Artificial Intelligence (AI) – Detailed Notes
(Extended)
13. AI vs Human Intelligence
Feature Human Intelligence Artificial Intelligence
Learns from experience, emotions, Learns from data, algorithms, and
Learning
and reasoning training
Decision- Sometimes influenced by emotions,
Purely logical and data-driven
Making ethics, and social values
Can think abstractly, create art, Limited creativity (only mimics
Creativity
music, new ideas patterns)
Slow in calculations, fast in Extremely fast in calculations, limited
Speed
imagination in imagination
Quickly adapts to unknown
Adaptability Needs retraining or reprogramming
situations
Prone to mistakes due to fatigue, Less prone if programmed well; errors
Error Rate
stress, or emotions mainly due to bad data
14. AI Techniques
1. Search and Optimization
o AI uses algorithms like Depth First Search (DFS), Breadth First Search
(BFS), A algorithm* to find solutions in a problem space.
o Example: Pathfinding in Google Maps.
2. Logic and Reasoning
o AI applies rules and logic to make conclusions.
o Example: Expert systems in medicine (if symptoms = fever + cough + cold →
possible flu).
3. Knowledge Representation
o AI stores knowledge in structured forms:
Semantic networks (nodes and links)
Frames (data structures for representing situations)
Ontologies (knowledge bases)
4. Machine Learning Algorithms
o Supervised Learning (uses labeled data – e.g., spam detection).
o Unsupervised Learning (finds hidden patterns – e.g., market segmentation).
o Reinforcement Learning (learning by rewards and punishments – e.g.,
training a robot to walk).
15. Popular AI Tools & Frameworks
TensorFlow – Deep learning framework by Google.
PyTorch – AI/ML library by Facebook.
Keras – User-friendly deep learning library.
Scikit-learn – Machine learning for beginners.
OpenCV – Computer Vision library.
NLTK & SpaCy – Natural Language Processing tools.
GPT & BERT – Advanced NLP models.
16. Real-World Case Studies
1. Healthcare: IBM Watson
o Used in cancer diagnosis by analyzing medical literature.
o Provides treatment recommendations.
2. Transportation: Tesla Autopilot
o Uses computer vision and deep learning for self-driving cars.
o Can detect pedestrians, traffic lights, and obstacles.
3. Finance: Fraud Detection
o AI systems detect unusual credit card transactions.
o Example: Mastercard & Visa use AI to monitor fraud in real-time.
4. Customer Support: Chatbots
o Companies like Amazon, Swiggy, Zomato use AI chatbots for 24/7 customer
queries.
5. Entertainment: Netflix Recommendations
o AI suggests movies/shows based on your past viewing behavior.
17. AI in Daily Life – Examples You Use Everyday
Google Search – AI ranks results using machine learning.
YouTube/Instagram/TikTok – AI recommends videos.
Spam Filters – Gmail uses AI to filter unwanted emails.
Face Unlock – Uses AI-based facial recognition.
Maps & Navigation – AI finds shortest/fastest routes.
E-commerce – Amazon recommends products.
Voice Assistants – Alexa, Siri, Google Assistant.
18. AI Ethics and Concerns
1. Bias and Fairness
o AI can be biased if training data is biased.
o Example: Face recognition AI showing higher errors for darker skin tones.
2. Job Loss & Automation
o Factory workers, drivers, clerical jobs may be replaced.
o Need for “reskilling” in AI-driven world.
3. AI in Warfare
o Autonomous drones and killer robots → ethical debates.
4. Privacy Issues
o AI surveillance may violate personal freedom.
5. Control Problem
o What if AI becomes too powerful? → Debate on AI safety.
19. Important AI Terms
Neural Network: A system inspired by human brain neurons for learning.
Training Data: The dataset used to teach AI models.
Overfitting: When AI memorizes data but fails to generalize.
Chatbot: AI software that simulates conversation.
Computer Vision: AI ability to analyze images.
Reinforcement Learning: AI learns by rewards/punishments.
Big Data: Large datasets used for training AI.
20. Future with AI – Predictions
AI Doctors – Virtual assistants diagnosing patients at home.
AI Teachers – Personalized education for every student.
AI Judges – Supporting courts with unbiased decisions.
Space AI – Robots exploring Mars without humans.
AI Creativity – AI writing books, music, and making films.
Brain-Computer Interfaces – Merging AI with human brains.
21. Diagram Descriptions (for Notes)
Since I can’t directly draw here, I’ll describe diagrams you can add to your notes:
1. Timeline of AI Development (1950 → 2025)
o Mark milestones: Turing Test, Expert Systems, Deep Learning, ChatGPT.
2. Types of AI (Tree Diagram)
o AI → Narrow AI → General AI → Super AI.
3. AI Branches Chart
o Machine Learning, NLP, Robotics, Computer Vision, Expert Systems.
4. Applications Pie Chart
o Healthcare, Business, Transportation, Education, etc.
22. Summary
AI = Human-like intelligence in machines.
Started in 1956, now everywhere in daily life.
Three levels: Narrow AI, General AI, Super AI.
Core branches: ML, DL, NLP, Robotics, Computer Vision.
Used in healthcare, business, education, finance, and daily life.
Pros: Fast, accurate, tireless.
Cons: Expensive, job loss, ethical risks.
Future: Powerful but must be controlled responsibly.