; higher values produce more unusual and
unexpected results.

Mode The default mode is Standard; the Raw mode applies

less automatic “beautification” and may more closely
match the prompt.

Version Lets you choose which version of Midjourney to use.

 Setting Description

The higher the number, the more recent the version.

Personalize When enabled, uses information from your likes and

ranking of other images to personalize the images you
create.

Speed Turbo mode is faster than the Fast mode but uses more
GPU credits, which are limited by your subscription
plan.

Once you’ve entered your prompt into the Imagine Box and set your
parameters, press Enter on your keyboard. Midjourney generates four
images to match your request, as shown in Figure 7.31.

FIGURE 7.31

Four images generated by a Midjourney prompt; click any image to view it larger.
Click any image to view it larger. You have the option to take more actions
on this specific image. The following options are available in the right-hand
pane, as shown in Figure 7.32:

Download an image by clicking the Download icon at the top of the

pane.
Generate more images based on this image by clicking either the Subtle
or Strong buttons in the Vary section. Subtle generates less dramatic
variations; Strong generates more extreme variations.
Create a larger version of this image suitable for printing and other uses
by clicking the Subtle or Create buttons in the Upscale section. Subtle
doubles the size of the image exactly; Creative doubles the size of the
image and adds new details.
Run the same prompt again by clicking the Rerun button.
Crop, zoom, or change the aspect ratio of the image by clicking the
Editor button.
Use the generated image as a reference image for a new prompt by
clicking the Image button.
Use the generated image as a style reference for a new prompt by
clicking the Style button.
Copy the existing prompt into the Imagine Bar by clicking the Prompt
button.
FIGURE 7.32

Available actions for fine-tuning an image.

As you can see, Midjourney offers a plethora of options for fine-tuning the
images it generates. This makes Midjourney a good choice for those who
want to generate high-quality images with specific parameters.

NightCafe

NightCafe (https://creator.nightcafe.studio) is unique in that it uses and lets

you choose from a variety of AI image-generation models. You can even
fine-tune any of these models to create your own custom model attuned to
your personal tastes and needs.
 Note

Different companies have created their own models for

generating AI images. For example, OpenAI created the
DALL-E model and Google created the Gemini model. Each
model uses unique programming and algorithms and is
trained on its own unique set of data. This is why you get
different results when you type the same prompt into
different AI image generators—they each work in their own
ways.

Like most of these AI image generators, there’s a free version available,

although it only gives you five credits per day. (A low-resolution image
costs one credit.) You purchase the Pro version via credit packs priced from
$5.99 for 100 credits per month to $49.99 for 1,400 credits per month.

The Create panel, shown in Figure 7.33, is where you get started. Click the
Model selector to select which model you want to use—at this writing, you
have the choice of Dreamshaper XL Lightning, DALL-E 3, Stable Core,
and Stable Video Diffusion. (If you scroll down, you can choose from a
variety of community-created models, too.) Enter your image description
into the Text Prompt box, select a style, (cinematic, color painting, CGI
character, and so on), and then click the Create button.

FIGURE 7.33

Choosing an AI model and generating an image with NightCafe.

The resulting image is displayed in the main pane, as shown in Figure 7.34.
From here you can duplicate (copy) the image, evolve the image by creating
another one anew, enhance the image by increasing its resolution (which
costs credits), download the image in JPEG format, or animate the image.
FIGURE 7.34

Viewing a newly generated image in NightCafe.

Note

NightCafe’s use of multiple AI image models makes it a

great choice to try out different models without having to
visit multiple websites.

OpenArt

OpenArt (https://openart.ai) is like NightCafe in that it lets you choose from

multiple AI image generation models. When you first join, you get 50
credits to use; the Start plan offers 5,000 credits per month for $12 per
month, the Hobbyist plan gives you 15,000 credits per month for $24 per
month, and the Pro plan delivers unlimited credits for $56 per month.

To use OpenArt, go to the Create panel, shown in Figure 7.35. To switch to

a different AI model, click the Switch button and make a selection. (Current
options include the site’s own OpenArt SDXL, Juggernaut XL, DALL-E 3,
DreamShaperXL, Stable Diffusion XL, and literally dozens more.) Like
NightCafe, you can also access models created by the user community or
fine-tune your own models.

FIGURE 7.35
Selecting model and generation options with OpenArt.

Once you’ve chosen a model, you can scroll down to upload a pose,
composition, style, or face reference. Click the Create button to generate
your image.

The image you generate is shown in the main part of the window, as shown
in Figure 7.36. Hover over an image to display it larger, download it in
JPEG format, or delete it.

FIGURE 7.36

Viewing generated images with OpenArt.

Crafting the Perfect Prompt to Generate Perfect Images

Crafting the perfect prompt to generate perfect images is just like crafting a
text prompt for an all-purpose AI tool, but with visual concerns added. It’s
really about describing exactly what you want to see—the main image, the
background, the style, the output type, and more. Here are some tips:

Be as clear as possible. Instead of words that have a low information

content, such as “beautiful” or “wonderful,” use specific adjectives to
describe appearance, mood, style, and other details. For example, if you
want to see a large, gray striped cat, don’t just prompt for picture of a
cat; include all the details, such as picture of a large gray striped cat
with green eyes.
But don’t supply too many details. As important as the details are,
including too many can overwhelm some AI image models. If you end
up with a real hodge podge of an image, try simplifying the prompt.
If there’s a style you seek, point it out. If you want a picture in the
style of 20th century impressionists, include that in your prompt. If it’s a
photo you want or even a specific type of photo, like high contrast
black and white portrait or dreamy color like a wedding photo,
structure your prompt accordingly.
If you want something real in the image, say so. Instead of saying you
want a picture of a beach, say you want a picture of Pacific Beach in
California.
Include motion and emotion. Don’t just describe who or what is in the
picture, but what they’re doing. Instead of children on a playground,
prompt children chasing each other around a playground. Instead of
saying a couple looking at each other, say a couple looking lovingly
at each other.
 Use the tool’s options. Some AI image generation tools only let you
input a text prompt. Others include various controls and options that help
you fine-tune your image. If there are options available, use them.

Remember, different AI image generators will generate different images.

You may need to try several different tools to get the results you want. And
don’t be afraid to fine-tune your query or “remix” an image if the first one
you get is close but not quite right. AI image generation is still in its
infancy, so experimentation is key.

Summary

This chapter was a lot of fun to write because creating images with AI is a
fun way to visualize what you could previously only see in your
imagination. You learned how AI generates images, how AI image
generators work, and all about the most popular AI image generators today.
You also learned all the various types of images you can create with AI and
the prompts to use to do so.

AI image generators are fun to play with for your own personal or casual
use. But using them for commercial use is another story. Professional artists
should not try to pass off AI-generated art as art they personally created.
The artistic community doesn’t look kindly upon that kind of
misrepresentation; if you use AI to create a commercial image (or even one
you share on social media), you need to say so clearly.
You also learned that you shouldn’t use AI to create fake images designed
to fool people. Even though AI can do that, and often does it quite
convincingly, it isn’t ethical to do so.

There’s a reason why there are so many AI image generators out there today
and why they’re so popular: They’re fun to use and can actually be quite
useful. Just be careful how you use them and what you use them for.

OceanofPDF.com
8

Using AI to Get a Job

IN THIS CHAPTER

Using AI to help find the job you want

Using AI to write a resume
Using AI to write cover letters
Using AI to prepare for interviews

AI can help you find your next job, from choosing a career path and
company that suits you to drafting your resume and cover letter. AI can
even help you prepare for the interview!

As with all AI-related activities, you’ll want to use caution with every step,
keeping in mind that AI can make mistakes and fill in blanks with
inaccuracies. So use AI to make your job hunt easier, but take what AI
outputs as a jumping off point and never a final say.

Using AI to Help Find the Job You Want

Sometimes the hardest part of the job search is determining the job you
want. Whether you’re looking for your first job, a new opportunity in your
field, or a different career, AI can help.
Identifying the Types of Jobs That Suit You

The first thing AI can help you with, career-wise, is identifying those jobs
you may be particularly suited for or interested in. In this regard, AI can
help you fine-tune a career path based on your experience, skill set,
interests, strengths, and weaknesses.

Identifying an appropriate career is easily done with all-purpose AI tools,

such as ChatGPT, Google Gemini, and Microsoft Copilot. To start, craft a
prompt that includes your key skills, past experiences, and preferred work
style, then ask the AI what jobs or careers you might be suited for. The
prompt can look something like this:

Please recommend some careers for me. My skills include [skill 1], [skill
2], and [skill 3]. I have experience doing [experience 1], [experience 2], and
[experience 3]. My previous jobs have included [job 1], [job 2], and [job 3].
I like to work [independently/as part of a team/remotely/in an office].

Include as many or as few variables as apply to your particular situation.

Obviously, if you’re just starting out, you won’t have as many (or any) prior
job experiences; if you’re further along in your career, you’ll have more
highlights to add.

When I entered a similar prompt into ChatGPT with my own skills and job
experience, I got the recommendations you see in Figure 8.1.
FIGURE 8.1

Some personalized career recommendations from ChatGPT.

Researching Potential Employers

When you’re considering where you might like to work, the more you know
about a potential employer, the better decision you can make. Fortunately,
most all-purpose AI tools have plenty of information about larger
companies, having ingested employee reviews, social media posts, news
articles, and information on the companies’ websites.

For a general overview of a company, use the following prompt:

What does [company] do?

To get an idea of what it’s like to work at that company, use the prompt

What is it like to work at [company]?

You can also ask more targeted questions, such as

What is the work culture like at [company]?

What is the work-life balance like at [company]?
What is the average employee age at [company]?
What are [company’s] values?
How satisfied are the employees at [company]?
What is [company’s] financial health?

You can also use AI to identify key decision-makers at a company—

information that might otherwise be relatively hidden from the general
public. Just enter the prompt

Who are the key decision-makers at [company]?

Or, for a more targeted search, enter

Who are the key decision-makers in the [name of department]

department at [company]?

This is one aspect of the job search where AI can really lead you astray.
Make sure you double-check its output on potential employers.
 Note

You can also use similar strategies to learn more about

people you can network with in your job search. If you know
what company a person works for, use AI to learn more
about that company and even that person’s department
before you start networking.

Using AI to Write a Resume

Just knowing what job you want or where you want to work isn’t enough.
You need to put together a resume that really sells your skills and
experience to potential employers, is well organized, and is clear and well
written—with no typos!

Fortunately, AI can do most of the work for you.

Writing a Resume with an All-Purpose AI Tool

Writing a resume is something that generative AI tools do well. All you

have to do is tell AI a little bit about yourself and that you need a resume
written. You can then cut the response from your AI tool and paste it into
your favorite resume template in Google Docs or Microsoft Word and—
voilà!—you have a professional resume in hand.

When it comes to crafting the perfect resume-creation prompt, you need to

enter the key information that employers will want to see. This information
should include

Desired position
Current and past work experience—list positions and employers,
including dates employed, along with your duties and key
accomplishments in each job
Important skills—include position-specific skills (for example, if you’re
a welder, welding would be an important skill) as well as more general
skills, such as teamwork, leadership, creative thinking, strategy, and so
forth
Education, including college degrees and professional training
Other accomplishments, including professional certifications, awards,
honors, and major projects

Such a prompt might look something like this:

Write a resume for a [position]. My skills include [skill 1], [skill 2],
[skill 3]. My employment history includes [position 1, employer 1, dates
employed], [position 2, employer 2, dates employed], and [position 3,
employer 3, dates employed]. I graduated from [name of school] in
[year] with a [name of degree and/or major] degree. I have
accomplished/received [accomplishment 1], [accomplishment 2],
[accomplishment 3].

The resulting resume might look something like the one in Figure 8.2

FIGURE 8.2

The start of a resume written by Google Gemini.

Warning
Make sure you proofread and edit any resume that AI writes for you. AI
engines like to read between the lines, so to speak, and fill in any gaps in
your resume. In doing so, they might make things up that you haven’t really
done. Always remember that AI regularly makes mistakes, so review the
resume to be sure it is accurate.

Writing and Fine-Tuning a Resume with a Resume-Writing Tool

There are also several AI tools specifically designed for writing resumes.
Some popular tools include Jobscan, Kickresume, Rezi, Teal, and Zety. This
section covers each in turn.

Improving Your Resume and LinkedIn Profile

You can also use all-purpose AI tools to get tips on how to improve an
existing resume. Just cut and paste your resume into the tool along with the
prompt provide recommendations to improve my resume.

All-purpose AI tools can also help you optimize your LinkedIn profile,
which many employers look at when they’re considering job applicants.
Just copy and paste your resume into an all-purpose AI tool, accompanied
by the prompt Write a LinkedIn “About” section using my resume as a
guide.
Jobscan

Jobscan (www.jobscan.co) can use AI to generate resumes based on your

inputs, as well as match your resume to open positions. It can also analyze
your current resume, score it, and suggest improvements. The free version
offers two resume scans a month and limited functionality. For $49.95 per
month, you get unlimited resume scans, resume editing and management,
job matching, and more. (Figure 8.3 shows the result of a sample Jobscan
resume scan.)

FIGURE 8.3

The result of a resume scan by Jobscan.

Kickresume

Kickresume (www.kickresume.com) uses OpenAI’s GPT-4 AI model to

power its resume-creation service. You pick a template, enter key
information about yourself, and Kickresume does the writing for you. The
site also includes a resume checker for your existing resume, a cover letter
builder, and a personal website builder. The free plan includes four basic
resume and cover letter templates; the $19 per month plan provides more
than 40 resume and cover letter templates, the resume checker, and the AI
Resume & Cover Letter Writer. Figure 8.4 shows how Kickresume’s AI
Wizard works.
Figure 8.4

Building a new resume with Kickresume’s AI Wizard.

Rezi

Rezi (www.rezi.ai) offers AI resume building, editing, checking, and

summary generator tools. (The summary analyzes your entire resume to
generate a summary or introductory paragraph.) Rezi also offers an AI
cover letter writer, resignation letter writer, and interview practice.

You can do a free trial that includes limited access to some (but not all)
features. The Pro plan, which offers full access to all features and unlimited
AI usages, costs $29 per month. Figure 8.5 shows a “Rezi Score” for a
sample resume—not bad but could use some improvement.

FIGURE 8.5
Scoring a resume with Rezi.

Teal

Teal (www.tealhq.com/tools/resume-builder) uses AI to help you write and

design your resume. Teal’s Resume Builder, shown in Figure 8.6, also
analyzes and scores your resume in terms of effectiveness and generates an
introductory summary.

FIGURE 8.6

Designing a resume with Teal’s Resume Builder.

The basic version of Teal, which offers limited access, is free. The Teal+
plan offers unlimited access and more advanced features for $29 per month.
Zety

Zety (https://zety.com) is a resume and cover letter building tool that uses
AI technology. You can use Zety to check your current resume, build a new
one, and find the perfect resume template. Figure 8.7 shows a typical
resume created by Zety. Zety costs $1.95 for 14-day limited access or $5.95
per month for full access to all features.

FIGURE 8.7

A resume created by Zety.

Using AI to Apply for a Job

Some of the AI resume-building tools just discussed offer job matching,

which identifies open positions that match the skills, experience, and
accomplishments on your resume. In most instances, however, you apply on
either a company website or a web-based job platform, such as
CareerBuilder, Indeed, or ZipRecruiter. (LinkedIn, while technically a
social network, is also a significant source for jobs and lets you apply for
many positions directly from its site.)

In addition, some AI-assisted job platforms can automate the application

process by filling out your basic information across different companies and
tracking the status of your applications. You can find this type of “easy
apply” functionality at Indeed, LinkedIn, ZipRecruiter, and other sites.

Using AI to Write Cover Letters

Most of the dedicated resume-building tools just discussed also draft, or

help you draft, cover letters. Plus, you can use just about any all-purpose AI
tool to write a cover letter for you.

To have AI write your cover letter, prompt it with the appropriate

information. In particular, you need to tell it to create a cover letter and
include the following:

The name of the company to which you’re applying

The title of the job for which you’re applying
Why you’re interested in the job and the company
 Basic information about yourself that relates to the job requirements,
such as prior work experience, applicable skills, and notable
accomplishments
Anything else that expresses why you would be a good choice for the
position

The AI tool should be able to take this information, fill in any gaps, and
craft an effective cover letter. You can then cut and paste the AI text into
your preferred cover letter format in Google Docs or Microsoft Word, along
with your name, phone, email, and, if you have one, website or LinkedIn
URL.

Warning

You may hear about strategies to manipulate AI applicant tracker systems

into favoring your materials. Tactics include things like keyword stuffing,
where you front load your resume or cover letter with words that might
cause your materials to be more appealing, or white fonting, which involves
putting attractive words in a white font so that they aren’t visible but do get
picked up by an AI tool. Be warned: These strategies are strongly
discouraged by professionals in the hiring world and may even cause your
application to get rejected.
Using AI to Prepare for Interviews

As you’re aware, many general-purpose AI tools let you carry on a real-

time conversation with the AI chatbot. You can use these conversations to
practice for an interview. Just tell the chatbot you want to do a mock job
interview, provide some details about the employer and the job, and then
start chatting. The prompt may look something like this:

Can we do a mock job interview with [employer] for [position]?

I entered that prompt into ChatGPT for a content writer position and started
the interview. Here’s a transcript of one of the questions asked and my
response.

ChatGPT:

Can you describe a project or piece of content you worked on recently?

What was your approach, and what were the outcomes?

This question is designed to understand your workflow and how you

measure success in your work.

My response:

In my previous employment, I recently wrote a 70-page response to a

patient complaint to the state medical board. I provided detailed responses
to all 16 points in the complaint and backed it up with information from
various sources.

ChatGPT:

That’s a very detailed and high-stakes project showcasing your ability to

handle complex and sensitive content effectively. Here’s some feedback:

Strengths:

Detail-oriented: Your response to a multifaceted issue demonstrates strong

attention to detail and thoroughness.

Research Skills: The ability to back up claims with information from

various sources is crucial for a content writer and shows your commitment
to accuracy and reliability.

Areas to Enhance:

Outcomes: It would be helpful to mention the outcomes of your efforts. For

instance, was the complaint resolved favorably? What was the impact of
your document on the case?

Relevance: If possible, relate this experience back to the content writer role
you’re applying for. For example, explain how these skills translate into
writing compelling and well-researched content for different audiences or
purposes.
Note that ChatGPT prefaced the question with why the interviewer was
asking that question and then assessed the strengths and weaknesses of my
response. It also suggested areas where I could improve my answers. It’s all
quite useful and really helps you prepare for what you’ll face in a real
interview.

