Introduction to AI

Artificial Intelligence (AI) refers to the development of computer systems that can perform tasks
typically requiring human intelligence.
These tasks include :-
✓ Visual perception
✓ Speech recognition
✓ Decision-making
✓ Language translation
✓ Problem-solving
✓ Learning and adaptation
AI systems are designed to mimic human cognitive functions, but often use different approaches to
achieve similar results.
NITI Aayog Definition: "AI refers to the ability of machines to perform cognitive tasks like thinking,
perceiving, learning, problem solving and decision making."

Key Features of AI:-

• Learning from data
• Problem-solving
• Pattern recognition
• Decision making

History of AI
Important Events :-
→ 1950 - Alan Turing proposed the "Turing Test"

→ 1956 - Term "Artificial Intelligence" coined by John McCarthy at Dartmouth Conference

→ 1960s-70s - Early AI programs and expert systems

→ 1980s - Machine Learning development

→ 1990s - Internet growth boosted AI research

→ 2000s - Big Data and improved computing power

→ 2010s - Deep Learning breakthrough

→ 2020s - ChatGPT and advanced AI models

 Types of AI

Based on Capabilities :-
1. Narrow AI (Weak AI)
 Performs specific tasks
 Examples :- Siri, Google Translate, Chess programs
2. General AI (Strong AI)
 Human-level intelligence across all domains
 Currently doesn't exist
3. Super AI
 Beyond human intelligence
 Theoretical concept

Based on Functionalities :-
1. Reactive Machines
 No memory, responds to current situations
 Example :- Deep Blue (Chess computer)
2. Limited Memory
 Uses past data for decisions
 Example: Self-driving cars
3. Theory of Mind
 Understands emotions and beliefs
 Under development
4. Self-Awareness
 Conscious AI
 Theoretical

Application of AI
 Healthcare :- Medical diagnosis, drug discovery

 Transportation :- Self-driving cars, traffic management

 Education :- Personalized learning, smart tutoring

 Entertainment :- Netflix recommendations, gaming

 Finance :- Fraud detection, algorithmic trading

 Agriculture :- Crop monitoring, precision farming

 Smart Homes :- Alexa, Google Home

 Domains of AI
1. Natural Language Processing (NLP) :- Helps computers understand and process human
language.
Types of NLP :-
1. Natural Language Understanding (NLU)
o Converts human language to computer-understandable format
o Examples: Voice assistants understanding commands, sentiment analysis
2. Natural Language Generation (NLG)
o Converts computer data to human language
o Examples: Chatbots generating responses, automated report writing

Note :- The NLU is difficult than NLG.

Applications :-
→ Google Translate

→ ChatGPT

→ Voice assistants (Siri, Alexa)

→ Grammar checkers

2. Computer Vision (CV) :- Enables computers to interpret and understand visual information.
Applications :-
→ Face recognition (Facebook, phone unlock)

→ Medical imaging (X-ray analysis)

→ Self-driving cars (object detection)

→ Quality control in manufacturing

3. Data Science/Statistical Analysis :- Extracts insights and patterns from large datasets.
Applications :-
→ Business analytics

→ Market research

→ Weather forecasting

→ Sports analytics
 AI Subdomains
Machine Learning (ML) :- Algorithms that learn from data without explicit programming.
Types :-
→ Supervised Learning :- Learns from labeled data
(Email spam detection)
→ Unsupervised Learning :- Finds patterns in unlabeled
data (Customer segmentation)
→ Reinforcement Learning :- Learns through trial and
error (Game playing AI)
Examples: Recommendation systems, image recognition, fraud detection
Deep Learning (DL) :- ML technique using artificial neural networks with multiple layers.
Key Features :-
→ Mimics human brain structure

→ Excellent for complex pattern recognition

→ Requires large amounts of data

Examples :-
→ Image recognition (Google Photos)

→ Speech recognition (Siri)

→ Language translation (Google Translate)

→ Self-driving cars
Neural Networks :- Computing systems inspired by biological neural networks.
Components :-
→ Neurons (Nodes) :- Basic processing units

→ Layers :- Input, Hidden, Output layers

→ Weights :- Connection strengths between

neurons
→ Activation Function :- Determines neuron
output
Types :-
→ Artificial Neural Networks (ANN): Basic neural network

→ Convolutional Neural Networks (CNN): For image processing

→ Recurrent Neural Networks (RNN): For sequential data

How it works :-
1. Input data enters through input layer
2. Data processes through hidden layers
3. Each neuron applies weights and activation functions
4. Final output produced at output layer

Examples :-
→ Image classification

→ Speech recognition

→ Natural language processing

→ Medical diagnosis