ChatGPT

What is ChatGPT?
ChatGPT (Generative Pre-trained Transformer) is a natural language
processing (NLP) model that uses deep learning to generate human-like
responses to text-based conversations. It is a large-scale neural network
architecture that was pre-trained on vast amounts of textual data to
learn how to understand and generate language.
 Samuel H. Altman
Founder

OpenAI
OpenAI is an American artificial intelligence (AI) research laboratory consisting of the
non-profit OpenAI Incorporated (OpenAI Inc.) and its for-profit subsidiary corporation
OpenAI Limited Partnership (OpenAI LP). OpenAI systems run on the fifth most powerful
supercomputer in the world
Co-founders of ChatGPT
There are several legal and ethical issues surrounding the use of ChatGPT and other large language models. Few of them
include :

Ilya Sutskever Greg Brockman Andrej Karpathy Elon Musk

Language Model
Language models are statistical methods predicting the succession of tokens in sequences, using natural text. Large language
models (LLMs) are neural network-based language models with hundreds of millions (BERT) to over a trillion parameters (MiCS),
and whose size makes single-GPU training impractical

GPT-3
Generative Pre-trained Transformer 3 (GPT-3) is an autoregressive
language model released in 2020 that uses deep learning to produce
human-like text. The architecture is a decoder-only transformer
network with a 2048-token-long context and then-unprecedented
size of 175 billion parameters, requiring 800GB to store. The model
was trained using generative pre-training; it is trained to predict
what the next token is based on previous tokens
 Natural Language Processing (NLP)
Natural language processing (NLP) is an interdisciplinary subfield of linguistics, computer science, and artificial intelligence
concerned with the interactions between computers and human language, in particular how to program computers to process
and analyze large amounts of natural language data. The goal is a computer capable of "understanding" the contents of
documents, including the contextual nuances of the language within them

Natural Language Processing Pipeline

Sentence Word Stop Word Dependency Part-of-Speech

Stemming Lemmatization
Segmentation Tokenization Analysis Parsing Tagging

STEP 1 STEP 2 STEP 3 STEP 4 STEP 5 STEP 6 STEP 7

 Conversational AI
The term conversational AI (artificial intelligence) refers to technologies, like virtual assistants or chatbots, that can “talk” to
people (e.g., answer questions). Conversational AI applications are often used in customer service.

Automatic Speech Recognition

VOICE (Speech to text)
Text to Speech VOICE

Listening

Dialog Management and Natural

TEXT Natural Language Understand
Language Generation TEXT

Comprehending Forming Response

Machine Learning and Deep Neural Networks

Training Data
Training data is the data you use to train an algorithm or machine
learning model to predict the outcome you design your model to
predict. If you are using supervised learning or some hybrid that
includes that approach, your data will be enriched with data
labeling or annotation.
 Welcome to the new Bing
Ask complex questions Get better answers Get creative inspiration

What is ChatGPT Bing and how does it How does ChatGPT Bing use ML to What industries and use cases is
differ from other chatbot services? improve its responses over time? ChatGPT Bing best suited for?

Bing is a web search engine developed and operated by Microsoft, launched in 2009 . The revamped search engine is significantly
more powerful than the previous version of Bing, thanks to a new interface, ChatGPT AI tech, and some other changes. This could
be the beginning of a new era of searching the web, one in which you tell your search engine what you want in a far more natural
and intuitive way.
 Performance Evaluation of ChatGPT
Performance evaluation of ChatGPT is typically done using metrics such as perplexity, accuracy, fluency, and human evaluation.

Perplexity is a measure of Accuracy measures how Fluency measures how well Human evaluation involves
how well ChatGPT can often ChatGPT provides ChatGPT can generate natural- having human evaluators
predict the next word in a correct responses to user sounding responses that are assess the quality of ChatGPT's
sequence of text. It inputs. This metric is often grammatically correct and responses. This can be done
measures the level of used in task-oriented coherent. This metric is through user surveys or by
uncertainty in the model's chatbots where the goal is to important for open-domain having evaluators rate the
predictions, with lower provide accurate information chatbots that are designed to responses based on criteria
perplexity scores indicating or complete a specific task. generate human-like responses such as relevance, coherence,
better performance. in a conversational context. and overall quality.

Perplexity Accuracy Fluency Human Evaluation

AI Chat Bots
An AI chatbot is a computer program that uses artificial intelligence
techniques such as natural language processing (NLP), machine
learning, and cognitive computing to simulate human conversation
or communication. The purpose of an AI chatbot is to provide
automated assistance or support to users through a messaging
interface, such as a chat window, text message, or voice chat. AI
chatbots can be used for a variety of purposes, such as

• Customer service

• Technical support

• Information retrieval

• Personal assistance

• Entertainment
Ethical Concerns of ChatGPT
There are several legal and ethical issues surrounding the use of ChatGPT and other large language models. Few of them
include :

• The model may generate text that is similar or identical to existing copyrighted content, potentially causing a copyright
infringement issue.

• Model may generate personal information or sensitive data that could be used to identify or harm individuals.

• Could lead to the model perpetuating harmful stereotypes or discrimination

• GPT-3 and other similar models have been trained on a massive amount of data, which may have come from questionable
sources, such as scraping of websites or other data collection methods which could have legal and ethical issues.
Future of ChatGPT
While chatbots certainly exist, ChatGPT is one of the first to produce such in-depth answers as clearly and human-like as it
does. This opens up a whole new opportunity for the business world.

1 Streamlining operations 2 Streamlining operations

• Data entry Because ChatGPT is programmed to analyze immense
• Appointment scheduling amounts of data from its neural network and generate
insights and suggestions based on that data, businesses can
• Reduce the risk of errors
make informed, cutting-edge, data-driven decisions.
• Improve the accuracy of data
• Ensuring businesses' efficiency and effectiveness More information on consumers can also spark creativity in
• Saving costs business professionals, allowing them to stay ahead of the
• Improving productivity curve with new ideas that their customers will love.
• Strengthening the bottom line
 Free/Paid versions of ChatGPT
OpenAI has recently rolled out its new subscription plan for the AI chatbot - ChatGPT Plus. The Plus subscription plan of ChatGPT
gets additional capabilities and features. Notably, the company confirmed that it will not stop the free to use ChatGPT services and
the new Plus plan comes with more features for the users.

Free Plan ChatGPT Plus

• Available when demand is low • Available when demand is high

• Standard response speed • Faster response speed

• Regular model updates • Priority access to new features

Thanks
Instructions for use
You must credit Slidechef in order to use this template

What you are allowed to do :

Modify this template
Use it for personal or commercial projects

What you are not allowed to do :

Offer Slidechef templates for download
Sublicense, sell or rent any Slidechef content
Distribute Slidechef content unless it has been authorized
Include Slidechef content in an online or offline file
Acquire the copyright of Slidechef content
Slide Chef