There are also several interview-focused websites that employ AI chatbots

in the same fashion. These sites include

aiApply (https://aiapply.co/ai-job-interview), free trial or $30 per month

for additional functionality
Final Round AI Interview Copilot (www.finalroundai.com), free trial
that includes a five-minute interview or $49 per month for 60 minutes of
interviews
Interview Prep AI (https://interviewprep-ai.com), free trial or about $11
for five interview simulations
Interviews by AI (www.interviewsby.ai), three interviews per month free
or $9 per month for an unlimited number of interviews

All these platforms work in a similar fashion. You provide the job title for
which you’re interviewing, and the AI asks you typical questions. You
answer and the AI evaluates your answers and offers suggestions for
improvement.
 Note

One more thing. After you have a job interview, it’s good
form to send thank-you letters to the people you talked with
at the company. You can use AI to write these thank-you
letters; just prompt write a short thank-you letter to
[person’s name] at [company] thanking them for meeting
with me on [date]. The AI tool will do the rest.

Summary

In this chapter, you learned the many ways you can use AI to help you find,
apply to, and interview for a job. General-purpose as well as specific AI
tools do a good job with these tasks.

Using AI tools will help you conduct a more targeted and efficient job
search, build effective resumes and cover letters, and prepare for job
interviews. Just be sure to use AI as a starting point and check its output for
accuracy.

OceanofPDF.com
9

Using AI at Work

IN THIS CHAPTER

Using AI to improve productivity

Using AI to make meetings more manageable

AI can speed up many tasks that used to consume a lot of time and make
others less frustrating, such as scheduling meetings, taking notes, or
creating memos and reports. AI can also help with some creative work, like
putting together presentations, newsletters, and marketing materials. All you
have to do is tell the AI tool what you want to do and how you want it done,
and it will do the rest.

Using AI to Improve Productivity

AI can help you improve your overall productivity at work—all types of

businesses are already doing that. There are a number of ways that you can
use AI to automate routine tasks and make your work life easier.

Generating Content

As you’ve been learning throughout this book, generative AI is pretty good

at generating many types of content. While there are some business-specific
content-creation tools, the all-purpose tools I’ve covered work just fine for
generating all sorts of work-related content, especially creative content.

How might you use AI to generate content in your job? Table 9.1 shows
some of the ways, along with the prompts you might use to do each task.

TABLE 9.1 Types of Work Content AI Can Generate

Content Sample Prompt

Project ideas Generate 3 ideas on how to improve our customer

retention rate

Memos Write a short memo informing staff that the office

will be closed early this Friday for the long holiday
weekend

Reports Use the following information to create a 3-page

report explaining the difference between projected
and actual sales for the month of May [include
information]

Charts and Graph the following monthly sales data [include

graphics data]
 Content Sample Prompt

Speaker notes Create notes for the following presentation [include

presentation text]

Presentations Create a 15-minute presentation promoting our

services to the auto repair industry [include info
about products]

Blog posts Write a 500-word blog post about current economic

conditions and how they will impact demand for
our products [include relevant data]

Product Create a short promotional description of our

descriptions product for our website [include product details]

Press releases Write a press release in standard format announcing

the opening of our new location [include location
information]

Any all-purpose AI tool such as ChatGPT, Google Gemini, or Meta AI can

generate this type of content. You only have to provide enough information
for AI to properly do its job.
 Note

As you learned in Chapter 4, “Using AI to Find the Right

Words,” you can also use AI to improve the grammar, style,
and clarity of your work-related writing. Even if you’re a
natural writer, AI can help you improve the writing you have
to do for work.

Managing Projects

Managing team projects can be a challenge, especially on large projects and

when you work with a team of many remote workers. Fortunately, you can
use AI to help manage even your most complex projects. AI can take over
many of the tedious and complicated tasks involved in project management,
including

Predicting project risks

Optimizing resource allocation and automating project workflows
Scheduling meetings
Sending reminders and alerts
Providing real-time project monitoring and updates
Generating reports
AI is beneficial for project management in that it learns from user behavior
and adjusts project assignments and deadlines accordingly. AI can analyze
large amounts of project data to predict future outcomes, such as project
completion dates, budget overruns, and resource shortages.

AI is also able to break a large project into smaller tasks and take a large
number of tasks and other inputs and quickly make sense out of them. With
the help of AI-based project management tools, managers can make better-
informed decisions, make on-the-fly course corrections, and keep even the
most complex projects on track.

That said, all-purpose AI tools are not the best approach for project
management. Instead, you want to use one of several AI-based project
management tools designed for business. The most popular of these tools
include

Asana (https://asana.com)
ClickUp (https://clickup.com)
Monday.com (www.monday.com)
Smartsheet (www.smartsheet.com)
Trello (https://trello.com)
Wrike (www.wrike.com)

Most of these tools work in a similar fashion:

1. Start by creating a new project and giving it a name.

2. Configure the relevant options for that project: owner, description, due
date, budget, notifications, and so on.
3. Add specific tasks to the project and assign those tasks to individuals on
your team.
4. Add due dates for each task.
5. As the project proceeds, each team member notes their progress on the
timeline, as shown in Figure 9.1.

FIGURE 9.1

AI-based project management with Monday.com.

The task management tool notifies team members when tasks are coming
due and notifies the team leader when tasks are completed. You can also
automate certain tasks, such as moving a task to a different group when
completed or notifying team members when a new task is added. You can
also program the tool to generate progress and other types of reports.

Warning

Before using any AI tool for work, examine your organization’s data
privacy policy or consult with a superior. Your company may not want its
proprietary information entered into a public AI tool.

Communicating and Collaborating

Another important use of AI at work is enhancing communication and

collaboration between team members. Some of these uses of AI can be
accomplished with all-purpose AI tools; others require activity-focused or
purpose-built tools.

Table 9.2 details some of the ways AI can enhance communications and
collaborations, along with appropriate tools for each task.

TABLE 9.2 Ways AI Can Enhance Your Day-to-Day Communication and Collaboration
Task Appropriate AI Tool

Intelligently filter and AI built into email or messaging program

prioritize email and
other messages

Automatically answer Dedicated AI responder tools, including

routine queries Ellie (https://tryellie.com), EmailTree
(https://emailtree.ai), Flowrite
(www.flowrite.com), and superReply
(https://superreply.co)

Analyze the All-purpose AI tools; copy message(s) into

emotional tone of the tool and prompt analyze the emotional
people with whom tone of these messages
you’re corresponding
and suggest
appropriate responses

Summarize large Dedicated email/chat summary tools,

volumes of emails including Hiver (https://hiverhq.com),
and messages to Shortwave (www.shortwave.com), and
highlight and Zapier (https://zapier.com); also can cut and
paste multiple messages into an all-purpose
 Task Appropriate AI Tool

prioritize key themes AI tool and prompt it to summarize

and issues contents

Organize, categorize, Dedicated AI summarization tools, including

and summarize large Dokkio, DOMA (www.domaonline.com),
values of documents and Nanonets AI (https://nanonets.com)

Whatever tools are used, AI can help streamline collaboration processes,

improve communications, and facilitate knowledge sharing—all of which
can lead to more efficient and effective teamwork with your co-workers.

Warning

Remember to verify any AI-generated or sourced content before you

include it in your documents, presentations, and plans. AI is sometimes
prone to “hallucinations” and just making stuff up, so make sure to fact
check anything you use that’s AI-generated.
Translating Documents and Conversations

If your work is global—or if you interface with teammates or suppliers in

other countries—you know that the language barrier can be challenging.
Fortunately, AI can help break that language barrier by interpreting
conversations and translating documents in real time.

You can find value from these and other AI-based translation tools:

DeepL (www.deepl.com)
Google Translate (translate.google.com)
Lingvanex (https://lingvanex.com)
Microsoft Translator (https://translator.microsoft.com)
QuillBot (https://quillbot.com/translate)
Systran (www.systransoft.com)
Wordvice AI (www.wordvice.ai/tools/translate)
Yandex Translate (https://translate.yandex.com)

These and other AI-based translation tools use neural machine translation
and other advanced techniques to deliver accurate and natural translations
in real time. If you operate in a multilingual environment or interface with
people who don’t speak your language, these tools can make your job that
much easier.
Creating Presentations

If presentations are part of your job, AI can help. AI can take much of the
stress out of creating presentations—even if it can’t actually give
presentations for you.

How can AI help prepare your next presentation? Consider that AI can

Generate general presentation ideas based on the topic you select

Create an outline for your presentation, incorporating any key points you
enter
Write the content for individual slides
Source relevant data
Create graphics and images for your slides, or identify appropriate
images from other sources
Suggest and apply design templates that best match the presentation’s
content and your audience
Adjust slide layouts to maximize the visual impact of information on
individual slides
Write a script or talking points based on your slides for you to use when
delivering the presentation

An all-purpose AI tool can do much of this for you. For example, you could
enter a prompt to generate a 15-minute presentation on how to use
Microsoft Teams for group collaboration. Figure 9.2 shows how
ChatGPT interprets this command. Just cut and paste the text from
ChatGPT into Microsoft PowerPoint or Google Slides and then add the
recommended graphics.

FIGURE 9.2

The recommended contents of a presentation created by ChatGPT.

source: www.chatgpt.com

There are also AI tools designed specifically for creating presentations,

including

Beautiful.AI (www.beautiful.ai)
Decktopus (www.decktopus.com)
 Sendsteps (www.sendsteps.ai)
SlidesAI (www.slidesai.io)
Tome (https://tome.app)

Each of these tools works a little differently but they all follow the same
general approach. For example, Sendsteps asks you what you want your
presentation to be about. You can answer with as little or as much detail as
you like. The tool presents several possible titles; choose one and specify
the desired length of the presentation in terms of minutes and number of
slides. Sendsteps generates an entire presentation for you, which you can
edit to your liking, including choosing slide themes and layouts. Figure 9.3
shows one such presentation.

FIGURE 9.3
An AI-generated presentation from Sendsteps based on a simple prompt: benefits of AI for office
workers.

Note

Microsoft PowerPoint includes the AI-powered Designer

feature that suggests design ideas based on slide content. It
also includes Presenter Coach, which provides feedback on
your presentation’s pacing and language.

In addition, some AI presentation tools can help you practice delivering

your presentation. You can get feedback on pacing and delivery that will
help you deliver an engaging presentation to your audience.

Using AI to Make Meetings More Manageable

AI can help you schedule meetings, transcribe and take meeting notes, and
even summarize the contents of meetings. It’s like having a virtual assistant
doing all the meeting dirty work for you.

Scheduling Meetings

Let’s start with one of the more challenging aspects of meeting

management: getting the darned things scheduled. Scheduling a meeting
involves juggling the individual schedules of the invitees—and the more
potential attendees, the more difficult that task becomes. Trying to find a
free hour on the schedules of a half-dozen or more individuals is a thankless
proposition.

That’s where AI comes in. Provide AI appropriate access to all employee

calendars, and it can immediately find that one hour everyone has free on
their schedules. No more shuffling back and forth between different
coworkers’ schedules to see if that person is free at 1:00 pm on Monday; AI
does all that grunt work for you.

That’s not all AI can do, either. When it comes to scheduling meetings, the
appropriate AI-powered meeting management tool can

Access multiple calendars to find open time slots across all participants,
even those in different time zones
Manage work, personal, and social media calendars—including Teams
and Slack
Automatically send meeting invitations and—closer to the meeting—
reminders to all attendees
Automatically suggest new times for meetings if conflicts arise
Book physical meeting rooms and equipment, if needed
Help you prepare for the meeting by creating agendas, gathering
necessary documents, and even suggesting topics based on current
projects and previous interactions with chosen participants
In addition, AI-powered meeting tools can learn individual preferences for
meeting times, lengths, and attendees. This lets the AI tool fine-tune
meetings to suit employees’ work habits.

All of these tasks are beyond what you might expect from an all-purpose AI
tool. Instead, you want to check out an AI-powered meeting scheduling
tool, such as

Calendly (https://calendly.com)
Clara (www.claralabs.com)
Clockwise (www.getclockwise.com)
Doodle (https://doodle.com)
Motion (www.usemotion.com)
Reclaim (https://reclaim.ai)

Some of these tools are cloud-based, whereas others reside on your

company’s servers or on your home computer. All work in pretty much the
same way; after you link the tool to your personal or work calendar, you tell
the tool you want to schedule a meeting of a given length with specified
individuals, and the tool does the rest.

Scheduling a meeting with one of these AI-powered meeting tools is

relatively simple. You start by connecting the tool to your company’s email
or messaging program, such as Microsoft Outlook or Microsoft Teams; this
gives the program access to employees’ email addresses and personal
schedules. You then tell the program what meeting you want to schedule,
who should be invited, and when would be some good times for the
meeting. (You may also be able to specify that you need the meeting
scheduled by such and such date or time.) The tool works through the
schedules of all invitees, determines the best time for all involved, and
sends out invitations. Figure 9.4 shows how Reclaim does it.

FIGURE 9.4

Scheduling a Smart Meeting with Reclaim.

Taking Notes and Summarizing Meetings

AI isn’t just for scheduling meetings; it can also capture and summarize the
content of meetings. This makes meetings more durable and functional
without any extra effort on your part.
Here are just some of the ways you can use AI-powered meeting apps to get
the most out of your meetings:

Transcribe the minutes of a meeting using voice-to-text transcription

technology
Organize and categorize meeting notes based on topics, keywords, or
tags
Analyze and summarize the content of a meeting and then distribute that
summary
Highlight the key points of a meeting and create to-do lists for meeting
attendees
Suggest related articles or notes based on meeting contents
Store meeting notes for future search and retrieval

Note

You need to notify meeting attendees, especially those

outside your company, if you’re using AI to record or
transcribe a meeting.

Some of the more popular AI-powered note-taking platforms include

Claap (www.claap.io)
 Fathom (https://fathom.video)
Fireflies (https://fireflies.ai)
Otter AI (https://otter.ai)
tl;dv (https://tldv.io)

Several of these tools, including Claap and Fathom, make video recordings
of your online meetings and then work from those recordings to transcribe
and summarize the meetings. Others work from meeting audio.

If you’re using one of these tools in a physical meeting, you need to open
the app on your computer or smartphone and manually start recording in the
app when the meeting starts, using your device’s camera and microphone. If
you’re recording an online meeting, you need to configure the app to work
with your meeting app, such as Google Meet, Microsoft Teams, or Zoom;
the app starts recording when the meeting starts. Some tools also let you
upload audio or video files from previously recorded meetings.

When the meeting starts, the AI tool does its job automatically. Some
transcribe the meeting in real time; others create the transcription when the
meeting is over. You typically get the option of editing a transcript before
you save it or send it to meeting participants or others in your company.
Figure 9.5 shows a typical meeting transcript generated by Otter AI.
FIGURE 9.5

An Otter AI meeting transcript.

By leveraging these AI-powered note-taking tools, you can enhance your

productivity, organization, and knowledge management capabilities. This
will make it easier for you to capture, retain, and later retrieve information
—and get the most of your many meetings.

Summary

This chapter covered just some of the many ways you can use AI at work.
In particular, you learned how AI can automate routine tasks, manage
projects, generate content, enhance communication and collaboration, and
create presentations. You also learned how AI can schedule, manage, and
summarize meetings.
That’s a lot of ways AI can help you at work, but it’s just the tip of the
iceberg. Many companies are using purpose-built AI tools to perform
specific tasks in their businesses. Don’t be surprised to sit down at your
desk one morning and be presented with a new AI tool you can use to help
optimize your work!

OceanofPDF.com
10

Using AI for Travel and Transportation

IN THIS CHAPTER

Using AI to create travel plans

Using popular AI-powered travel planning tools
Using AI to prepare for a trip
Using AI during a trip
Using AI in local transportation

Thanks to the dynamic duo of predictive and generative AI, travel is getting
easier. From determining your itinerary and booking your tickets to
navigating your route and acting as a real-time translator, AI is set to play a
major role in your travel and transportation needs.

That said, AI is not perfect, especially when it comes to travel planning. AI

can and sometimes does make mistakes that can include travel
recommendations that don’t exactly fit what you’re looking for. In addition,
AI doesn’t always possess the most up-to-date data, which means it might
not have the latest information about flights, lodging, and routes. It’s always
best to supplement AI recommendations with your own research and
information from established travel and navigation sites.
Using AI to Create Travel Plans

AI is already playing a major role in trip planning. You can use AI to

determine where to vacation, get personalized recommendations for your
next trip, and book the right flights and rooms—at the right price. AI can
streamline and personalize the entire travel planning process.

Much travel planning can be done using the all-purpose AI tools I discuss in
Chapter 3, “Getting Started with All-Purpose AI Tools,” including
ChatGPT, Claude, Google Gemini, and Microsoft Copilot. There are also
several AI-powered travel-specific tools available that provide even more
travel-related functionality; I cover those in the next section.

Whatever tool you use, AI can assist at every stage of the travel planning
process.

Making Personalized Travel Recommendations

One of the things that AI is good at is making personalized travel

recommendations. If you feed AI the right information—your past travel
history, things you like to do, your budget, and the trip timing and duration
—it can suggest new destinations for you to enjoy. AI matches your
preferences with the vast amount of data on which it was trained, resulting
in personalized recommendations that you might not think of otherwise.
You can also use AI to research where you’d like to go. Let’s say that, based
on your past experiences, AI recommends a trip to Florida. You can then
ask AI about the area in which you’re interested, things to do there, best
times to go, and the like.

Providing recommendations in this fashion is something that both all-

purpose AI tools and AI-powered travel-specific tools are good at.

Booking Flights and Accommodations

Most AI-powered travel tools (but not all-purpose AI tools) can handle all
aspects of your travels, including booking flights, finding accommodations,
and even making restaurant reservations. These tools use AI to analyze
historical data and predict pricing trends to determine the best time to book
your reservations. This could potentially save you significant money.

Creating an Itinerary

Once you’ve chosen a location for your travels, you can have AI create a
detailed day-by-day itinerary for your trip. Both all-purpose AI tools and
travel-specific tools can do this, although you might get more detailed and
accurate results from a travel-specific tool.

To have AI create an itinerary, just feed it information about where you’re

going, for how long, and what you like to do. You can, if you want, include
information about yourself and your travel companions (if any), physical
restrictions you might have, and specific activities or locations that must be
included. For even more detailed recommendations, feed the AI details
about your flights and hotel reservations.

