Artificial Intelligence

Marks distribution��Quiz/Hourly/Midterm 20%�Lab Activities / Assignments 30%�Final Paper 50%

Classes / week (2+1)�Course Outline�2 hrs theory

2 hrs Lab

• Automation?
• How it helps?
• Automation is enough?
• Why need intelligence

Artificial Intelligence?

• Machines performing tasks requiring human intelligence
• Learning, reasoning, problem-solving, perception
• Simulation or augmentation of human intelligence

Formal Definitions of AI

• John McCarthy: Intelligent machines
• Systems that think like humans
• Systems that act rationally

Why Study AI?

• Transforms industries and society
• High demand for AI professionals
• Automation and efficiency

AI in Healthcare

• Disease diagnosis
• Medical imaging
• Drug discovery

AI in Education

• Smart tutoring
• Automated grading
• Personalized learning

AI in Business & Finance

• Fraud detection
• Algorithmic trading
• Chatbots

AI in Transportation

• Self-driving cars
• Traffic systems
• Route optimization

AI in Agriculture

• Crop monitoring
• Precision farming
• Yield prediction

AI in Entertainment

• Game AI
• Content recommendation
• Music and art generation

Ethical Issues in AI

• Bias
• Privacy
• Job loss

Careers in AI

• AI Engineer
• Data Scientist
• ML Engineer

AI Tools & Languages

• Python
• TensorFlow
• PyTorch
• Scikit-learn

Types / Division of AI�� Thousands of ways……

By Capability (what it can do)

Narrow AI (Weak AI) ← everything we have today

--Built for one specific task

--No real understanding outside its domain

​

Examples

• ChatGPT
• Face recognition
• Spam filters
• Recommendation engines

By Capability (what it can do)

General AI (AGI) ← not achieved yet

• Human-level intelligence across many domains
• Can learn, reason, and transfer knowledge like a person

Status: Theoretical / research goal

Superintelligent AI ← purely hypothetical

• Smarter than humans in all areas
• Self-improving

Status: Sci-fi + long-term speculation

By Functionality (how it behaves)

Reactive Machines

• No memory
• Respond only to current input

Example

• IBM Deep Blue (chess)

Limited Memory

• Uses past data to make decisions
• Most modern AI lives here

Examples

• Self-driving cars
• Chatbots
• Recommendation systems

By Functionality (how it behaves)

Theory of Mind (experimental)

• Would understand emotions, beliefs, intentions

Status: Research stage

Self-Aware AI

• Has consciousness and self-identity

Status: Hypothetical

By Learning Method (how it learns)

Machine Learning (ML)

• Learns patterns from data

Subtypes

• Supervised learning (labeled data)
• Unsupervised learning (finds patterns)
• Semi-supervised learning

Deep Learning

• Uses neural networks with many layers
• Great for images, speech, language

Reinforcement Learning

• Learns by trial and error
• Uses rewards and penalties