Basics of AI
Welcome! 👋
This app will teach you the basic ideas of Artificial Intelligence (AI) in very simple language.
You do not need any coding skills or AI background to use this app.
Just click on the tabs on the left to start learning!
Where to Start?
Basics
Learn what AI, ML, and other terms mean
Internals
See how AI learns from data
Applications
Discover AI uses in Physics, Math, Biology
Understanding AI Basics
Let's learn some important terms. Don't worry - we'll explain everything in simple words!
How AI Works (Simple View)
Data
Text, images, numbers
AI Model
Finds patterns
Useful Output
Answers, predictions
What is AI (Artificial Intelligence)?
Simple Definition:
AI means making computers do tasks that normally need human thinking. The computer can learn from examples and make decisions on its own.
Real-Life Examples:
- 📱 When your phone recognizes your face to unlock - that's AI!
- 🎵 When Spotify suggests songs you might like - that's AI!
- 📧 When Gmail filters spam emails automatically - that's AI!
What is Machine Learning (ML)?
Simple Definition:
Machine Learning is a type of AI where computers learn from examples instead of being given exact rules. It's like teaching a child by showing many pictures, not by giving a rulebook.
Real-Life Examples:
- 📸 Photo apps that group pictures of the same person together
- 🛒 Online shops that show "You might also like..." products
- 📝 Your phone keyboard predicting the next word as you type
Machine Learning is like how you learned to recognize fruits. Nobody gave you rules like "if it's red and round, it's an apple." You just saw many apples and learned!
What is Generative AI (GenAI)?
Simple Definition:
Generative AI is AI that can create new things - like writing text, making images, or composing music. It doesn't just recognize or classify; it generates something new!
Real-Life Examples:
- 💬 ChatGPT writing answers to your questions
- 🎨 DALL-E or Midjourney creating pictures from your descriptions
- 📝 AI helping you write essays or summarize long articles
What are AI Agents?
Simple Definition:
AI Agents are AI systems that can take actions on their own to complete tasks. They can plan steps, use tools, and work towards a goal - like a helpful assistant that can actually do things for you.
Real-Life Examples:
- 📅 An AI that can book appointments for you by checking your calendar and sending emails
- 🔍 An AI assistant that can search the internet, read articles, and give you a summary
- 🛒 A shopping assistant that compares prices across websites and finds the best deal
Think of AI Agents like a very smart helper. Instead of just answering questions, they can actually do tasks step by step!
What is RAG (Retrieval-Augmented Generation)?
Simple Definition:
RAG is a method where AI first searches for relevant information (like looking in a library) and then uses that information to give you a better answer. It combines searching with generating.
Real-Life Example:
- Imagine asking a question about your college syllabus. RAG would:
- First, search your syllabus documents
- Find the relevant parts
- Then write an answer using that information
Retrieval = Finding/fetching information
Augmented = Made better/enhanced
Generation = Creating the answer
How AI Learns (Training)
Let's see what happens inside an AI when it learns. We'll use a simple step-by-step visual!
Understanding Training in Simple Words
AI sees lots of examples - like thousands of sentences, pictures, or numbers.
AI finds patterns - it notices what things have in common.
AI stores these patterns as numbers - not as full sentences or pictures, but as mathematical patterns inside the model.
The AI doesn't store text like a notebook. It learns patterns and stores them as numbers. It's like how you don't memorize every cat you've seen, but you learned the pattern of what makes something a cat!
Interactive Training Simulation
Click "Start Training" to see how AI learns step by step!
Step 1: Input Data
These are example sentences the AI will learn from.
Key Takeaway
Training is like teaching by showing many examples. The AI learns the patterns in the data, not by memorizing everything, but by understanding what usually goes together!
How AI Creates Answers (Inference)
After training, AI can answer questions. This is called "inference" - let's see how it works!
What Happens When You Ask AI a Question?
You give input - your question or prompt goes into the AI.
AI uses its learned patterns - it thinks about what usually comes next based on its training.
AI predicts the next word - one word at a time, building a full answer!
Next-Word Prediction Explained
If you type: "The sun is"
AI thinks: "What word usually comes after 'The sun is'?"
Based on patterns, it might predict: "shining", "bright", or "hot"
Try It Yourself!
Type a short phrase and see how AI would build an answer word by word.