For example, you might use prompts like the following:

Create an itinerary for a family with three elementary-age kids

visiting the Orlando area for a week in April.
Suggest a list of activities for my husband and I to do in New York
City over a weekend. We both like Broadway shows and museums.
We are traveling to California wine country for three days.
Recommend an itinerary that includes as many wineries as possible.
Suggest a good route for a weekend getaway in New England in
early October.
Create an itinerary for our upcoming trip to Europe considering the
following flights and hotel reservations [include information].

AI will consider the information you provided and suggest the best things to
do each day you’re there.

Warning

All-purpose AI tools may not always have up-to-date information, so you

should verify all recommendations they make. In addition, you may find
recommendations that include seasonal establishments or those that have
recently closed. Travel-specific sites and tools are apt to have more accurate
and up-to-date information.

Functioning as a Virtual Travel Assistant

It’s important to know that interacting with AI is more like dealing with a
real travel agent than just filling out a few fields on a web form. AI chatbots
can interact with you in a conversational manner, understand your queries,
and provide real-time responses. They may even be better than human
travel agents in handling complex travel arrangements.

Using Popular AI-Powered Travel Planning Tools

Although you can use an all-purpose AI tool such as ChatGPT, Claude, or

Meta AI to plan your next trip, using a focused AI-powered travel planning
tool is probably the better bet. As you can see in Table 10.1, there are
several of these tools available, all focused on helping both leisure and
business travelers get the most out of their next trips.

TABLE 10.1 AI-Powered Travel Planning Tools

Tool URL Price

GuideGeek https://guidegeek.com Free

 Tool URL Price

Layla https://layla.ai Free

Roam Around https://roamaround.app 30 tokens/$10

Trip Planner AI https://tripplanner.ai Free

Wonderplan https://wonderplan.ai Free

GuideGeek

GuideGeek (https://guidegeek.com) is an AI-based travel service for your

Android or Apple iOS smartphone. It uses the OpenAI engine and travel
information obtained over the Internet—and curated by human experts—to
answer your travel-related questions.

You use GuideGeek from within either Instagram, Facebook Messenger, or

WhatsApp apps; it’s totally free. Just ask GuideGeek a travel-related
question and you’ll get your answer.

Figure 10.1 shows GuideGeek in action within Instagram; I asked it to

make some recommendations for outdoor activities in Madison, Wisconsin,
over a long weekend. The interaction looks just like a normal Instagram
messaging session.

FIGURE 10.1

Obtaining travel recommendations from GuideGeek via Instagram.

Note
 Because GuideGeek is smartphone based, it’s a good tool to
use on the go while you’re traveling.

Layla

Layla (https://layla.ai) is a multifunction AI-based travel planning tool. You

can ask Layla for travel inspiration, to find specific types of attractions or
destinations, to plan road trips and other itineraries, and to book flights and
hotel rooms.

You can use Layla from its website or via its Android or Apple iOS apps.
It’s free.

As you can see in Figure 10.2, all you have to do is enter your question or
prompt into the Ask me anything box and click the Ask button. Figure
10.3 shows a typical response to a typical question.
FIGURE 10.2

Asking Layla a travel-related question.

FIGURE 10.3

Layla’s travel recommendations.

Roam Around

Roam Around (https://roamaround.app) lets you enter information about

where you’re going, and it generates what it calls a “hypercustomized”
travel plan. To use Roam Around, enter your destination, the number of
days you’ll be there, and any information about the type of trip and what
you might like to see, as shown in Figure 10.4. Roam Around generates a
fairly detailed itinerary for your trip, as shown in Figure 10.5.
FIGURE 10.4

Planning a trip with Roam Around.

FIGURE 10.5

A detailed itinerary from Roam Around.

Unlike some of the other AI-powered travel planners, Roam Around isn’t
free. Each travel plan you create costs one travel token; you can purchase
30 tokens for $10. Roam Around is available on the web or as an Android
or Apple iOS app for your smartphones.

Trip Planner AI

Next up is Trip Planner AI (https://tripplanner.ai), a free AI-powered trip

planner. Just tell Trip Planner AI where you want to go and when, as well as
the kind of activities you’re looking for (as shown in Figure 10.6), and Trip
Planner AI does the rest.

FIGURE 10.6

Planning a trip with Trip Planner AI.

Figure 10.7 shows the results from Trip Planner AI. You get a day-by-day
itinerary with each activity detailed on a map of your destination. You can
also use Trip Planner’s Hotel Finder to find hotels and Airbnb locations
convenient to your trip activities.

FIGURE 10.7

How Trip Planner maps out your travel itinerary.

Wonderplan

The final AI-powered travel planning site I’m covering is Wonderplan

(https://wonderplan.ai). This is a free travel planning site, shown in Figure
10.8. Just tell it your destination, when you’re traveling, your budget, and
your favorite activities, and Wonderplan generates an itinerary for you.
FIGURE 10.8

Entering key information about your trip into Wonderplan.

As you can see in Figure 10.9, Wonderplan’s results look a little like those
of Trip Planner, in that all your activities are plotted on a map of your
destination. Click any item to read more.
FIGURE 10.9

Viewing Wonderplan’s suggested itinerary for a weekend trip to Chicago.

Note

Many major travel booking sites have also integrated AI into

their sites for trip recommendations, booking, and more.
Among the sites using AI in whole or in part are
Booking.com, Expedia, Kayak, Google Travel, and Hopper.
Using AI to Prepare for a Trip

AI is useful for more than just suggesting where to go and booking tickets
to get there. You can use AI at almost every stage of your travel planning—
and even to suggest some tasks you might not think of on your own.

Researching Documentation Needs

If you’re just taking a weekend trip down the road, you don’t need anything
special in terms of documentation. But if you’re traveling out of the
country, you might need a passport, visa, or other documents—and
requirements can vary from country to country.

Ask your AI a question like this:

I’m travelling to Europe in March. I’m going to England, France, Italy,

Switzerland, and Poland. What documents do I need to take with me and
how do I get them?

Figure 10.10 shows what ChatGPT recommends.

FIGURE 10.10

Using ChatGPT to determine what documents you need to procure before traveling to Europe.

You can also use AI to discover medical recommendations for traveling to

some destinations outside the country, such as vaccinations for malaria and
yellow fever. Just enter a prompt like this:

What medical recommendations do you have for a trip to Nigeria?

Figure 10.11 shows what Claude AI recommends.

FIGURE 10.11

Claude AI’s medical recommendations for a trip to Nigeria.

Learning Local Languages

If you’re traveling to a non-English speaking country, it helps to know a

little of the local language before you arrive. Fortunately, several AI-
powered language learning tools can help you become fluent in a variety of
foreign languages—or at least learn enough to get by. These tools include

Babbel (www.babbel.com)
Busuu (www.busuu.com)
 Duolingo (www.duolingo.com)
Memrise (www.memrise.com)
Mondly (www.mondly.com)
Rosetta Stone (www.rosettastone.com)

These tools have incorporated AI technology to create personalized learning

experiences and provide immediate feedback on your progress. It can make
learning a new language easier than it used to be.

Getting Packing Assistance

Don’t know what to pack for your upcoming trip? A handful of AI-powered
apps can help. They take into account where you’re going, the length of
your trip, and the activities you have planned, as well as historical weather
trends and the upcoming forecast to recommend the type of clothing and
accessories to pack.

These AI-powered personalized packing apps include

PackPoint (www.packpnt.com)
Packr (www.packr.app)
WhatToPack (www.whattopack.ai)

These apps can help you avoid over- or under-packing, ensure you have
everything you need for your planned activities, and prepare you for
varying weather conditions. They’ll also help you stay more organized
when packing, so you won’t forget anything important.

Predicting the Weather

Part and parcel of smart packing is taking weather conditions into account.
While it’s still impossible to precisely predict the weather more than a day
or two out, AI-powered weather apps analyze past weather patterns and
current forecasts to provide more accurate forecasts for your desired
location.

Among the most popular of these AI-powered weather apps are

Atmo (https://atmo.ai)
Rainbow Weather (www.rainbow.ai)
Tomorrow.io (https://weather.tomorrow.io)

Enhancing Your Personal Security

Traveling can bring with it risks related to health, safety, and (depending on
the destination) political instability. Use these AI-powered apps to get
updates on real-time threats at your travel destination, including violence
and disease outbreaks:

GeoSure (www.geosureglobal.com/individuals)
Sitata (www.sitata.com)
 TravelSmart (www.travelsmartapp.com)

Using AI During a Trip

AI’s utility for travelers doesn’t stop when you get to your destination.
There are plenty of ways you can use AI during your trip—on either your
smartphone, tablet, or computer.

Getting Personalized Recommendations

While you can (and should) use AI to plan your itinerary and activities in
advance, you probably still want to do some exploring on the fly, even if it’s
just finding a good restaurant for dinner. Use general AI tools or travel-
specific tools to recommend nearby restaurants, clubs, attractions, and
experiences. Remember to include details of your tastes and expectations in
your prompts.

Providing Cultural Insights

Traveling to different places can expose you to different customs and

cultures. Use AI to learn about the customs and cultural sensitivities of
where you’re traveling, such as what time dinner is typically eaten, whether
tipping is expected, how formal the people are, how punctual they are, how
to greet others (Handshake? Cheek kisses? Bowing from the waist?), and
other things that the locals know and you might not. You want to conduct
yourself with the proper etiquette for your location.
Finding this information is as easy as asking any all-purpose AI tool the
question what are the cultural sensitivities of [country]. You’ll get a very
useful response such as the one shown in Figure 10.12.

FIGURE 10.12

Learning about the culture and customs of Japan, courtesy of ChatGPT.

Navigating Your Route

When you’re traveling, especially out of the country, you don’t often know
the best way to get from point A to point B. Should you walk? Take a taxi or
ride-share service? Use public transportation? Rent a car or drive your own
vehicle?

You can use AI to help you determine the best form of transportation and
the best routes to take.

As always, you can ask general navigation questions of just about any all-
purpose AI tool, but using a navigation-specific tool will probably be easier
and generate more targeted results. These AI-powered navigation tools will
help you decide what type of transportation to take, optimize your route,
provide turn-by-turn instructions, and offer real-time traffic updates. They
also provide up-to-date information about public transit systems (buses,
trains, and subways), including routes and schedules—all useful
information when you’re visiting an unfamiliar city.

The most popular of these AI-powered navigation tools include

Apple Maps (www.apple.com/maps/)

Google Maps (https://maps.google.com)
HERE WeGo (https://wego.here.com)
MapQuest (www.mapquest.com)
Waze (www.waze.com/live-map/)

That’s right, the most popular AI navigation tools just happen to be today’s
most popular map and navigation apps. That’s because they have evolved to
include AI assistance, particularly in providing predictive routing based on
historical data and current conditions, real-time road conditions and traffic
updates, safety alerts, and more. Just fire up your favorite map app on your
smartphone and let AI help you get to where you want to go!

Translating on the Fly

When you’re travelling out of the country, you won’t always have a firm
grasp on the native language. It helps to have some sort of on-the-fly
translation assistance, which AI can provide.

Several AI-powered translation apps can help you communicate more

effectively in places where you don’t speak the language. These apps can
translate spoken and written text in real time and even work offline if you
download the necessary languages ahead of time. These smartphone-based
apps include

DeepL Translate (www.deepl.com)

Google Translate (https://translate.google.com)
iTranslate (https://itranslate.com)
Microsoft Translator (https://translator.microsoft.com)
Papago (https://papago.naver.com)

There are also several portable AI-powered language translation devices.

These devices (from Anfier, Javisen, Vasco, and other companies) look like
small smartphones and use AI to translate words and photos from one
language to another. In this instance, AI is embedded in a handheld device
with a microphone, speaker, and camera. Figure 10.13 shows the handheld
Vasco Translator device.

FIGURE 10.13

The portable Vasco Translator device.

Keeping You Safe While Traveling

AI-powered personal safety apps can help keep you safe when you’re
traveling. Some of these apps send out alerts when you’re in trouble and
some also provide location-specific safety tips and ratings and alert you to
potential dangers in your area.
The most popular of these personal safety apps include

bSafe (www.getbsafe.com)
GeoSure (www.geosureglobal.com/individuals)
Smart Traveler (https://2009-2017.state.gov/r/pa/ei/rls/dos/165020.htm)

Create a Personal Safety Plan

Before any trip, it’s a good practice to put together a personal safety plan
for yourself that you can share with family and friends back home. This
plan should include all details of your itinerary—where you’re staying,
flights you’re taking, etc.—and local contact numbers. This way the people
will know how to contact you if they need to.

For your own use, this plan should include contact information for local
authorities, hospitals, and the like for each of your destinations. You can use
AI to assemble this information and store it on your smartphone and, if
you’re bringing it with you, your computer. It wouldn’t hurt to keep a print
copy of this plan in your luggage, too, just as a backup.

The whole point of this is to be prepared in the event that something

unexpected happens. If you’ve done your homework in advance, any
unfortunate incidents can be dealt with more efficiently and quickly.
Using AI in Local Transportation

AI can significantly enhance many aspects of your daily transportation,

making your various commutes more efficient, safer, and even more
enjoyable. As always, AI might not have the most up-to-date information
about routes and schedules, so verify details before you set out.

Getting Where You Want to Go

Just as Maps, Google Maps, Waze, and other navigation apps can help you
navigate cities to which you travel, they can also help you optimize your
travels locally. These AI-powered apps can help you choose the best route
to any destination on any given day and at any given time, taking into
account real-time incident reporting, road construction, and weather
conditions.

Don’t assume that your old familiar route will always be the
best one. Road conditions are constantly changing, and you
never know what traffic is going to be. These apps can help
predict the best route based on real-world conditions.
Optimizing Ride Sharing and Carpooling

Ride sharing services such as Lyft, Uber, and Via use AI algorithms to
match you with drivers in the shortest possible time. In addition, many
carpooling apps use AI to match you with other riders with similar routes
and destinations.

To find carpool apps in your area, search "AI-powered carpooling apps.

Finding Parking

In crowded metropolitan areas, it’s often a hassle to find a parking spot.

Fortunately, AI can work with real-time data to both book parking spots in
advance and find open parking spots. These apps analyze historical parking
patterns to suggest the best places and times to park, often at discounted
rates.

Some of the most popular AI-powered parking apps include

Parkopedia (www.parkopedia.com)
SpotHero (https://spothero.com)
Valet EZ (https://valetez.com)

Optimizing the Use of Public Transportation

If you use public transportation at home or when you travel, AI can help
you obtain real-time transit information and advice. You can use AI to
determine the best transit options, the fastest routes, and the lowest-price
fares.

The best way to do all this is with an AI-powered public transportation app,
such as one of the following:

Citymapper (https://citymapper.com)
Moovit (https://moovit.com)
MyTransit (www.mytrans.it)
Transit (https://transitapp.com)
Umo (https://umomobility.com)

Summary

In this chapter, you read about the many ways AI can help you plan and
enjoy your trips and travels. You saw how AI can help you choose your
destination, prepare an itinerary, and prepare for your trip—from learning
local languages to figuring out what to pack. You also discovered how to
use AI to make it easier to get around town.

Just remember to use common sense when using AI for travel. AI doesn’t
always have the most up-to-date local information, and it can sometimes
make recommendations that don’t mesh with your needs. As with all things
AI, double-check the advice you get before you act on it.
Travel and transportation are two good examples of how AI will touch all
aspects of your daily life. In fact, many such uses of AI are already
happening behind the scenes, helping to power and optimize the many
mapping apps and travel services you use on a regular basis.

OceanofPDF.com
11

Using AI for Health and Wellness

IN THIS CHAPTER

Using AI to create personal fitness and nutrition plans

Using AI for your mental health
Using AI to understand health and wellness information

As you’ve learned throughout this book, there are many uses for artificial
intelligence. One of the most personally impactful uses is to improve your
individual health and wellness. Although AI will never replace a trusted
medical provider, it can help you improve your physical and mental health
and better understand the healthcare options available to you.

Using AI to Create Personal Fitness and Nutrition Plans

AI can be a useful health and wellness helper. It’s particularly useful in

creating personalized plans for your fitness and nutrition. By feeding it
information about your age, gender, weight, height, activity levels, medical
history, food preferences, and weight/fitness goals, any all-purpose AI tool
can create plans tailored to your personal needs and wants.
 Warning

You should consult with a healthcare professional before undertaking any

new fitness or nutrition regimen.

Creating a Fitness Plan

Let’s start with using AI to create a fitness plan. Everybody’s bodies are
different, and we each have unique goals for working out and staying fit.
You may be looking to maintain your current health, increase your
flexibility, build up your muscles, or something entirely different. Whatever
your personal goals, AI can help.

You can use any all-purpose AI tool introduced in Chapter 3, “Getting

Started with All-Purpose AI Tools,” such as ChatGPT, Meta AI, or
Microsoft Copilot, to create personalized fitness and workout plans. When
you feed it the proper information, the AI tool can recommend specific
exercises, workout routines, and intensity levels. It’s all a matter of feeding
AI the proper information through a detailed prompt.

What should your prompt include? For best results, make sure it includes
your

Age
 Weight
Height
Gender
Health status
Any physical challenges, injuries, or disabilities (such as bad knees,
asthma, or osteoporosis)
Desired type of workout (low impact, weight training, cardio, high-
intensity interval training, and so on), if you have one
What kind of exercise equipment you own (if any)

Warning

Be cautious when providing personal information to AI tools. Any

information you provide can be used for further training and responses by
that tool; your privacy is not ensured. If you’re not comfortable providing
personal details like your height, weight, and health issues to others, you
shouldn’t provide it to an AI tool, either.

Your prompt should also include any specific goals you might have. Here
are some examples:

Create a workout plan that helps me increase my mobility

List 20 stretches I can do to start my day
 Create a six-month weightlifting plan to help me build my leg
muscles
Create a yoga routine for cardio
Prepare an exercise routine for balance

Putting it all together, you’d enter a somewhat complex prompt that looks
something like this:

I’m a [age] year-old [gender]. I weigh [weight] pounds and am [height] tall.
I have bad knees and can’t walk long distances. Create a workout plan to
improve my cardio activity that takes less than 30 minutes a day.

When I fed this particular prompt into Meta AI, I got the plan shown in
Figure 11.1.
FIGURE 11.1

A personalized fitness plan from Meta AI.

