Introduction to Artificial Intelligence

Prepared by: Dr T. J Alalibo & Prof S. Orike

Date: June 2025
 What is Artificial Intelligence?
• AI simulates human intelligence using machines.
• Definition: AI is the science of making machines perform tasks that normally
require human intelligence.
• AI refers to the autonomous capability of machines to simulate intelligent
behaviour.
• AI is both a scientific field and a set of applied technologies across
various sectors (medicine, transport, education, etc.).
• AI allows machines to learn from data, , adapt to new inputs and perform
tasks.
• Examples: Chatbots, spam filters, facial recognition.
 Defining AI – Key Features
• AI is a technology built on computer systems. It is not human brains,
which are natural intelligence

• It can be a product/process. That is, it Can be a system (e.g., chatbot), or

the algorithm behind it.

• It is a non-human entity that performs human-like tasks.

• It mimics or imitates human intelligence. It Tries to simulate aspects of

human thinking
 Purpose and Goal of AI
• AI studies intelligent agents/entities and their behaviors in
environments.
• These agents observe, learn from, and interact with environments to
achieve specific goals.
• It enables computers to perform tasks that would typically require
the cognition of human or other living things
• The goal: Maximize performance through continuous learning and
self-improvement.
 Core Processes in AI
• Learning: Acquiring data and understanding patterns.
• Reasoning: Drawing conclusions based on rules and
information.
• Self-Correction: Adjusting and improving decision-making
over time.
• AI systems adapt and evolve through iterations of these
processes.
 The concept of ‘Training" AI
• AI does not “think” like humans—it follows learned patterns.
• Training involves exposing the machine to large datasets so it can
recognize patterns and make decisions.
• Thus, AI learns through exposure to data and feedback.
• Training involves feeding data + algorithms.
• Like training a pet to behave in specific ways, with reward and
correction (feedback)
 Data and Algorithms in AI
• Big Data are huge amount of data. They fuel AI performance: the more
data, the more accurate the outcomes.

• More data = better performance (Big Data).

• Algorithms are step-by-step rules for processing data.

• Better algorithms enable solving complex tasks.

• Thus, Performance improves with better algorithms and a larger knowledge

base (data repository).
 AI Conceptual Framework

• Perception: Input via sensors, cameras and microphones to

collect real world data.
• Reasoning: Making logical decisions based on input and goals.
• Learning: Adapts or Improves over time with data and new
experiences.
• Interaction: Communicates with users (voice, text, gestures,
etc).
• Autonomy: Acts independently.
 Classification of AI (by Capability)
• AI is broadly divided into two categories:
o Narrow AI (Weak/Applied AI)
o Strong AI (Artificial General Intelligence)
o Super AI
• Each type differs in scope, complexity, and capability.
 Narrow AI (Weak AI)
• Task-specific; Designed for one task (e.g., facial recognition, voice
assistants).
• Simulates intelligence but lacks consciousness or emotional
understanding.
• Operates within predefined boundaries and rules.
• Examples: Siri, Alexa, self-driving cars, spam filters.
 General AI and Super AI
General/Strong AI:
▪ Human-level cognition (future): AI that can perform any intellectual
task a human can do and possesses beliefs, reasoning, learning, and
consciousness.
▪ In other words, this is future AI that can be fully human-like in
cognitive, emotional and moral emulation.
Super AI:
• Surpasses human intelligence hypothetical);
• Hypothetical and not yet developed; currently only exists in theory and
sci-fi.
Classification of AI by Functionality
• Reactive Machines: No memory; No memory or learning—just
responds (e.g., Deep Blue chess).
• Limited Memory: Temporary memory; Remembers recent data for a
short time (e.g., self-driving cars).
• Theory of Mind: Emotional intelligence; AI that understands human
feelings and beliefs (in development).
• Self-Awareness: Hypothetical conscious AI; Future concept—AI with
consciousness and self-reflection
AI tree
 The AI Tree
• AI includes multiple branches and technologies:
❑ Machine Learning: Systems learn from data and make predictions; deep learning
❑ Natural Language Processing: Understanding and generating human language
❑ Computer Vision: AI that "sees", like image and video recognition
❑ Speech: Converting between spoken words and text.
❑ Robotics: Machines that perform physical tasks autonomously.
❑ Expert Systems: Decision-making systems that mimic expert reasoning.
❑ Planning & Optimization: Choosing the best actions to achieve a goal.
• These key branches serve different purposes and rely on varying techniques.
 AI in Everyday Life
• Voice Assistants (Siri, Alexa): Understand voice commands and
provide information.
• Self-driving Vehicles: Use sensors and AI to navigate
• Recommendation Engines (Netflix, Amazon): Analyze user behavior
to suggest content.
• Medical Diagnostics, Social Media Feeds: AI predicts diseases from
medical images or data.
• ChatGPT: Conversational AI that answers questions and generates
text.
AI’s Role in Innovation and Research
❑ AI enhances problem-solving across multiple disciplines.
❑ Promotes innovation by developing novel approaches to challenges.
❑ Its learning capability allows it to contribute knowledge to evolving fields.

Importance of AI in Expert Tasks

❑ AI systems function similarly to human experts.
❑ They assess the environment and make decisions accordingly.
❑ Used in scenarios where expert-level reasoning and decision-making are
needed.
 The Future of AI and Challenges
❑ AI is becoming smarter and more independent.
❑ Challenges:
❖ Bias in data
❖ Privacy concerns
❖ Job automation/job loss
❖ Existential risks from uncontrolled AI
❑ Experts like Elon Musk and Stephen Hawking warned of AI risks.
❑ The future depends on ethical development, regulation, and human oversight
 FYI- Early History of AI
❑ 1943: McCulloch & Pitts created the first model of an artificial
neuron.
❑ 1950: Alan Turing proposed the Turing Test to evaluate machine
intelligence.
❑ 1956: “Artificial Intelligence” officially named during the
Dartmouth Conference.
 Founding Fathers of AI

❑ John McCarthy: Coined the term AI; developed LISP programming

language.
❑ Alan Turing: Theorized that machines could simulate human
reasoning.
❑ Marvin Minsky: Worked on machine perception and cognition.
❑ Allen Newell & Herbert Simon: Created Logic Theorist, first AI
program.
 Key Milestones in AI Development

▪ 1960s: ELIZA chatbot mimicked a therapist. Shakey navigated

space as a robot.
▪ 1980s: AI became formalized in academia and industry.
▪ 1995: Intelligent agents emerged, laying the groundwork for
assistants.
▪ 2010s: Rise of commercial AI – smartphones and assistants went
mainstream. Siri and Google Now.
 Generative AI Revolution
❑ Generative AI: Creates original content—text, code, art, etc.
❑ ChatGPT, DALL·E, and Midjourney are examples.
❑ Used in education, design, entertainment, and software development.
❑ OpenAI’s ChatGPT (2022) became the fastest-growing public AI tool
in history.