Large Language Models like ChatGPT predict one word (or token) at a time, very fast! They might generate hundreds of words per second, each based on what came before.
AI in Your Field of Study
AI is being used in Physics, Mathematics, and Biology in exciting ways. Let's explore!
🔭 AI in Physics
How AI Helps Physics Students & Researchers:
- Analyzing experiment data - AI can quickly find patterns in large datasets from physics experiments.
- Running simulations - AI helps simulate complex physical systems faster than traditional methods.
- Reducing errors - AI can identify and correct measurement errors in lab data.
- Visualizing graphs - AI tools can create beautiful plots and help interpret results.
- Discovering new physics - AI has helped find new particles and patterns in particle physics data!
📖 Student Story
"Priya is a physics student doing her final year project on pendulum motion. She collected lots of data but making graphs took hours. She used an AI tool to quickly plot all her graphs, fit curves to her data, and even find small errors she missed. Her project was completed in half the time!"
📐 AI in Mathematics
How AI Helps Mathematics Students:
- Solving problems step-by-step - AI can show how to solve equations, integrals, and more with explanations.
- Checking your answers - Quickly verify if your solution is correct.
- Visualizing functions - AI can draw 2D and 3D graphs of mathematical functions.
- Generating practice questions - AI can create unlimited practice problems for any topic!
- Proving theorems - Advanced AI is even helping mathematicians prove new theorems!
📖 Student Story
"Rahul was struggling with differential equations. He asked an AI chatbot to explain the concept like he was 10 years old. The AI gave him a simple water-flow analogy that made everything click! Then it generated 20 practice problems from easy to hard, and checked all his answers instantly."
AI tools like Wolfram Alpha and ChatGPT can solve most undergraduate math problems! But remember - use them to learn, not just to copy answers.
🧬 AI in Biology
How AI is Transforming Biology:
- Predicting protein shapes - AI can predict how proteins fold, which is crucial for understanding diseases.
- Drug discovery - AI helps find new medicines faster by predicting which molecules might work.
- Analyzing medical images - AI can detect diseases in X-rays, MRIs, and microscope images.
- Classifying cells - AI can identify different cell types automatically from images.
- Gene analysis - AI helps analyze DNA sequences and find genetic patterns.
📖 Student Story
"Sneha is a biology student interested in proteins. She heard about AlphaFold - an AI that can predict how proteins fold into 3D shapes. Using AlphaFold's database, she could see the predicted structure of a protein she was studying for her assignment, without doing months of lab work!"
🏆 Famous AI in Biology: AlphaFold
What is it? AlphaFold is an AI made by Google DeepMind.
What does it do? It predicts the 3D shape of proteins from their amino acid sequence.
Why is this important? Knowing protein shapes helps us understand diseases and create new medicines. This used to take years in a lab - AlphaFold can do it in minutes!
Impact: AlphaFold has predicted structures for almost all known proteins - over 200 million! Scientists worldwide use this for free.
How AI Chatbots Work
You've probably used ChatGPT or similar chatbots. Let's understand what happens behind the scenes!
Inside a Chatbot: Step by Step
You Type a Question
You write your question or request in the chat box.
Chatbot Reads & Understands
The chatbot converts your text into numbers it can process.
AI Model Thinks
The AI uses patterns from training to figure out a good response.
Chatbot Sends Answer
The response is converted back to text and shown to you!
Let's Understand Each Step Better
Step 1: You Type a Question
This is your "input" or "prompt". For example: "What is photosynthesis?"
Step 2: Chatbot Reads & Understands
Your words are broken into small pieces called "tokens" (words or parts of words). Each token is converted to numbers because computers only understand numbers!
"What is photosynthesis?" → [What] [is] [photo] [synthesis] [?] → [numbers]
Step 3: AI Model Thinks
The AI model looks at your question and all the patterns it learned during training. It asks itself: "Based on everything I learned, what would be a helpful answer?"
Then it predicts the answer one word at a time, very fast!
Step 4: Chatbot Sends Answer
The generated text is sent back to you through the chat interface. You see the complete answer on your screen!
Try the Demo Chatbot
This is a simple demo to show the idea. Type a question and see a sample response!
Modern chatbots like ChatGPT have been trained on billions of words from the internet, books, and more. But they don't remember your previous conversations unless specifically designed to!