Creating Nutrition and Meal Plans

Similarly, an all-purpose AI tool can create personalized nutrition and meal

plans that can help you maintain your weight, lose or gain weight, or just
eat healthier. Feed the same personal information into the AI tool along with
your dietary goals, using prompts like this:

Create a meal plan to help me lose 10 pounds in the next three

months.
 Create a meal plan to help me maintain my weight during the winter
months.
I currently eat a lot of snack foods and fast food and don’t have a lot
of time to prepare meals. Create a nutrition plan to help me eat
healthier.
I’m a type 2 diabetic and need to watch my carbs. Create a nutrition
plan to help me eat properly for my condition.

Note

You’ll get more medically accurate results if you also tell AI

what medications you’re taking.

Put all this together and you might create a prompt like the following:

I’m a [age] year-old [gender]. I weigh [weight] pounds and am [height] tall.
I don’t always have time to cook my own meals, so I eat out a lot. Can you
create a meal plan to help me lose 20 pounds in the next 6 months?

I fed this prompt into ChatGPT and got the recommendations you see in
Figure 11.2, along with sample meal plans for each meal and tips for eating
out more wisely.
FIGURE 11.2

A personalized nutrition plan from ChatGPT.

Beware of Unsafe Goals

AI seems to have limited guardrails built in to ensure that its advice is safe.
For instance, when I asked ChatGPT for a plan to lose 20 pounds in 10
days, it responded: “Losing 20 pounds in 10 days is an extreme goal that is
generally not considered safe or sustainable. A more realistic and healthy
weight loss rate is about 1-2 pounds per week. Rapid weight loss can lead to
muscle loss, nutritional deficiencies, and other health problems. However, I
can suggest a more gradual and sustainable plan to help you lose weight
effectively and healthily.” Still, it is always advisable to consult an expert
before starting a new regime.

Using AI-Powered Fitness and Nutrition Tools

All-purpose AI tools do a good job of creating personalized fitness and

nutrition plans, but specialized fitness and nutrition tools can do even better.
These tools not only provide personalized plans but also monitor your
progress and, often, serve as virtual coaches for your workout routines.

These tools all work in similar fashion. You input your vital statistics and
your goals, answer some questions about your lifestyle and preferences, and
get back a personalized plan. Some of these tools interface with personal
fitness devices, such as smartwatches and fitness trackers, to monitor your
performance in real time. They then track your progress and adjust your
routines as necessary.

Most of these AI-powered fitness and nutrition tools are mobile apps you
use with your smartphone. This way you always have the tool with you
when you’re eating out or working out. Some of the more popular of these
AI-powered apps include

ArtiFit (https://artifit.app) creates personal workout plans, uses your

phone’s camera to track body movement, and provides real-time
feedback.
Coachify.AI (https://coachify.ai), which is tailored for fitness enthusiasts,
provides personalized workout experiences, tracks progress, and ensures
proper exercise technique.
Fitbod (https://fitbod.me) creates personalized workouts and learns from
each workout to enhance your continuing progress.
FitnessAI (www.fitnessai.com), which is designed for weight training,
optimizes sets, reps, and weight each time you work out.
GymBuddy AI (www.gymbuddy.ai) creates personalized workout
programs tailored your fitness level and goals.
Lifesum (https://lifesum.com) lets you track your meals (including carbs,
protein, and fat) versus your personal meal plan and uses AI to identify
foods via image recognition.
MikeAI (www.mikeai.co) offers personalized AI-generated fitness
assessments, workout plans, and meal plans.
MyFit-AI (https://myfit-ai.com) creates personalized fitness programs
and meal plans.
MyFitnessPal (www.myfitnesspal.com) is an all-in-one fitness, food, and
calorie tracker.
Noom (www.noom.com) is another all-in-one app that creates
personalized meal plans and tracks meals and exercise; it provides
statistics and insights to help you stay motivated over time.
Planfit (https://planfit.ai) has personalized workout plans and access to a
personal AI trainer.
 TempoFit (https://tempo.fit) offers personalized weight training plans
that work with your smartphone camera to monitor your form and
progress.

While some of these apps are free, most offer some sort of subscription plan
with monthly or yearly payment options.

Using AI for Your Mental Health

AI isn’t just for your physical health. You can use AI to help maintain and
improve your mental health. This section covers some of the ways you can
use AI in this fashion. But always work with your healthcare provider if you
have mental health concerns. AI is not a substitute for professional
guidance.

Warning

ChatGPT, Google Gemini, Microsoft Copilot, and other all-purpose AI tools

were not specifically designed to address mental health needs. They can
lack diagnostic accuracy and often misinterpret symptoms. While these
tools can provide useful assistance, they should not be used as a substitute
for professional counseling and psychotherapy.
Providing Information and Resources

The first way that AI can assist your mental health needs is by providing
information about mental health disorders, coping strategies, and resources.
Just ask any all-purpose AI tools questions such as

What are the symptoms of depression?

What causes bipolar disorder?
How can I cope with anxiety?
What resources are there to help with eating disorders?

Encouraging Journal Writing

One approach to better mental health is to keep a daily journal about your
feelings, thoughts, and experiences. Journal writing can help you process
confusing emotions and gain self-awareness.

All-purpose AI tools can help you start journaling and continue with it on a
regular basis. Just prompt an AI tool with something along the following
lines:

Can you help me start writing a mental health journal?

Once you get started, you may have trouble putting your thoughts into
words. This is another area where AI can help. Just prompt the AI that you
need to create a daily journal entry and then enter some of what you felt or
experienced that day. You might start with something like this:

I need to create a daily journal entry.

Figure 11.3 shows how Claude AI responded. Just continue answering the
AI’s questions and you’ll have your entry.

FIGURE 11.3

How Claude AI can help you create a journal entry.

Guiding Cognitive Behavioral Therapy Exercises

Cognitive Behavioral Therapy (CBT) can help you identify and deal with
negative thought patterns that may be causing stress and affecting your
mental health. Any all-purpose AI tool can lead you through CBT exercises
with a prompt like this:

Guide me through some CBT exercises.

Figure 11.4 shows how Google Gemini responded.

FIGURE 11.4

Letting Google Gemini guide you through CBT exercises.

Guiding Relaxation Exercises

Similarly, all-purpose AI tools can provide instructions for various

relaxation techniques that can help improve your mental health. Here are
some example prompts:

Guide me through some relaxation exercises

or

Help me relax

Figure 11.5 shows what Microsoft Copilot recommended.

FIGURE 11.5

Getting relaxation advice from Microsoft Copilot.

Using Chatbots for Automated “Therapy”

While interfacing with an AI chatbot will never replace facetime with a

human counselor or therapist, you can still get some value by talking about
your problems with AI. For people who might otherwise be hesitant about
seeing a therapist, AI chatbots, such as those offered by most all-purpose AI
tools, can serve as a gateway to in-person therapy. In addition, some
younger people may be more comfortable with a technological solution
than talking to someone in person.

You can initiate a discussion with an AI chatbot with a simple prompt, such
as

I’d like to talk about a problem I’m having.

The chatbot should respond, asking for more information. You then respond
to that response and continue the conversation. (Figure 11.6 shows the start
of such a conversation with ChatGPT.) It’s not the same level of counseling
you’d get from a trained professional, but it might be enough to help a little.
FIGURE 11.6

Starting a “counseling” session with ChatGPT.

Note

If having a text-based discussion with an AI chatbot feels too

impersonal, consider a chatbot from a multimodal AI tool,
such as that offered by the ChatGPT mobile app. These
voice-activated chatbots let you talk to the AI and then
 respond via computer-generated voices. It’s a lot more
natural approach that some people will appreciate.

Using Purpose-Built AI for Mental Health

In addition to the functionality offered by all-purpose AI tools, several AI

apps are designed especially for mental health needs. Some of the more
popular apps include

Calm (www.calm.com) offers stress relief, breathing and grounding

exercises, and other tools to relieve anxiety.
Happify (www.happify.com) provides exercises, activities, and games to
improve your emotional well-being.
MindDoc (https://minddoc.com) offers real-time mood tracking and
analyzing patterns to provide useful insights into your emotional well-
being.
Replika (https://replika.com) is an AI chatbot that many use to share
their thoughts and emotions.
Rootd (www.rootd.io) is designed to help individuals deal with panic
attacks.
Youper (www.youper.ai) uses CBT and AI chatbots to provide mental
health assessments, monitoring, and support.

In addition to these consumer-focused apps, many healthcare systems and

employers offer AI-based mental health solutions for their enrollees and
employees. Check with your employer or healthcare provider to see what’s
available to you.

Using AI to Understand Health and Wellness Information

One of the challenges in dealing with your health and wellness is

understanding symptoms, disorders, diagnoses, and the like. There’s always
technical medical lingo involved, and it isn’t always easy to understand. AI,
fortunately, can help.

Researching Medical Conditions and Diagnoses

You can use AI to find out a lot of useful information before you contact
your physician or healthcare provider. Sometimes this information can save
you a trip to the doctor or clinic—or convince you that you need an in-
person appointment.

In this regard, any all-purpose AI tool can do the job. Simply prompt the AI
with your symptoms and receive a list of what could be causing your
problem. For example, you might use one of these prompts:

I’m experiencing neck and upper back pain when I wake up in the morning.
What could be causing this?

or
I get stomach cramps and nausea after I eat ice cream. What could cause
this?

ChatGPT’s response to the latter prompt is shown in Figure 11.7.

FIGURE 11.7

ChatGPT’s analysis of some distressing symptoms.

AI can also help you learn more about specific conditions—causes,

common symptoms, seriousness, treatment, and the like. Just use a prompt
like this:

Tell me about bone spurs

Figure 11.8 shows what Meta AI knows about that condition.

FIGURE 11.8

Learning about a given condition from Meta AI.

source: www.meta.ai

You can also ask AI how to treat a specific condition. Use a simple prompt
like this one:

How can I treat an ingrown toenail?

Figure 11.9 shows what Google Gemini recommends.

FIGURE 11.9

Treatment advice from Google Gemini.

Use AI Health Advice with Caution

As with all things AI, you should treat AI’s medical advice with caution.
Just as with any other online research, AI sometimes serves up all scenarios
and not just likely ones. So don’t panic if AI brings up extreme conditions
or diseases in offering possible causes of your symptoms—you should
focus on the most common causes, not the extremely rare ones.

This is why it’s essential to verify any health-related information you

receive from AI with what’s available from trusted healthcare websites,
such as Healthline (www.healthline.com), Mayo Clinic
(www.mayoclinic.org), and WebMD (www.webmd.com). Just as important,
you should always discuss vital health issues with your physician or
clinician before taking any action on your own. AI isn’t a trained medical
professional; your doctor is.

Researching Medications

You can also use AI to research medications. If you simply want to find out
more about a given medication, ask AI a question like this:

Tell me about [medication]

Maybe you want to know the risks and side effects of a given medication, in
which case you’d use a prompt like

What are the side effects of [medication]?

(Figure 11.10 shows such a response from Microsoft Copilot.)

FIGURE 11.10

Researching medication side effects with Microsoft Copilot.

You can also use AI to uncover unfavorable interactions between

medications with a prompt like this:

Can I take [medication 1] along with [medication 2]?

If you want to find some alternatives to a given medication (maybe you’re

allergic to the original or it’s not covered by your health insurance), use the
prompt
What are some alternatives to [medication]?

Finally, if you just want to find out what drugs are used to treat a given
condition, use the prompt:

What medications are used to treat [condition]?

Interpreting Medical Results

We depend on physicians and clinicians to keep us healthy, but sometimes

they use medical jargon, acronyms, and language we don’t understand. It’s
becoming increasingly common for test results to be sent home to you via a
letter or patient portal without a call to explain the results. It’s even worse
when you examine your invoices or insurance documents; it’s often difficult
to figure out just what your results were or what you’re paying for.

Fortunately, AI can help translate medicalese into language that anyone can
understand. Enter confusing terms into the AI tool, along with the
appropriate prompt, like this:

What does [medical term] mean?

The AI should be able to explain it to you, as shown in Figure 11.11.

FIGURE 11.11

Google Gemini translates a fancy medical term into plain English.

Even better, ask AI to explain medical lingo or your doctor’s instructions in

plain English. Consider this prompt:

Explain hypertension to me like I’m an 8th grader.

When prompted in this fashion, Meta AI provided the easy-to-understand

description shown in Figure 11.12.
FIGURE 11.12

How Meta AI explains hypertension to an 8th grader.

Communicating with Healthcare Providers

You can use all your AI-based research to better communicate with your
healthcare providers. Based on what you find out from AI, you can prepare
more intelligent and targeted questions to ask your providers, as well as
direct their attention to specific issues identified by AI.

For example, if your doctor says you need to cut back on carbohydrates,
you can ask AI for help, like this:
My doctor told me I need to cut back on carbs. I don’t know what this
means. What do I need to do?

Figure 11.13 shows how ChatGPT responds.

FIGURE 11.13

How ChatGPT explains a doctor’s instructions.

Similarly, if you’re facing a surgical procedure and have some qualms and
questions, ask AI for assistance:

I am preparing for sinus surgery. Can you give me some questions for me to
ask my doctor about the surgery?
Figure 11.14 shows the questions Google Gemini provided.

FIGURE 11.14

Google Gemini prepares you for an upcoming surgery.

How the Healthcare Industry Is Incorporating AI

Not surprisingly, AI is being adopted throughout the healthcare industry for

a variety of tasks. Don’t be surprised if you find yourself interfacing with
AI the next time you schedule a visit.

Medical professionals are currently using or evaluating the use of AI for

 Scheduling appointments
Checking in for appointments
Assisting providers
Managing medications
Remotely monitoring patients
Analyzing patient data
Enhancing online patient portals
Answering patient questions
Assisting in diagnosing conditions and recommending treatments
Automating billing and other administrative tasks

In short, expect AI to touch just about every aspect of your future

healthcare. Ideally, this will help provide more efficient and accurate care—
and reduce frustrating wait times!

Summary

In this chapter, you learned many of the ways you use AI to improve your
physical and mental health. You learned how to use AI tools to create
personal fitness and nutrition plans, relieve stress, and better understand
medical conditions, diagnoses, and medications.

AI promises to be a boon to the medical profession and those of us who rely

on it for our health and well-being. Providing quality healthcare is a
complex proposition that can benefit from the “intelligence” and
automation provided by AI technology. AI will help you better manage your
own care and improve the care you receive from medical professionals.

OceanofPDF.com
12

Using AI to Help Caregivers

IN THIS CHAPTER

Using AI to help caregivers with health tasks

Using AI to help caregivers with financial and legal matters
Using AI for emotional support for caregivers
Using AI to provide virtual companionship and assistance
Examining other AI tools for caregivers

As you learned in the previous chapter, AI can be a boon for your personal
health and wellness. Not surprisingly, AI can also help family caregivers in
navigating their daily tasks, monitoring those in their care, and even
providing companionship for those in care.

Using AI to Help Caregivers with Health Tasks

Generative AI can be a real help to family caregivers. Although caregiving

can be rewarding, it is often something new and quite stressful for family
members, and they have lots of questions about what to and what not to do.
AI can provide the answers.

If you’re a caregiver and not sure what to do in a given situation, an all-

purpose AI tool such as ChatGPT, Google Gemini, or Microsoft Copilot can
help. When you run into any of the following situations, here are some
prompts to help.

Getting Smarter about Medical Information

As discussed in Chapter 11, “Using AI for Health and Wellness,” AI can be

a boon for anyone trying to understand medical conditions and medical
terminology. You can use AI much like an encyclopedia, asking it questions
about various topics or having it explain complex medicalese in everyday
terms.

This capability helps caregivers better understand the medical conditions of

those receiving care and the medical implications of the care they receive. It
also helps caregivers make more informed decisions and improve their
caregiving skills.

Extracting useful medical information from an all-purpose AI tool is as easy

as asking the right questions. For example, if you want to learn more about
a specific medical condition, you might use the prompt

Tell me more about [condition].

If you’re having trouble comprehending a doctor’s care instructions, use the

prompt

Please explain the following instructions in language a child would

understand [include instructions].
You can even ask AI for advice on what to do in a given situation, such as
when a person appears confused or has fallen and bruised a hip. Use the
following prompt:

The person in my care has [describe condition or situation]. What should I

do?

Warning

Remember, AI cannot substitute for advice from a trained medical

professional. Asking AI for help is a good first step, but always consult a
physician for all important medical matters. Remember, AI makes mistakes.

Working with Doctors

As you also learned in Chapter 11, AI can help you prepare those in your
care for upcoming doctor’s visits or hospital stays. Before accompanying
someone to an upcoming visit, use AI to prepare a list of questions for the
doctor with a prompt like this:

The person in my care is [describe the patient’s age, weight, and other
vitals] and has been experiencing [describe any current health issues or
conditions]. What questions should we ask the doctor during an upcoming
visit?
Also, as stated previously, you can use AI to translate anything the doctor
says that you don’t grasp into easier-to-understand language.

Providing Personalized Care Plans

AI can analyze the data you provide about the person for which you’re
caring to create a personalized care plan. This plan might include daily
schedules, medication maintenance, dietary guides, and exercise routines.
Just provide the AI tool with key information about the person and ask for a
care plan:

Please create a personalized care plan for a person for whom I’m caring.
They are [describe the patient’s age, weight, and other vitals] and have the
following conditions: [describe any current health issues or conditions].
They are in generally [excellent/good/fair/poor] health and require
[constant/intermittent] care. Include information about [any or all of the
following: daily schedules, medication maintenance, diet, and exercise].

Again, be wary of sharing personal information because it will feed into

AI’s learning database. And verify AI’s output, since it’s not always
accurate.

Monitoring Health Conditions and Identifying Trends and Issues

AI excels at analyzing data, identifying trends, and predicting future

outcomes. This capability can prove useful to caregivers monitoring the
well-being of those in their care.

If you track the person’s vitals on a regular basis, you can enter that
information into an all-purpose AI tool and ask the tool to identify trends or
predict possible oncoming medical conditions. You can do this for the
person’s weight, blood pressure, blood sugar, heart rate, and other key
metrics. Just use the following prompt (and be prepared to either cut and
paste data from another application or enter it manually):

Analyze the following [metric] data for a [enter the person’s age, weight,
and other vitals] individual. Please identify any trends and alert me to any
possible developing conditions.

Note that you don’t have to track this information manually. Most
smartwatches, such as the Apple Watch (www.apple.com/watch/), Google
Pixel Watch (store.google.com/us/category/watches_trackers), and Samsung
Galaxy Watch (www.samsung.com/us/watches/), can monitor heart rate,
blood oxygen level, and other vitals and report results back to caregivers or
medical professionals.

There are also numerous wearable personal pendants and other safety
devices designed with caregivers in mind. These devices typically include
much of the functionality of a smartwatch paired with remote functionality
so a caregiver can access the device from their smartphone. Some of these
devices also include two-way audio so the caregiver can communicate with
the person in their care.

Companies offering these personal safety devices include

Alert 1 (www.alert-1.com)
Bay Alarm Medical (www.bayalarmmedical.com)
LifeAlert (www.lifealert.com)
Lifeline (www.lifeline.com) (learn about discounts for AARP members
at www.aarp.org/membership/benefits/)
Medical Guardian (www.medicalguardian.com)
MobileHelp (www.mobilehelp.com)
Theora Care (https://theoracare.com)
UnaliWear (www.unaliwear.com)

Note

Some of these devices require a paid subscription to a

monitoring service for full functionality. There are many
health apps available for your smartphone that track an
individual’s vitals, including some that tie into a smartwatch
 or personal monitoring device to obtain key information.
You can find these apps in your phone’s app store.

Then there’s the Together app, shown in Figure 12.1, that uses your
smartphone to video selfies of a person to determine their vitals and share
that information with caregivers or loved ones. It’s easy to use; just have the
individual stare into the phone’s camera for 60 seconds or so and it uses AI
technology to determine their blood pressure, pulse rate, heart rate
variability, and respiratory rate. It’s pretty nifty. Learn more at
www.togetherapp.com.
FIGURE 12.1

Taking heart pulse and respiratory rate via selfie with the Together app.

Improving Nutrition

As also discussed in Chapter 11, you can use all-purpose AI tools to provide
nutrition advice or meal plans for the person in your care. Just prompt the
AI with information about that individual, any dietary restriction they have,
and any health goals they may want to achieve. Use a prompt like the
following:

Please provide a seven-day meal plan for a [describe person’s age, weight,
and other vitals] [male/female] with [describe any health conditions]. This
person needs to restrict carb consumption and is allergic to soy. They’d like
to maintain or increase their weight by a few pounds over the next month.

Using AI to Help Caregivers with Financial and Legal Matters

AI can help caregivers with more than just health-related issues. AI can also
be a good source of information and advice about the financial and legal
issues facing those caring for family members.

Managing the Financial Waters

One of the most challenging parts of being a family caregiver concerns

finances. Fortunately, AI is able to help answer a lot of financial questions.
Here are some prompts you may want to try to help you navigate healthcare
costs, insurance claims, and financial planning related to the person in your
care; some of these prompts should result in the AI tool asking you
additional questions:

Create a budget for managing caregiving expenses for a [describe

age and gender] living on their own with minimal care
What financial assistance programs are available for caregivers?
How do I apply for Medicaid or other government aid for
caregiving?
How do I claim caregiving expenses on my taxes?
What strategies can I use to manage my [enter age] father’s medical
expenses?
What is the best type of long-term care insurance for my [enter age]
father?
Are there any employer benefits or support programs available for
caregivers?
How much should I set aside in an emergency fund for caregiving
expenses?

Understanding Legal Issues

There are a plethora of legal issues surrounding long-term care. As a family

caregiver, you can use any all-purpose AI tool to ask questions such as
these:
 What legal documents should I have in place as a caregiver?
What do I need to know about estate planning for my [enter age]
parents?
My parent has approximately [enter dollar amount] in savings. How
can I best manage her finances on a weekly basis?
I’m on a fixed budget and my mother needs long-term care. How
can we manage this?
How do I obtain power of attorney for my elderly parent?
What is a healthcare proxy and how do I designate one for my
father?
What is a living will, and how do I create one for my mother?
What are advance directives, and why are they important in
caregiving?
What is guardianship, and when is it necessary for a caregiver to
seek it?
How do I become a legal guardian for my elderly parent?
What legal steps should I take to manage my parent’s finances
responsibly?

Warning

AI cannot substitute for professional legal or financial advice. Check with a

professional who can review your unique circumstances.
Receiving Useful Tips and Personalized Advice

You can ask any AI tool for caregiving tips. Use a prompt like this:

Can you give me some useful tips for caregiving for a family member?

When you enter information about the person receiving care, AI can
provide personalized care advice. Use a prompt like this:

I’m caring for my father. He’s [age] years old and living by himself in a
[house/apartment]. He has trouble walking and is no longer capable of
driving. His mind is still sharp, but his memory is starting to fade. He
doesn’t like people looking after him but needs assistance with many things.
Can you give me some advice on caring for him?

You can also ask for personalized advice for dealing with specific
situations. For example:

My mother is no longer capable of maintaining her house and yard by

herself. We’re afraid she’d react adversely to being put into an assisted care
facility. What options do we have in dealing with this situation?

Discovering Other Resources

Want to know more about resources available to you as a caregiver? AI can

locate and recommend local resources, support groups, specialists, and
more, based on your specific needs and the needs of the person receiving
care. All you have to do is ask:

What resources are available to caregivers in the [enter your

location] area?
Are there any local support groups for children caring for parents
with Alzheimer’s?
My [enter age] mother is having foot problems. Can you recommend
a specialist in our area?
Can you recommend someone to help me navigate Medicare options
for my [enter age] mother?
I’m a caregiver for my parents and I need more help, especially
when I’m at work. What can you recommend?

Just ask the AI tool for whatever you need wherever you may be.

Using AI for Emotional Support for Caregivers

AI can provide caregivers a sympathetic ear and can help reduce stress and
avoid burnout. AI chatbots can provide some of the emotional support
caregivers might need.

Most all-purpose AI tools provide interactive chat capability. Some, like

ChatGPT, provide voice chat, so you can talk with the AI instead of just
typing text messages. Sometimes after a long, frustrating day of caregiving,
you just need to talk a little bit about how it’s affecting you.
 Warning

AI chat tools should never replace genuine human contact and conversation.
If caregiving is really getting you down, consider engaging the services of a
therapist or counselor, or finding a support network to help you get through
the tough times. Learn about AARP’s caregiving resources and support at
aarp.org/caregiving.

Smart Technology for Caregiving

You can use various smart home devices to make your tasks easier and
automate tasks for the person in your care. While most of these smart
devices don’t currently use AI (or, if they do, only rudimentarily via
predictive AI), expect companies to more fully integrate generative AI in
the future. Future smart home technology can potentially use AI to better
predict user behavior and integrate information between devices to enhance
remote caregiving.

Consider the following:

Smart lighting helps turn lights on or off without having to get off the
couch.
 Smart thermostats control temperature on a preconfigured schedule or
learn an individual’s heating/cooling needs.
Smart doorbells let those in care see who’s at the door without having to
open it.
Smart door locks control who can enter a house or apartment—and
notify caregivers when a resident leaves unexpectedly.
Smart cameras let caregivers or family members monitor loved ones
from a distance.

These smart devices can be operated on preconfigured schedules, via

smartphone apps, or via voice-controlled smart speakers, such as the
Amazon Echo or Google Nest devices. Smartphone operation lets
caregivers operate these devices remotely, which is useful for when you
can’t be there in person. Smart speakers also let caregivers communicate
with and remotely listen in on those receiving care.

Using AI to Provide Virtual Companionship and Assistance

A newer and perhaps more significant use of AI in caregiving is the ability

to provide virtual companionship. Being able to converse with an AI
chatbot can help relieve loneliness and boredom.

As you learned in Chapter 11, virtual companionship is available today.

Most all-purpose AI tools enable text-based conversations if the person
receiving care is able to use a computer keyboard.
With other generative AI tools, such as ChatGPT, an individual can speak to
the AI tool and have the AI respond in a human-like voice. This enables
more natural conversation, especially as AI gets better at reading a person’s
mood and emotions and can respond with some semblance of empathy and
personality.

Expect AI-enabled chatbots to get more human-like over time and be

integrated into more and different physical devices. For example, the ElliQ
“AI sidekick” is a device with a tablet-based display, speaker, and
microphone that uses AI technology to carry on real-time conversations.
Think of it as a very smart Amazon Echo Show fine-tuned for an older
market. Figure 12.2 shows ElliQ in action; learn more at https://elliq.com.
FIGURE 12.2

ElliQ providing AI-powered companionship. (Photo courtesy of ElliQ.)

You can also expect future versions of Amazon Alexa, Apple Siri, and
Google Assistant to provide more realistic and sophisticated conversations.
The current versions of these voice assistants use predictive AI to guess at
what you want or ask for; newer versions will incorporate generative AI to
provide true one-on-one conversations.
Examining Other AI Tools for Caregivers

In addition to the apps and devices discussed elsewhere in this chapter,

there are a handful of AI-powered tools designed specifically for caregivers
that bear further examination, including

Arti is a home healthcare platform that integrates with third-party

devices and healthcare organizations to provide caregivers and assisted
living facilities with a variety of useful services. The Arti platform
enables remote monitoring and can also predict and uncover previously
hidden health problems. (Learn more at www.caredaily.ai.)
CareFlick is an AI-powered management platform for assisted living
organizations. It includes Yana, an AI-powered co-pilot that offers advice
and insights to caregivers, as well as care monitoring and insightful
reporting. (Learn more at www.careflick.com.)
CarePredict is a wearable device that uses AI to learn a person’s daily
activity patterns and alerts caregivers to changes that might indicate
health issues. (Learn more at www.carepredict.com.)
Medisafe is a medication management app that uses AI to provide
personalized medication reminders and tracks adherence. (Learn more at
www.medisafe.com/download-medisafe-app/.)
Vera is a smartphone app that uses AI-tailored music to improve
memory care. (Learn more at www.veramusic.com.)
While some of these tools, such as Arti and CareFlick, are targeted at larger
facilities and organizations, it’s likely that much of the functionality they
offer will filter down to individual caregivers in the future. The other tools
are available to individual caregivers today.

The Future of AI in Caregiving

The potential impact of AI on caregiving is enormous. It’s all a matter of

integrating advanced AI models into easy-to-use tools and devices.

In some ways, generative AI will make current devices better. Instead of

merely tracking a person’s heartrate and breathing, for example, AI will
enable health and fitness trackers to analyze a person’s health over the long
term and suggest needed care. Instead of requiring caregivers to physically
monitor surveillance cameras and audio feeds, generative AI will automate
the monitoring and know when a human caregiver needs to intervene.
Instead of just responding to simple questions (such as “what’s the weather
today?”), generative AI will turn today’s voice assistants into humanlike
companions.

Generative AI will do all this and more, naturally and unobtrusively. Those
receiving care will interface with these apps and devices via natural-
language voice commands, just like talking with a human caregiver.
Caregivers, loved ones, and healthcare professionals will be able not only to
monitor the individual but also respond to subtle trends and provide more
personalized care. The changes are likely to be incremental but significant.

It’s important, then, to keep abreast of new developments in AI-assisted

caregiving. We’re likely to see a lot of new products, services, and
functionality over the next few years that promise to make easier the lives
of caregivers and those they care for. And for certain tasks, people may
come to prefer a robot to a human to help maintain a level of dignity or
privacy.

Summary

In this chapter, you learned ways that AI, however peripherally, is used in
caregiving today. In particular, you learned all the different ways caregivers
can use AI to help them navigate their daily tasks, from better
understanding medical information to managing legal and financial issues.

In addition, you learned how generative AI chatbots can be used to provide

virtual companionship and assistance. You also learned about a few other
AI-assisted tools available to caregivers today.

While AI won’t singularly relieve the growing caregiver crisis, it will help
caregivers to do more with their limited time—and enable those receiving
care to live more independently without the need for in-person care. AI will
change the dynamics of caregiving—with major developments expected
over the next several years.

OceanofPDF.com
13

The Future of AI

IN THIS CHAPTER

What’s next for AI?

How will AI impact your life in the future?
What risks lie ahead?

If you’ve made it this far through this book, you’ve learned a lot about
artificial intelligence and how to use it in everyday life. What we haven’t
yet examined is what’s next for AI, what it’s likely to look like and do in the
near and not so near future.

The reality is that AI today is in its infancy, not unlike how the Internet was
in the mid-90s. Back then we had no idea that the Internet would lead to a
boom in online shopping (and the corresponding decline in physical retail
stores), personalized news feeds and all the information in the world at our
fingertips (including a new universe of disinformation), and the domination
of social media (often at the expense of real-world connections).

That’s where we are with AI today. We know it’s here, we know it’s
evolving, we know its initial impact, but we have no idea where it’s going
to be five, ten, or more years in the future. What we do know is that AI is
going to have a big impact—and how people use AI will inevitably evolve
as it becomes more a part of our everyday lives.

What’s Next for AI?

If you had a crystal ball back in 1965, you would have bought shares in
IBM at about $1.50 per share. Today that stock is going for almost $200 per
share.

That same crystal ball, at different times, might have predicted the advent of
global communications satellites, humans walking on the moon, the rise of
the Internet, and the end of the Cold War. Think what you could have done
with any of that information if you had it.

And if you had a crystal ball today and were looking at the AI market,
you’d… well, what would you do? What does the crystal ball say about
what’s next for artificial intelligence?

AI technology is advancing so rapidly that it’s impossible to know what

directions it might take in the months and years ahead. But let’s take a
guess.

AI Gets Smarter, Faster, and Less Expensive

It’s safe to say that AI technology will continue to advance. Maybe not at
today’s accelerated rates, but still rapidly, at least for the foreseeable future.
We can also predict that AI will get smarter and more accurate over time,
due to increases in computing power, the amount and quality of data fed
into large language models, and the sophistication of the algorithms and
programming behind the technology. The more data AI has to work with,
the more informed—and at least apparently “smarter”—it will become.

In practical terms, this means you can expect AI models to provide more
accurate answers to questions when queried for information. For example,
AI-aided healthcare may provide more personalized health
recommendations. Movies and TV shows are likely to take advantage of
faster, better, and lower-priced special effects, leading to onscreen fantasy
elements that put today’s computer-generated imagery to shame. And AI
image generators will no doubt inundate us with fantastic images that look
as real as a photograph but picture places and characters that only existed in
our imagination.

In addition, the way we interact with AI will become more natural. Instead
of typing text into a prompt box, we’ll talk with devices that can see and
hear everything around them. You’ll speak to these AI-powered devices and
they’ll respond, sounding as real as your neighbor next door. You won’t be
able to tell the difference between an AI chatbot and a human being. And, as
conversational-AI assistants can access calendars and other real-time data,
they can become even more helpful.
You can also safely predict that AI will get faster as computing power
continues to increase. Big AI tasks, like creating visually sophisticated
images, will be completed in the blink of an eye, instead of waiting minutes
to generate results. Complex mathematical and scientific calculations will
be completed as soon as you hit the Enter button.

How fast AI will evolve is less predictable. Will it be twice as smart or

twice as fast in one year, two years, or longer? There are too many variables
to even venture a prediction. Just know that history shows that technology
always gets faster and smarter and cheaper. The same will happen with AI.

From Artificial Intelligence to Artificial Superintelligence

Today’s artificial intelligence has not yet advanced to the level of human
intelligence. That’s the next step in AI development, to match and,
ultimately, exceed that of human beings.

Experts consider today’s artificial intelligence to be weak AI, or what some

call narrow artificial intelligence (ANI). AI that works like and performs at
the level of human intelligence is called strong AI or artificial general
intelligence (AGI). AI that exceeds the constraints of human intelligence is
called superintelligent AI or artificial superintelligence (ASI).

Put another way, ANI is typified by machines that imitate human behavior,
typically one task at a time. AGI involves machines that can continuously
learn from their experiences and thus approach true human intelligence. ASI
builds on that to create machines that are smarter than humans in all aspects
and measurement of intelligence.

ANI is what we have today and AGI could be achieved in the near future—
some say within the next few years. ASI, however, may take a lot longer to
achieve, if it can be achieved at all. ASI machines would have enormous
capabilities and the ability to learn and grow in an exponential fashion,
eventually achieving what some call the technological singularity. (We’ll
talk more about the singularity later in this chapter—and there’s a lot to talk
about.)

Bottom line: When betting on AI growth, it’s artificial general intelligence

that’s next up to bat—AI systems that approach and match that of us flesh
and blood humans. It’s likely that we’ll achieve AGI in our lifetimes, but
what happens after that is nearly impossible to predict. Figure 13.1 shows
one estimate of how AI will progress in the future, based on information
presented in philosopher Nick Bostrom’s book Superintelligence: Paths,
Dangers, Strategies.
Figure 13.1

A possible timeline to AGI and ASI. (Based on Superintelligence: Paths, Dangers, Strategies by
Nick Bostrom. Estimates of the timing and duration of takeoff were not included on Bostrom’s
original graph.)

AI Gets Embedded in Other Devices

Today, most of us use AI on dedicated websites or smartphone apps. Going

forward, AI will become part of other devices, either used seamlessly by the
device or easily accessible by users.

Consider, for example, self-flying delivery drones. Instead of having to be

remotely piloted at the source, embedded AI will help the drone navigate to
the desired address while dealing with any real-world obstacles that might
appear in flight. You won’t see the technology in use but will experience its
benefits.
Microsoft, as another example, is introducing a new generation of personal
computers that include a neural processing chip and embedded AI
capabilities. These Copilot+ PCs, as Microsoft is calling them, will be able
to enhance available lighting and visual effects for better-looking video
calls, provide real-time live captions and subtitles when playing videos, and
recall anything and everything you’ve ever done on that device. Plus you’ll
be able to use the embedded AI to generate any text or images you imagine.

What other devices might include embedded AI? A short list might include
smartphones, smart watches, fitness trackers, televisions and streaming
media players, smart home devices, and self-driving cars. The longer list
includes just about any device you can think of. (Yes, there will be—and, in
fact, already is—an AI-enabled toaster. It apparently makes perfect toast,
every time.)

AI and Self-Driving Cars

One of the most anticipated uses of AI is in self-driving vehicles. While

there are various AI-assisted features in cars today (such as adaptive cruise
control, lane-keeping assistance, and automatic emergency braking), most
of us aren’t quite ready to jump into a self-driving vehicle—at least not yet.

Experts refer to a vehicle that can operate without any human intervention
under all possible driving conditions—no steering wheel or other driving
controls needed—as a Level 5 autonomous vehicle. Unfortunately, the
industry isn’t quite there yet.

Whenever the auto industry develops the first practical Level 5 autonomous
vehicle, expect the initial uses to be for fleet vehicles. The trucking
industry, for example, is likely to be upended by self-driving vehicles that
replace the current ranks of long-haul truck drivers. The ridesharing
industry is also likely to replace human drivers with autonomous vehicles,
thus significantly impacting the millions of individuals currently driving for
Lyft, Uber, Via, and other ridesharing services.

The technical challenges in developing a true self-driving vehicle are

daunting and beyond what the tech and auto industries can accomplish
today. Most experts predict that we won’t see widespread use of Level 5
autonomous vehicles until the year 2040 or later.

AI Merges with Other Technologies and Services

Embedding AI in freestanding devices is just the first step. Experts are also
working on how to integrate AI with other technologies to improve their
operation and efficiency.

One potential merging of technologies involves AI and robotics. By

themselves, current-generation robots are pretty dumb; they have to be
guided by an external human operator. In contrast, AI-enabled robots (what
some call autonomous mobile robots, or AMRs) can learn to “think” on
their own in real time, freely navigating their environment and performing
more advanced tasks.

For example, Ameca (www.engineeredarts.co.uk/robot/ameca/) is an

autonomous humanoid robot (see Figure 13.2) that its manufacturer bills as
“the world’s most advanced human shaped robot.” Ameca merges artificial
intelligence with a human-like artificial body, enabling it to respond to
commands and questions in real time with lifelike facial expressions.

FIGURE 13.2

The Ameca AI-powered humanoid robot. (Photo courtesy Engineered Arts.)

Imagine an AI-enabled robot like Ameca in a retail store or restaurant. The
robot could greet customers by name, direct them to what they’re seeking,
and prepare and deliver fancy cocktails or coffee drinks—all without human
intervention. An AI-enabled robot could even make small talk with
customers and recommend specific products or specials based on their
conversations.

Note

AI-enabled robots also have a future in caregiving, as

discussed in Chapter 12, “Using AI to Help Caregivers.”

What other technologies can benefit from using AI? How about these:

AI and the Internet of Things—devices with sensors and software—

which could improve existing smart home technologies
AI with health trackers that could detect critical health issues
AI with digital assistants—which is already happening, with Amazon
planning to release an AI-powered version of Alexa and Google
integrating Google Gemini AI into its Google Assistant technology
AI with virtual reality (VR), such as AI-powered characters that appear
in your field of vision and interact with you when you’re wearing VR
 glasses
AI with videogame tech, to create more lifelike characters that interact
with you in real time
AI with holographic technology, to put a face and a body to your AI
assistant’s voice—and show you how to do things, just like a real person
in the room would

And there’s more. AI can merge with both technologies and services to
provide additional benefit to individuals.

For example, AI and robotics can merge with health coaching to create
empathetic artificial companions or with mental health therapy to provide
24/7 support and care. Both of these advancements could significantly
improve patient care, especially in a field that is already experiencing a
shortage of qualified workers.

Predicting this future is easy. AI becomes even more useful when it works
together with these and other technologies.

AI Gets Personalized

One other expected path for AI going forward is in creating personalized

experiences of all sizes and shapes. As AI advances, it will obtain more
personal data about each and every one of us. While this raises some
privacy concerns, it also opens up the potential to create products, services,
and experiences that are tailored to your individual likes and needs.
Imagine a world that is crafted just for you—not just your Internet
newsfeed or streaming video recommendations, but all your entertainment,
shopping, and education needs. You’ll literally experience a world of your
own, created by AI for your enjoyment.

How Will AI Impact Your Life in the Future?

How will all these possible developments affect you directly? It’s a given
that AI will touch many aspects of your life, whatever your job or social
standing. How you take advantage of AI’s benefits are totally up to you.

AI and You at Home

It’s a few years from now. You’re woken up at the designated time by the
soothing voice of your personal AI assistant coming to you from a nearby
speaker or in a holographic image. (Think of Amazon’s Alexa or the Google
Assistant, except much more advanced.) Your AI assistant provides the
weather forecast and your schedule for the day and recommends what
clothes to wear. It has also started cooking your breakfast to be ready when
you’re done in the shower, which is already running at your desired
temperature.

AI has already planned out your day and programmed your self-driving car
to take you where you need to go. Not a moment is wasted; everything is
ready for you at the exact moment you need it. If something new crops up
—maybe the carpool fails and one of your children needs to be picked up
from school—AI automatically adjusts everything in real time to
compensate.

Perhaps midway through the day you begin to feel unwell. You tell AI your
symptoms, and your AI assistant automatically diagnoses the problem. If
it’s something simple, like a cold, AI tells you how to treat it. If it’s more
serious, AI schedules a virtual visit with your doctor and transmits its
diagnosis to the physician’s office. The doctor uses AI tools to confirm the
diagnosis and prescribe the proper treatment; if a prescription is needed, it
is automatically transmitted to your pharmacy and delivered to your home
via AI-powered drone.

While you’re out doing things, your AI is busy making sure your home is
running in tip-top order. Your AI-powered self-driving lawn mower is
cutting the grass in perfect patterns, the house is automatically heated or
cooled to the right temperature in each room for whomever is (or isn’t) in it,
and every square inch of your floors are being cleaned by your AI-powered
robotic vacuum cleaner.

That evening, after a dinner planned by AI (but still prepared by you), you
settle down for some relaxing entertainment. AI knows what you like to
watch and puts it onscreen for you. It knows not to interrupt you while
you’re watching, but gently reminds you the best time to go to bed for your
personal sleep cycle.
And that’s a day with AI in the home.

AI and You at Work

In the future, AI will do a lot of your old work for you. If your
responsibilities include a lot of repetitive, time-consuming tasks, AI will be
a blessing.

For example, if you currently sit around all day plugging numbers into a
spreadsheet, AI can do that boring work for you. If you sit in front of a
computer screen conducting research or retrieving information, AI can do it
for you—much faster than you can. If you spend most of your day trying to
come up with new marketing or product ideas, AI can generate a number of
new ideas for you to choose from.

A lot of what you do will be directed by AI systems. AI will essentially be

running the day-to-day aspects of your company, only pulling people in
when they’re necessary. You’ll get your assignments from AI, so your work
will be targeted precisely where you’re needed.

Not only will AI help you get your work done faster, but it will help you be
more accurate. Run your work through an AI tool, and it will find any
mistakes you might have made. It will even find factual and logical
inconsistencies. Not that you can ever achieve perfection, but AI will help
you get close.
If you’re on the road, AI will plan the perfect routes and make all necessary
reservations. You probably won’t be on the road a lot, however, because
you’ll be conducting most of your meetings virtually. AI will make it look
and sound like everyone is in the same room, even though you’re all in
separate locations. Initially you’ll see a conference room onscreen with
everyone around the same virtual table; in time, your colleagues will appear
together holographically in 3D space.

For upper management, AI will monitor key metrics in real time and
determine when certain actions need to be taken. AI will feed the right
information to the C-suite so that executives can make better-informed
decisions. In fact, AI will recommend certain strategies and courses of
action, and those recommendations are likely to be followed. AI will, in
many ways, know more about the business than management does.

Bottom line (and business is all about the bottom line), AI will take over all
the mundane, routine, utterly mind-numbing manual tasks common in
businesses today. It will help run the business and make decisions for the
future. That will result in higher productivity and higher profits.

On the other side of the equation, some workers may get left behind in the
upcoming AI revolution. AI will replace some existing workers if their
skills are no longer needed. If AI can do what you currently do better and
cheaper, your workday may consist of training and searching for a new job
—which AI can also help you do.
AI and You at Play

All work and no play makes AI a dull technology, so it’s a good thing that
AI will play a significant role in your extracurricular life.

I’ve never considered myself an artist. In fact, I can’t draw a perfect circle,
and I have trouble coloring inside the lines. But with generative AI, I’ll be
able to create all sorts of images in all manner of styles just by telling it
what I want. That’s exciting.

AI will let you be creative in all sorts of ways. Want to create a new piece
of artwork to hang on your wall? AI can do it. Want to write a short story, or
your family’s memoirs? AI can do it. Want to compose that perfect piece of
music that’s been banging around your head for years? AI can do that, too.

With generative AI, you no longer need innate skills to be creative. AI can
take your ideas and run with them, in whatever medium you desire. Thanks
to AI, we could be on the cusp of a new creative era, one that enables
individuals of all types to pursue their creative muses.

And that’s not all. AI will help you get better at your favorite sports by
observing your performance and offering personalized advice. Want to
improve your golf swing? AI will show you how. Want your kids to be
better at baseball or soccer? AI will help train them. Have a free half hour
and want to get a little exercise? AI will come up with the perfect exercise
routine on the spot.
Let’s not forget more passive entertainment. As previously noted, AI knows
your viewing and listening habits and will recommend new movies, TV
shows, and music for your entertainment pleasure. You’ll get personalized
playlists and viewing queues, with really good recommendations for new
stuff you’ll probably like.

AI and You All Day, Every Day

Look forward even just a few years and you can see how AI will integrate
with almost all facets of your daily life. You’ll use AI to plan your day, to
communicate with others, to assist in work, and to help entertain you. If all
goes as it could, AI will improve your life in a multitude of both small and
big ways. It will be everywhere.

What Risks Lie Ahead?

The future of AI may not be altogether rosy. There are many ways that
artificial intelligence can be used for evil as well as for good—and many
experts fear that out-of-control AI could pose significant risks to our
society.

Some people are already using AI-generated images to mock, to taunt, to

sexualize, to ridicule, and to poke fun at both public and private figures. AI
makes it easier for people to violate others’ privacy and exploit biases; it
also makes it difficult to determine what is real and what is a deepfake.
While AI has the potential to create wonderful changes in our daily lives
and societies, it’s important to be aware of some of the large-scale risks
connected with letting the technology get too far ahead of regulators and
law.

We Lose Oversight and Control

As AI systems become more autonomous and capable of making decisions

on their own, there is a very real risk that we’ll lose oversight and control of
that technology and other systems. When we let AI make too many
decisions for us, we might come to rely on AI too much and lose the
capacity to make our own decisions—and the accompanying accountability.

Along the same lines, as we let AI manage more and more activities, our
own skills required to perform those activities may atrophy. This could be
as simple as not being able to add, subtract, multiply, and divide in our
heads (already a problem since the advent of the handheld calculator) or as
worrisome as diminishing interpersonal skills because we don’t interact
with real live human beings as often as before. And after years of blindly
following your GPS app, can you still read a map? If this happens, it won’t
necessarily be AI’s fault but rather our own, for letting AI take over too
much of our lives.
We Let AI Become Weaponized

It is an unfortunate fact: many of society’s most important technological

advancements have either come about as part of a war effort or been
enlisted to support a war effort. One need look no further than the use of
motorized vehicles and airplanes in World War I, radar and nuclear weapons
in World War II, and spy satellites during the Cold War.

Knowing this, it’s quite likely that AI will be co-opted by the military for
both defensive and offensive purposes. Consider these potential uses of AI
by domestic and foreign militaries—some of which are already in use:

Using swarm intelligence technology to guide hundreds of decentralized

drones in coordinated attacks to overwhelm an enemy’s defense systems
Using AI to make targeting decisions in warfare
Transferring attack decisions from humans to Lethal Autonomous
Weapons Systems (LAWS)
Developing armies of autonomous, so-called killer robots to fight on the
battlefield

Okay, that last one may seem a little far-fetched—until, that is, you realize
that the Marines have already tested an autonomous utility robot equipped
with an anti-tank weapon for battlefield use. It shouldn’t surprise you to
discover that the United States and China are currently in a LAWS arms
race, each side aggressively pursuing research and development of AI
systems for military purposes.
The possible weaponization of AI isn’t the sole purview of the world’s
militaries, however. Hackers can use (and are probably already using) AI to
develop sophisticated and dangerous malware to use in cyberattacks.
Machine learning can help hackers more quickly develop malicious code,
thus staying a step ahead of legitimate cybersecurity efforts. AI can help
automate cyberattacks, increasing their speed and scale. And AI can assist
malicious individuals in developing more targeted and effective phishing
scams.

Malicious actors may also be able to use AI to manipulate essential systems

for their own purposes. Imagine AI being used to infiltrate and disrupt
critical infrastructure, such as a nation’s power grid or water supply.
Whatever bad things humans can do, chances are AI can do it even better.

The Ultimate AI Risk: Achieving the Singularity

Many scientists and AI experts believe that, based on continuing increases

in computing power and advancements in AI models, at some point in the
future, AI will reach and surpass the level of human intelligence and
become cognizant. This point, where AI exceeds the intelligence of humans,
is called the singularity—and it could bode ill for us humans left behind.

Although some experts fear that a truly sentient AI could turn against its
human creators, others dismiss the entire concept of the singularity as either
implausible, improbable, or not quite as dire as others predict. To ease any
concerns you might have, know that there are just as many arguments
against the singularity occurring as there are for it happening. The chief
arguments against the singularity occurring include the following:

Technological advancements tend to level off over time. AI is

experiencing significant growth now, but that growth is likely to slow in
the coming years, meaning artificial intelligence will never become
artificial superintelligence.
AI will never be as smart as humans—because it isn’t human.
Artificial intelligence today relies almost totally on a few discrete forms
of input—essentially text, images, audio, and video fed into large
language models. Human intelligence, in contrast, is informed by a range
of sensory inputs that aren’t available to today’s AI models, including
touch, taste, smell, and other senses. Computers simply don’t and can’t
experience the world in the many ways that humans do and thus won’t
become human.
Resources are not unlimited. The fact that all resources are limited
includes those resources necessary to develop and power AI. We may
simply run out of the raw materials required to build enough powerful
servers—or not have enough electricity available—to reach the
singularity.
A single, large, universal artificial superintelligence is unlikely. A
singularity event requires a huge, universal AI model. It’s more likely
that AI will evolve into multiple smaller, more task- or industry-specific
models rather than a single model that tries to serve everyone and
 everything. In addition, it’s highly unlikely that multiple countries will
link their AI models to those of their enemies.
Human intelligence can evolve, too. Other experts think that the
singularity can be avoided by advancements in human intelligence.
Continuing developments in bioengineering, genetic engineering, and
mind-altering drugs could lead to an explosion in human intelligence
levels, tapping into previously unused or underused brain capacity.
We can pull the plug. Some believe the simplest defense against a
universal ASI model that might try to take over the world is to simply
turn it off—if, indeed, that’s possible. When experts sense that a given
ASI model might be edging close to desiring world domination, they can
figuratively flip the switch or pull the plug. AI needs power to work.
Deprive the computers that drive AI of that power and they quit working
—and AI isn’t very smart at all when its brains shut down.

All that said, it’s important to know that many experts in the field are
seriously concerned about the possibility of the singularity and how it might
possibly affect humankind. While the probability of the singularity
occurring may be low, it’s not something to be taken lightly.

Summary

The precise future of artificial intelligence is difficult to predict, other than

it will continue to get smarter, faster, and cheaper. The path of evolution
will likely move from today’s weak narrow artificial intelligence to stronger
human-like artificial general intelligence to the ultimate artificial
superintelligence—although at what speed remains to be seen.

So far we’ve only tapped the surface as to how we can use artificial
intelligence technology. Going forward, we’ll discover new and currently
unimagined uses for AI. As with any developing technology, how we
envision AI today is likely to be much different than how it actually ends up
being used.

We know that AI can automate repetitive operations. We know it can

conduct research and inform decision-making. We know it can generate
text, images, audio, and video to inspire creativity.

We don’t know what else AI can do or how we can use it. We will
eventually discover new and exciting uses for AI that will make our lives
better. We just don’t yet know what those uses might be. That will come
with time and experience, of which we’ll have much. As I know you’ll
discover on your own, the more you use AI, the more ways you’ll think of
to use it.

Whatever the future holds, AI will continue to develop and to become

smarter. How much smarter is up to us humans, as is how we choose to use
everything AI promises to deliver.

OceanofPDF.com
Glossary

AGI See artificial general intelligence.

AI See artificial intelligence.

AI generator A software or service that uses generative AI to generate

content—text, images, video, audio, code, or other media.

AI image generator A type of AI software or service designed to create

visual imagery from textual or other forms of input.

algorithm A set of rules or instructions that perform a task or solve a

problem, commonly used by AI.

all-purpose AI tool A versatile AI generator designed to perform a wide

variety of tasks. Popular all-purpose AI tools include ChatGPT, Claude AI,
Google Gemini, Meta AI, and Microsoft Copilot.

ANI See artificial narrow intelligence.

artificial general intelligence (AGI) An AI system that works like and

performs at the level of human intelligence.

artificial intelligence (AI) The ability of a computer or machine to mimic

human intelligence.
artificial narrow intelligence (ANI) AI systems designed to address
specific tasks but that do not meet human capabilities.

artificial superintelligence (ASI) A hypothetical AI that exceeds the

constraints of human intelligence.

ASI See artificial superintelligence.

automation The use of technology or processes to perform a task or

activity with minimal human input.

autonomous vehicle A self-driving car or other vehicle, powered by AI

technology, that can navigate and operate without human intervention.

chatbot A program designed to simulate conversations with human users;

AI-powered chatbots employ conversational AI (either written or spoken).

computer vision A technology that teaches computers to extract

meaningful information from images.

conversational AI A type of AI that enables machines to understand,

process, and generate human language in the form of a conversation.

data training set A collection of information and other inputs used to teach
an AI model.
deep learning A type of machine learning that uses artificial neural
networks to simulate human thought.

deepfake Artificially created media in which people’s likenesses—

including their voice, image, or statements—have been manipulated
without their permission, typically using AI.

digital assistant An app or service designed to assist users by answering

simple questions and performing simple tasks on demand.

GenAI See generative AI.

generative AI A newer type of AI designed specifically to generate or

create new content.

GPT Generative pretrained transformer, a type of AI language model

trained to answer questions and generate new content.

hallucination In the world of AI, an incorrect response to a prompt.

large language model (LLM) An extensive deep learning computational

model trained on large amounts of data.

LLM See large language model.

machine learning (ML) A subset of AI that uses algorithms to

autonomously learn processes without specific human programming.
ML See machine learning

multimodal AI An AI system that can process and integrate information

from multiple data types, such as text, sound, and images.

narrow AI See artificial narrow intelligence.

natural language processing (NLP) The subset of AI that enables

computers to analyze, understand, and derive meaning from human
language.

neural network A type of machine learning program that works in a similar

fashion to the human brain.

NLP See natural language processing.

personalization The act of designing something to meet a person’s

individual requirements or desires.

predictive AI AI designed to predict trends and based on past patterns and

algorithms; predictive AI has been around for a while.

prompt Text-based input given to an AI system to generate a specific

output or response.

sentiment analysis The process of identifying whether the emotional tone

of a message is positive, negative, or neutral.
singularity A hypothetical future point at which AI gains sentience and
technological growth and becomes uncontrollable and irreversible, resulting
in unforeseeable changes to human civilization.

smart devices Devices that can be controlled remotely via Internet-

connected apps or AI algorithms.

smart home A home equipped with multiple smart devices.

smart speaker A type of speaker and voice command device with an

integrated virtual assistant that offers interactive actions and hands-free
activation. Today’s most popular smart speakers include the Amazon Echo
and Google Nest devices.

speech-to-text (STT) A technology that translates the spoken word into

written text.

strong AI See generative AI.

STT See speech-to-text.

technical singularity See singularity.

text-to-speech (TTS) The process of translating written text into spoken

word.

TTS See text-to-speech.

voice assistant A digital assistant that interacts with users through voice
recognition and natural language processing.

weak AI See predictive AI.

OceanofPDF.com
Index

Symbols

2001: A Space Odyssey (film), 10

accommodations, booking with AI, 228

activity suggestions with AI, 148-149
adjustment stage, 13
Adobe Firefly, 178-180
advertising, AI’s impact in, 22
AGI (artificial general intelligence), 281
agriculture job losses, 35
Ahrefs Content Idea Generator, 108
AI (artificial intelligence). See also generative AI; predictive AI; super
AI
benefits of, 26-30
augmenting creativity, 30
automating repetitive processes, 26
improving business productivity, 26
improving digital assistants, 28-29
improving healthcare, 29
increasing availability, 27
 managing messages, 29
personalizing learning, 30
personalizing recommendations, 28
personalizing user experience, 28
reducing risk, 27
companies in, 15
defined, xviii, 2
ethical issues, 47
best practices for content creation, 54
bias, 32-34, 49-50, 131, 143
inaccurate content, 50-51, 93, 219
job application manipulation, 211
malicious usage, 49
plagiarism, 48-49
privacy, 34-35, 51, 94
schoolwork and AI, 47-48, 93
sharing AI-generated content, 47
in writing, 98-99
in fiction, 10
future of. See future of generative AI
in Google search results, 128-129
hallucinations, 130
history of, 7-9
identifying AI-generated content, 38
 music, 45-46
photos, 40-43, 168-169
propaganda, 46
verifying accuracy, 93
videos, 44
writing, 38-40
impact of, xviii, xix
legal issues
copyright, 51-53, 174, 177
liability, 38
misconceptions about, 6
professional writers and, 99
purpose of, 2
risks of, xix, 30-37, 130-131
bias, 32-34, 49-50, 131, 143
job loss, 35-36
misinformation, 31-32, 130-131, 159
mistakes, 36
privacy violations, 34-35, 51, 94
resource usage, 37
search engines and chatbots versus, 2-3
search engines versus, 125-127
AI aggregators, 81
AI boom
 1980-1987, 8
2000-present, 8-9
AI generators
all-purpose AI tools
ChatGPT, 64-67
choosing, 83-87
Claude, 67-69
defined, 58
freestanding versus embedded, 61-64
Google Gemini, 69-72
how to use, 59-61, 64-82
Meta AI, 73-74
Microsoft Copilot, 74-77
Perplexity, 77-79
Pi, 79-80
Poe, 81-82
prompts, 87-92
purpose of, 58
verifying content, 46, 93
for content outlines, 109
for dictation and transcription, 124
for editing content, 111-123
how to use, 16-18
for caregiver emotional support, 274
 for caregiver financial information, 272
for caregiver information, 268-277
for caregiver legal information, 272-273
for communication with healthcare providers, 263-264, 269
for cover letters, 210-211
for fitness and nutrition plans, 248-252
for health and wellness research, 258-263
for improving productivity at work, 216-222
for improving writing, 107-112
for interview preparation, 211-213
for job applications, 210-211
for job searching, 202-204
for managing meetings, 222-226
for mental health resources, 252-258
for navigation, 241-243
for resume writing, 204-209
for thank-you letters, 213
for transportation, 243-244
for travel planning, 228-236
for trip preparations, 236-239
during trips, 240-243
for types of writing, 99-107, 216
for virtual companionship, 275-276
for poetry, 107
 for search and research
all-purpose tools, 132-134
research-specific tools, 134-140
for topic suggestions, 108
for writing/rewriting content, 111
AI image detectors, 43
AI image generators
Adobe Firefly, 179-180
comparison of, 178-179
copyright, 174, 177
DALL-E, 180-181
Deep Dream Generator, 183-184
DeepAI AI Image Generator, 182-183
DreamStudio by Stability AI, 184-186
ethical usage, 176-178
functionality example, 166
Google Gemini, 186-187
Hotpot AI Art Generator, 187-188
Image Creator from Microsoft Designer, 189-190
Midjourney, 190-194
models for, 194
NightCafe, 194-196
OpenArt, 196-197
prompts, 167-168, 197-198
 types of images, 169-176
AI or Not, 43
AI Overviews, 72-73, 128-129
AI Poem Generator, 107
AI Poem Writer, 107
AI Text Classified, 40
AI text detectors, 40
AI winter (1987-1993), 8
aiApply, 212
Akkio AI Outline Generator, 109
Alert 1, 270
Alexa, 9, 28
algorithms, 3, 12
all-purpose AI tools
for caregiver emotional support, 274
for caregiver financial information, 272
for caregiver information, 268-274
for caregiver legal information, 272-273
ChatGPT, 64-67
choosing, 83-87
Claude, 67-69
for communication with healthcare providers, 263-264, 269
conversations via text, 160-161
conversations via voice, 161-164
creating presentations, 221
defined, 58
fitness plans, creating, 248-249
freestanding versus embedded, 61-64
Google Gemini, 69-72
for health and wellness research, 258-263
for hobbies
finding new, 153-155
online communities, 158
pursuing existing, 155-157
how to use, 59-61, 64-82
interview preparation with, 211-212
job searching with, 202-203
for mental health resources, 252-257
Meta AI, 73-74
Microsoft Copilot, 74-77
nutrition plans, creating, 250-251, 271
Perplexity, 77-79
for personal connections
activity suggestions, 148-149
sharing memories, 149
on social media, 158-159
understanding others, 149-152
Pi, 79-80
 Poe, 81-82
prompts, 87-92
purpose of, 58
researching employers, 203-204
resume writing, 204-205
for search and research, 132-134
search engines versus, 125-127
travel planning with, 228
verifying content, 46, 93
for virtual companionship, 275-276
for work-related content, 216
Alphabet (Google), 15
Amazon, 9, 15
Ameca, 283-284
AMRs (autonomous mobile robots), 283
Ancestry, 149
ANI (narrow artificial intelligence), 281
Anthropic, 15. See also Claude
Apple, 15
Apple Maps, 241
Apple Watch, 270
applying for jobs with AI, 210-211
art. See also images
AI’s impact in, 19, 30, 165
 commercial artwork, 177
creating fine art with AI, 172
art styles, AI-generated portraits, 174-176
Arti, 276
artificial general intelligence (AGI), 281
artificial intelligence. See AI (artificial intelligence)
artificial narrow intelligence (ANI), 281
artificial superintelligence (ASI), 281
ArtiFit, 251
Asana, 217
ASI (artificial superintelligence), 281
Asimov, Isaac, 10
AskYourPDF, 145
assessment stage, 13
Atmo, 239
atrophy of skills, 288-289
AudioPen, 124
automation of repetitive processes, 26
autonomous mobile robots (AMRs), 283
availability, increasing, 27

Babbel, 238
Baidu, 15
banking
AI’s impact in, 20
job losses, 35
Bard. See Google Gemini
Bay Alarm Medical, 270
Beatles, 22
Beautiful.AI, 221
benefits of AI, 26-30
augmenting creativity, 30
automating repetitive processes, 26
improving business productivity, 26
improving digital assistants, 28-29
improving healthcare, 29
increasing availability, 27
managing messages, 29
personalizing learning, 30
personalizing recommendations, 28
personalizing user experience, 28
reducing risk, 27
best practices, AI-generated content creation, 54
bias in AI, 32-34, 49-50, 131, 143
Bing AI. See Microsoft Copilot
Bing search engine, 76-77
booking travel with AI, 228
Booking.com, 236
books, AI in, 10
Bradbury, Ray, 10
bSafe, 243
business, AI’s impact in, 20
benefits, 26-27
risks, 35-36
business letters, writing with AI, 103-104
Busuu, 238

Calendly, 223
Calm, 257
Cameron, James, 10
CareFlick, 277
caregivers. See also healthcare
AI tools for, 276-277
communication with healthcare providers, 269
emotional support, 274
financial information, 272
future of AI, 277
generating personalized care plans, 269
legal information, 272-273
monitoring health conditions and trends, 269-271
 personalized advice, 273
researching medical conditions, 268
researching resources available, 274
smart devices for, 275
CarePredict, 277
carpooling with AI, 244
cartoons, drawing with AI, 169-170
CBT (Cognitive Behavioral Therapy), 254-255
central processing unit (CPU), 14
chatbots, 160-161
AI versus, 2-3
ELIZA, 7-8
managing messages, 29
for mental health therapy, 256-257, 275-276
ChatGPT
compared to Google Gemini and Meta AI, 83-87
DALL-E image generator, 180-181
how to use, 16-18, 64-67
interview preparation with, 211-212
pricing, 64
search and research with, 132-133
voice activation, 162-163
ChatGPT-4o engine, 162
choosing AI generators, 83-87
citing sources, 141
Citymapper, 244
Claap, 225
Clara, 223
Clarke, Arthur C., 10
Claude, 67-69
clickbait, 159
ClickUp, 217
Clockwise, 223
Coachify.AI, 251
Cognitive Behavioral Therapy (CBT), 254-255
collaboration at work with AI, 219
collages, creating with AI, 171
comic book art, drawing with AI, 169-170
commercial artwork, 177
communication
with healthcare providers, 263-264, 269
at work, 219
companionship, AI tools for, 256-257, 275-276
computer vision, 13
connections. See personal connections with AI
Consensus, 135-136
conspiracy theories, 131
control, loss over AI, 288-289
conversation starters, 148
conversations with AI, 160
via text, 160-161
via voice, 161-164
Copilot. See Microsoft Copilot
Copilot in Windows, 62-63, 77
Copilot+ PCs, 282
CopyAI Idea Generator, 108
Copyleaks, 40
copyright, 51-53, 174, 177
Corporate Headshots tool (Hotpot AI Art Generator), 188
cover letters, writing with AI, 210-211
CPU (central processing unit), 14
Creative Commons, 179
creativity
augmenting, 30
future impact of AI, 287-288
cultural insights with AI, 240
current news events, researching, 142

DALL-E, 9, 178-181
data collection stage, 11-12
data processing stage, 12-13
Decktopus, 221
Deep Blue, 8
Deep Dream Generator, 178, 183-184
deep learning, 12
DeepAI AI Image Generator, 17-18, 178, 182-183
deepfakes, 19, 31-32, 159, 177
DeepL, 220
DeepL Translate, 242
diagnoses, researching, 258-260, 268
dictation AI tools, 124
dietary plans, creating
with AI, 250-251, 271
for caregivers, 269-271
digital assistants
AI in, 284
improving, 28-29
documentation for travel, researching, 236-238
Dokkio, 219
DOMA, 219
Doodle, 223
DreamStudio by Stability AI, 178, 184-186
Duolingo, 238
 E

editing content with AI, 111-112

Grammarly, 112-114
Hemingway Editor, 114-116
HyperWrite, 116-117
ProWritingAid, 118-119
QuillBot, 119-120
Sudowrite, 121-122
Wordtune, 122-123
education
AI’s impact in, 22, 30
personalizing, 30
electricity consumption, 37
Elicit, 135-137
ELIZA, 7-8
Ellie, 219
ElliQ, 276
email messages
summarizing with AI, 219
writing with AI, 101-102
EmailTree, 219
embedded AI, 61-64, 282
emotional support for caregivers, 274
employers, researching, 203
entertainment, future impact of AI, 287-288
ethics of AI-generated content, 47
best practices, 54
bias, 32-34, 49-50, 131, 143
image generation, 176-178
inaccurate content, 50-51, 93, 219
job application manipulation, 211
malicious usage, 49
plagiarism, 48-49
privacy, 34-35, 51, 94
for schoolwork, 47-48, 93
sharing content, 47
in writing, 98-99
event planning, 148
exercise plans, creating
with AI, 248-249
for caregivers, 269
Expedia, 236

Facebook, 15
fact-checking sites, 46
factual information, researching, 143
fair use laws, 52
FamilySearch, 149
fantasy images, creating with AI, 171-172
Fathom, 225
fiction, AI in, 10, 106
films, AI in, 10
Final Round AI Interview Copilot, 212
finance
AI’s impact in, 20
job losses, 35
financial information for caregivers, 272
finding hobbies with AI, 153-155
finding online communities, 158
fine art images, creating with AI, 172
fine-tuning AI results for search and research, 144
Fireflies.ai, 124, 225
Fitbod, 252
fitness AI tools, 251-252
fitness plans, creating
with AI, 248-249
for caregivers, 269
FitnessAI, 252
flights, booking with AI, 228
Flowrite, 219
freestanding AI tools, embedded versus, 61-64
friends and family. See personal connections with AI
future of generative AI, 279-285
accuracy, speed, price, 280-281
embedded in devices, 282
in healthcare/caregiving, 277
impact of
at home, 285-286
in leisure and entertainment, 287-288
at work, 286-287
intelligence gains, 281-282
personalized experiences, 285
risks in, 288-291
loss of oversight and control, 288-289
singularity, 290-291
weaponization, 289-290
in robotics, 283-285
self-driving cars, 283

games, playing with AI, 20, 148

Gemini. See Google Gemini
Gemini for Google Workspace, 64
genealogy tools, 149
general-purpose AI tools. See all-purpose AI tools
generative AI
all-purpose AI tools
ChatGPT, 64-67
choosing, 83-87
Claude, 67-69
defined, 58
freestanding versus embedded, 61-64
Google Gemini, 69-72
how to use, 59-61, 64-82
Meta AI, 73-74
Microsoft Copilot, 74-77
Perplexity, 77-79
Pi, 79-80
Poe, 81-82
prompts, 87-92
purpose of, 58
verifying content, 46, 93
conversations with, 160
via text, 160-161
via voice, 161-164
defined, xix, xxii, 4
future of, 279-285
accuracy, speed, price, 280-281
embedded in devices, 282
 in healthcare/caregiving, 277
impact at home, 285-286
impact at work, 286-287
impact in leisure and entertainment, 287-288
intelligence gains, 281-282
personalized experiences, 285
risks in, 288-291
in robotics, 283-285
self-driving cars, 283
for hobbies
finding new, 153-155
online communities, 158
pursuing existing, 155-157
how to use
for caregiver emotional support, 274
for caregiver financial information, 272
for caregiver information, 268-277
for communication with healthcare providers, 263-264, 269
for content outlines, 109
for dictation and transcription, 124
for editing content, 111-123
for fitness and nutrition plans, 248-252
for health and wellness research, 258-263
how to use, 16-18
 for improving productivity at work, 216-222
for improving writing, 107-112
for legal financial information, 272-273
for managing meetings, 222-226
for mental health resources, 252-258
for navigation, 241-243
for poetry, 107
for topic suggestions, 108
for transportation, 243-244
for travel planning, 228-236
for trip preparations, 236-239
trips, usage during, 240-243
for types of writing, 99-107, 216
for virtual companionship, 275-276
for writing/rewriting content, 111
image generators
Adobe Firefly, 179-180
comparison of, 178-179
copyright, 174, 177
DALL-E, 180-181
Deep Dream Generator, 183-184
DeepAI AI Image Generator, 182-183
DreamStudio by Stability AI, 184-186
ethical usage, 176-178
 functionality example, 166
Google Gemini, 186-187
Hotpot AI Art Generator, 187-188
Image Creator from Microsoft Designer, 189-190
Midjourney, 190-194
models for, 194
NightCafe, 194-196
OpenArt, 196-197
prompts, 167-168, 197-198
types of images, 169-176
impact of, 19-23
in art, 19, 30, 165
in business, 20, 26-27, 35-36
in education, 22, 30
in finance and banking, 20
in gaming, 20
in healthcare, 21, 29, 247, 265
in marketing and advertising, 22
in music, 22
in software development, 23
in transportation, 23, 227
in writing, 21, 30, 36, 97
for job searching
cover letters, 210-211
 identifying job types, 202-203
interview preparation, 211-213
job applications, 210-211
networking, 204
researching employers, 203-204
resume writing, 204-209
thank-you letters, 213
for personal connections
activity suggestions, 148-149
sharing memories, 149
on social media, 158-159
understanding others, 149-152
predictive AI versus, 5
purpose of, 4
requirements, 14
for search and research
advantages, 129-130
all-purpose tools, 132-134
for current news events, 142
drawbacks, 130-131
for factual information, 143
fine-tuning results, 144
for historical events, 144
for performing tasks, 142
 prompts, 140-141
for purchases, 142
research-specific tools, 134-140
for scientific/technical information, 143
steps in, 10-14
adjustments, 13
assessment, 13
data collection, 11-12
data processing, 12-13
outcome prediction, 13
for summarizing information
for email messages, 219
for general information, 145
for large documents, 219
for meetings, 224-226
Generative Pre-trained Transformer (GPT), 64
GeoSure, 239, 243
Goldman Sachs, 36
Google, 15
AI in search results, 128-129
launch of, 9
Google Bard. See Google Gemini
Google Gemini
compared to ChatGPT and Meta AI, 83-87
 embedded in Google Workspace, 64
how to use, 69-72
image generator, 178, 186-187
previous name, 62
pricing, 69
search and research with, 133
Google Maps, 241
Google Pixel Watch, 270
Google Translate, 220, 242
Google Travel, 236
GPT (Generative Pre-trained Transformer), 64
GPT-3 AI language model, 9
GPU (graphics processing unit), 14
Grammarly, 111-114
GravityWrite Blog Outline Generator, 109
greeting cards, creating with AI, 173
GuideGeek, 230-231
GymBuddy AI, 252

hallucinations, 36, 130, 219

Happify, 257
Harrison, George, 22
health trackers, AI in, 284
healthcare. See also caregivers
AI’s impact in, 21, 29, 247, 265
communication with providers, 263-264, 269
fitness and nutrition AI tools, 251-252
fitness plans, creating with AI, 248-249
future of AI, 277
interpreting medical terminology, 262-263
job losses, 35
nutrition plans, creating with AI, 250-251, 271
researching conditions and diagnoses, 258-260, 268
researching medications, 261-262
Healthline, 260
Hemingway Editor, 111, 114-116
Hephaestus, 7
HERE WeGo, 241
Hero of Alexandria, 7
historical events, researching, 144
history of AI, 7-9
Hive Moderation, 43
Hiver, 219
hobbies
finding with AI, 153-155
online communities, 158
pursuing with AI, 155-157
holiday cards, creating with AI, 173
holographic technology, AI in, 285
home, future impact of AI, 285-286
Hopper, 236
Hotpot AI Art Generator, 178, 187-188
HubSpot Blog Ideas Generator, 108
human intelligence, surpassing of, 290-291
HyperWrite, 111, 116-117
Hypotenuse AI Outline Generator, 110
Hypotenuse AI Poem Generator, 107

I, Robot (Asimov), 10
“I Sing the Body Electric” (The Twilight Zone television episode), 10
IBM, 15
identifying AI-generated content, 38
music, 45-46
photos, 40-43, 168-169
propaganda, 46
verifying accuracy, 93
videos, 44
writing, 38-40
Illuminarty, 43
Image Creator from Microsoft Designer, 178, 189-190
ImageFX, 186
images. See also art
AI generators for
Adobe Firefly, 179-180
comparison of, 178-179
DALL-E, 180-181
Deep Dream Generator, 183-184
DeepAI AI Image Generator, 182-183
DreamStudio by Stability AI, 184-186
ethical usage, 176-178
functionality example, 166
Google Gemini, 71, 186-187
Hotpot AI Art Generator, 187-188
Image Creator from Microsoft Designer, 189-190
Midjourney, 190-194
models for, 194
NightCafe, 194-196
OpenArt, 196-197
prompts, 167-168, 197-198
identifying AI-generated content, 40-43, 168-169
legal issues, copyright, 174, 177
types of, 169-176
cartoons and comic book art, 169-170
collages, 171
 fantasy, 171-172
fine art, 172
greeting cards, 173
manufactured reality, 173-174
portraits, 174-176
improving
productivity, 216-222
communication and collaboration, 219
creating presentations, 220-222
generating content, 216
managing projects, 217-218
translation tools, 220
writing, 107-112
editing content, 111-112
Grammarly, 112-114
Hemingway Editor, 114-116
HyperWrite, 116-117
outlining content, 109-110
ProWritingAid, 118-119
QuillBot, 119-120
resumes and LinkedIn profiles, 206
rewriting content, 110-111
Sudowrite, 121-122
suggesting topics, 108-109
 Wordtune, 122-123
inaccurate AI-generated content, 50-51, 93, 219
Instagram, 15
instructions for tasks, researching, 142
Internet of Things, 284
Interview Prep AI, 212
interview preparation with AI, 211-213
Interviews by AI, 212
itineraries, creating with AI, 228-229
iTranslate, 242

job searching with AI

cover letters, 210-211
identifying job types, 202-203
interview preparation, 211-213
job applications, 210-211
networking, 204
researching employers, 203-204
resume writing, 204-209
thank-you letters, 213
jobs, loss of, 35-36
Jobscan, 206-207
jokes, 148
journal writing with AI, 253-254

Kasparov, Garry, 8
Kayak, 236
keyword matching, 126
keyword stuffing, 211
keywords in prompts, 140
Kickresume, 207-208
Kubrick, Stanley, 10

languages
learning, 238
translating, 241-242
large language models (LLMs), 12, 64
LAWS (Lethal Autonomous Weapons Systems), 289
Layla, 230-232
learning. See education
legal information for caregivers, 272-273
legal issues. See also ethics of AI-generated content
copyright, 51-53, 174, 177
liability, 38
legal services, job losses, 35
leisure time, future impact of AI, 287-288
Lennon, John, 22
Lenovo, 15
Leonardo da Vinci, 7
Lethal Autonomous Weapons Systems (LAWS), 289
letters, writing with AI
business letters, 103-104
personal letters, 101
Level 5 autonomous vehicles, 283
liability for AI, 38
LifeAlert, 270
Lifeline, 270
Lifesum, 252
Lingvanex, 220
LinkedIn profiles, improving with AI, 206
LLMs (large language models), 12, 64
local transportation with AI, 243-244
loss of oversight and control, 288-289
Lyft, 244

machine learning (ML), 12

malicious usage of AI-generated content, 49
managing projects with AI, 217-218
manufactured reality images, creating with AI, 173-174
manufacturing job losses, 35
MapQuest, 241
maps, AI usage in, 241, 243
marketing, AI’s impact in, 22
Mars rovers, 9
Martin, George, 22
Martin, Giles, 22
Mayo Clinic, 260
McCartney, Paul, 22
meal plans, creating with AI, 250-251, 271
medical conditions
monitoring trends, 269-271
researching, 258-260, 268
Medical Guardian, 270
medical terminology, interpreting, 262-263
medications, researching, 261-262
Medisafe, 277
meetings, managing with AI, 222-226
scheduling meetings, 223-224
taking notes and summarizing meetings, 224-226
memoirs, writing with AI, 104-105
memories, sharing with AI, 149
Memrise, 238
mental health
AI tools for, 257-258
CBT (Cognitive Behavioral Therapy), 254-255
chatbots for therapy sessions, 256-257, 275-276
emotional support for caregivers, 274
journal writing, 253-254
relaxation exercises, 255-256
researching information with AI, 253
messaging, managing, 29
Meta AI, 15
compared to ChatGPT and Google Gemini, 83-87
how to use, 73-74
pricing, 73
search and research with, 134
Meta Smart Glasses, 74
Microsoft, 15
Microsoft 365 applications, Copilot in, 77
Microsoft Bing search engine, 76-77
Microsoft Cooking Assistant, 75
Microsoft Copilot
embedded in Microsoft 365 applications, 77
embedded in Windows, 62-63, 77
how to use, 74-77
 previous name, 62
pricing, 75
Microsoft Copilot+ PCs, 282
Microsoft Designer, 75
Microsoft Fitness Trainer, 75
Microsoft PowerPoint, 222
Microsoft Translator, 220, 242
Microsoft Vacation Planner, 75
Midjourney, 179, 190-194
MikeAI, 252
MindDoc, 257
misinformation. See also inaccurate AI-generated content
identifying AI-generated content, 46
spread of, 31-32, 130-131, 159
mistakes in AI, 36
ML (machine learning), 12
MobileHelp, 270
models for AI image generators, 194
Monday.com, 217
Mondly, 238
monitoring health trends as caregiver, 269-271
Moovit, 244
Motion, 224
movies, AI in, 10
multimodal AI, 161-164
music
AI’s impact in, 22
identifying AI-generated content, 45-46
MyFit-AI, 252
MyFitnessPal, 252
MyHeritage, 149
MyTransit, 244

Nanonets AI, 219

narrow AI. See predictive AI
narrow artificial intelligence (ANI), 281
NASA Mars rovers, 9
natural language processing (NLP), 12
navigation, AI usage in, 241-243
Netflix recommendation system, 9
networking with AI, 204
neural networks, 12
news events, researching, 142
news media, job losses, 36
NightCafe, 179, 194-196
NLP (natural language processing), 12
nodes, 12
Noom, 252
notes, writing with AI, 100
nutrition AI tools, 251-252
nutrition plans, creating
with AI, 250-251, 271
for caregivers, 269-271
Nvidia, 15

online communities, finding, 158

OpenAI. See also ChatGPT; Microsoft Copilot; Perplexity
DALL-E image generator, 180-181
in Siri, 28
OpenArt, 179, 196-197
Oracle, 15
Otter.ai, 124, 225
outcome prediction stage, 13
outlining content with AI, 109-110
oversight, loss over AI, 288-289

packing assistance with AI, 239

PackPoint, 239
Packr, 239
Papago, 242
parking with AI, 244
Parkopedia, 244
performing tasks, researching, 142
Perplexity
how to use, 77-79
pricing, 77
search and research with, 134
personal connections with AI. See also conversations with AI
activity suggestions, 148-149
sharing memories, 149
on social media, 158-159
understanding others, 149-152
personal letters, writing with AI, 101
personal safety devices, 270-271
personal safety while traveling, 242-243
personality analysis, 150
personalizing
advice for caregivers, 273
care plans, 269
future of AI in, 285
learning, 30
personalized travel recommendations with AI, 240
 recommendations, 28
travel recommendations with AI, 228
user experience, 28
photos
creating with AI, 173-174
identifying AI-generated content, 40-43, 168-169
Pi, 79-80
plagiarism, 48-49
Planfit, 252
planning trips, AI usage in, 228-236
booking flights/accommodations, 228
creating itineraries, 228-229
with GuideGeek, 230-231
with Layla, 231-232
personalized recommendations, 228, 240
with Roam Around, 233
with Trip Planner AI, 234-235
with Wonderplan, 235-236
Poe, 81-82
poetry, writing with AI, 106-107
politics, risks of AI, 31-32
portraits, creating with AI, 174-176
PowerPoint, 222
predicting weather with AI, 239
predictive AI
defined, 3
examples of, xxii
generative AI versus, 5
private data usage, 34
purpose of, 3
preparing for trips, AI usage in, 236-239
learning local languages, 238
packing assistance, 239
researching documentation needs, 236-238
security information, 239
weather predictions, 239
prescription drugs, researching, 261-262
presentations, creating with AI, 220-222
pricing
Adobe Firefly, 178
aiApply, 212
ChatGPT, 64
Claude, 67
Consensus, 135
DALL-E, 178
Deep Dream Generator, 178, 183
DeepAI AI Image Generator, 178, 182
DreamStudio by Stability AI, 178, 184
Elicit, 135-136
Final Round AI Interview Copilot, 212
Google Gemini, 69, 178
Grammarly, 111
GuideGeek, 230
Hemingway Editor, 111
Hotpot AI Art Generator, 178, 187
HyperWrite, 111
Image Creator from Microsoft Designer, 178, 189
Interview Prep AI, 212
Interviews by AI, 212
Jobscan, 206
Kickresume, 207
Layla, 230-231
Meta AI, 73
Microsoft Copilot, 75
Midjourney, 179, 190
NightCafe, 179, 194
OpenArt, 179, 196
Perplexity, 77
Pi, 79
Poe, 81
ProWritingAid, 112
QuillBot, 112
 Rezi, 208
Roam Around, 230, 234
Scholarcy, 135, 137
Scite, 135, 139
Sudowrite, 111
Teal, 209
Trip Planner AI, 230, 234
Wonderplan, 230, 235
Wordtune, 112
Zety, 209
privacy
predictive AI and, 34
violations by AI, 34-35, 51, 94
productivity, improving, 26, 35-36, 216-222
communication and collaboration, 219
creating presentations, 220-222
generating content, 216
managing projects, 217-218
translation tools, 220
professional writers, AI and, 99
project management with AI, 217-218
prompts, 59-61, 87-92, 94
for AI image generators, 167-168, 197-198
for research, 140-141
 for current news events, 142
for factual information, 143
for historical events, 144
for performing tasks, 142
for purchases, 142
for scientific/technical information, 143
propaganda, identifying AI-generated content, 46
ProWritingAid, 112, 118-119
public transportation with AI, 244
purchases, researching, 142
pursuing hobbies with AI, 155-157

Q–R

QuillBot, 40, 112, 119-120, 220

Rainbow Weather, 239
Ray-Ban’s Meta Smart Glasses, 74
recipes, researching, 142
Reclaim, 224
recommendations
personalizing, 28
for travel, 228, 240
reducing risk, 27
relaxation exercises with AI, 255-256
repetitive processes, automating, 26
Replika, 257
requirements for generative AI, 14
research with AI
advantages, 129-130
all-purpose tools, 132-134
for current news events, 142
drawbacks, 130-131
for factual information, 143
fine-tuning results, 144
for historical events, 144
for performing tasks, 142
prompts, 140-141
for purchases, 142
research-specific tools, 134-140
for scientific/technical information, 143
researching
caregiver resources, 274
employers, 203
medical conditions and diagnoses, 258-260, 268
medications, 261-262
mental health resources, 253
travel documentation needs, 236-238
research-specific AI tools, 134-140
resource consumption, 37
responsibility for AI, 38
restaurant reservations with AI, 228
resumes
improving with AI, 206
writing with AI, 204-209
rewriting content with AI, 110-111
Rezi, 208
ride sharing with AI, 244
risks of AI, 30-37, 130-131
bias, 32-34, 49-50, 131, 143
in future of AI, 288-291
loss of oversight and control, 288-289
singularity, 290-291
weaponization, 289-290
job loss, 35-36
misinformation, 31-32, 130-131, 159
mistakes, 36
privacy violations, 34-35, 51, 94
reducing, 27
resource usage, 37
Roam Around, 230, 233
robotics, future of AI in, 283-285
Roomba, 9
Rootd, 257
Rosetta Stone, 238
Russakovsky, Olga, 32

safety while traveling, 242-243

Samsung Galaxy Watch, 270
scheduling meetings with AI, 223-224
Scholarcy, 135-138
schoolwork, usage of AI-generated content, 47-48, 93
Schwarzenegger, Arnold, 10
scientific information, researching, 143
Scite, 135, 138-140
Scribbr, 40
search engines. See also research with AI
AI versus, 2-3, 125-127
Google, AI in search results, 128-129
searches with AI
advantages, 129-130
all-purpose tools, 132-134
for current news events, 142
drawbacks, 130-131
for factual information, 143
fine-tuning results, 144
for historical events, 144
 for performing tasks, 142
prompts, 140-141
for purchases, 142
research-specific tools, 134-140
for scientific/technical information, 143
security information for travel with AI, 239
self-driving cars, 9, 23, 283
Sendsteps, 221
sentiment analysis, 150
sharing AI-generated content, 47
sharing memories with AI, 149
Sharly, 145
short notes, writing with AI, 100
short stories, writing with AI, 10, 106
Shortwave, 219
singularity, 290-291
Siri, 9, 28
Sitata, 239
SlidesAI, 221
smart devices for caregivers, 275
Smart Tissue Autonomous Robot (STAR), 29
Smart Traveler, 243
Smartsheet, 217
smartwatches, 270
social media
AI in, 9
creating posts with AI, 104, 158-159
misinformation on, 159
software development, AI’s impact in, 23
sources, citing, 141
SpeakAI, 124
SpotHero, 244
STAR (Smart Tissue Autonomous Robot), 29
Star Trek: The Next Generation (television series), 10
Starr, Ringo, 22
strong AI, 281. See also generative AI
Sudowrite, 111, 121-122
suggesting topics with AI, 108-109
SummarizeBot, 145
summarizing information with AI
for email messages, 219
for general information, 145
for large documents, 219
for meetings, 224-226
for web pages, 71
super AI, 5
superintelligent AI, 281
superReply, 219
Systran, 220

talking with AI, 160

via text, 160-161
via voice, 161-164
Teal, 209
technical information, researching, 143
television, AI in, 10
TempoFit, 252
TensorFlow, 9
The Terminator (film), 10
text
identifying AI-generated content, 38-40
talking with AI via, 160-161
thank-you notes, writing with AI, 102-103, 213
Theora Care, 270
therapy sessions with chatbots, 256-257, 275-276
tl;dv, 225
Together app, 271
Tome, 221
Tomorrow.io, 239
topics, suggestions from AI, 108-109
traditional AI. See predictive AI
training AI models, copyright, 52-53
transcription AI tools, 124
Transit, 244
translation with AI, 220, 241-242
transportation. See also travel, AI usage in
AI’s impact in, 23, 227
job losses, 36
local transportation with AI, 243-244
navigation with AI, 241-243
parking with AI, 244
public transportation with AI, 244
ride sharing and carpooling with AI, 244
travel, AI usage in
planning trips, 228-236
booking flights/accommodations, 228
creating itineraries, 228-229
with GuideGeek, 230-231
with Layla, 231-232
personalized recommendations, 228, 240
with Roam Around, 233
with Trip Planner AI, 234-235
with Wonderplan, 235-236
preparing for trips, 236-239
learning local languages, 238
 packing assistance, 239
researching documentation needs, 236-238
security information, 239
weather predictions, 239
during trips, 240-243
TravelSmart, 239
Trello, 218
trends (health), monitoring, 269-271
Trip Planner AI, 230, 234-235
Turing, Alan, 7
Turing Test, 7
The Turk, 7
The Twilight Zone (television series), 10

Uber, 244
Umo, 244
UnaliWear, 270
understanding others with AI, 149-152
United States Copyright Office (USCO), 53
user experience, personalizing, 28
 V

Valet EZ, 244

Vasco Translator, 242
Vera, 277
verifying AI-generated content, 46, 93
Via, 244
video games, AI in, 285
videos, identifying AI-generated content, 44
virtual companionship, 256-257, 275-276
virtual reality (VR), 284
virtual travel assistants, AI as, 229
voice, talking with AI via, 161-164
von Kemplelen, Wolfgang, 7
VR (virtual reality), 284

watermarks, 39
Watson, 9
Waymo, 9
Waze, 241
weak AI, 281. See also predictive AI
weaponization of AI, 289-290
weather predictions with AI, 239
web pages, summarizing, 71
web search. See search engines
WebMD, 260
weight-loss plans, creating with AI, 250-251, 271
Westworld (film/television series), 10
WhatToPack, 239
white fonting, 211
Winston AI, 43
Wonderplan, 230, 235-236
Wordtune, 112, 122-123
Wordvice AI, 220
workout plans, creating
with AI, 248-249
for caregivers, 269
workplace, future impact of AI, 286-287
work-related AI usage
for improving productivity, 216-222
communication and collaboration, 219
creating presentations, 220-222
generating content, 216
managing projects, 217-218
translation tools, 220
for managing meetings, 222-226
scheduling meetings, 223-224
 taking notes and summarizing meetings, 224-226
Wray, Christopher, 32
Wrike, 218
writing
AI’s impact in, 21, 30, 36, 97. See also AI generators
ethics of AI-generated content, 98-99
identifying AI-generated content, 38-40
improving with AI, 107-112
editing content, 111-112
Grammarly, 112-114
Hemingway Editor, 114-116
HyperWrite, 116-117
outlining content, 109-110
ProWritingAid, 118-119
QuillBot, 119-120
resumes and LinkedIn profiles, 206
rewriting content, 110-111
Sudowrite, 121-122
suggesting topics, 108-109
Wordtune, 122-123
types of, 99-107
business letters, 103-104
cover letters, 210-211
email messages, 101-102
 journal writing, 253-254
memoirs, 104-105
personal letters, 101
poetry, 106-107
resumes, 204-209
short notes, 100
short stories, 106
social media posts, 104
thank-you notes, 102-103, 213
work-related content, 216

X–Y–Z

Yandex Translate, 220

Youper, 257
Zapier, 219
ZeroGPT, 40
Zety, 209

OceanofPDF.com
OceanofPDF.com