AI Explorer Zone

Learn how AI works, and how to use it wisely

Free, simple lessons on artificial intelligence for kids and high school students, from "What is AI?" all the way to how chatbots predict words and how machines learn from data. Practise writing clear prompts, see how context changes an answer, teach a machine yourself, and learn to catch AI mistakes. Everything runs inside your browser. No sign-up, no real AI is connected, and nothing you type is saved or sent anywhere.

  • Beginner to advanced
  • Prompt writing
  • Machine learning basics
  • Checking AI answers
  • Safe school use
  • Works on phone
Your path

Three levels, one step at a time

Level 1 · Explorer

Understand AI

For beginners and younger students (about Classes 5 to 8). What AI is, how machines learn from examples, what a prompt is, and staying safe.

Level 2 · Builder

Prompt and check

For Classes 9 and 10. Data and models, the 5-part prompt, how context changes meaning, checking answers, and bias.

Level 3 · Expert

Inside the machine

For +1 and +2. Next-word prediction, neural networks, overfitting and testing, advanced prompting, and responsible AI.

Activity

Spot the unsupported claim

Compare two AI answers against a trusted source and flag the claims that the source does not back up.

Go to activity
Lessons

Understand AI: simple lessons

Pick your level, open a lesson, read it in about five minutes, and answer the two quick questions at the end. Each lesson has a "Try it" button that opens a matching tool.

Interactive tools

Try it yourself

These tools are teaching models made for this page. They copy the ideas behind real AI in a small, simple way, so every "AI" reply here is simulated. Real AI tools are much bigger, but they follow the same ideas.

Prompt Lab: how does the AI read your prompt?

Simulated: no real AI is used, and nothing is sent

Type a prompt you might give an AI chatbot. The lab shows which parts the AI can find in it and what it would have to guess. Try to get a score of 80 or more.

Prompt clarity

0

    A clearer version to try

    Fill in the [brackets] with your own details.

    Context Switcher: same words, different meaning

    Simulated replies, written for this lesson

    Many words have more than one meaning. An AI uses the words around your question (the context) to guess which meaning you want. Pick a question, then add different context and watch the reply change.

    What the AI sees

    Teach the Machine: sort mangoes and lemons

    A real (tiny) machine learning model, running in your browser

    Each dot is one fruit the machine has seen. Left to right is size, bottom to top is sweetness. Add training examples by tapping the board, then switch to "Test" and tap anywhere to place a mystery fruit. The machine looks at the nearest examples it knows and votes. This method is called k-nearest neighbours.

    Tap to:

    Machine says

    Switch to "Test a mystery fruit" and tap the board.

    Experiments to try

    • Empty the board and add only 2 examples. Is the machine good at guessing?
    • Add a wrongly labelled fruit, set k to 1, and test near it. Then set k to 5. What changes?
    • Add mangoes only on the right. Now test a small sweet fruit. Is the guess fair?

    Next-Word Predictor: how chatbots write

    A tiny real language model, trained only on the text below

    Chatbots write one word (or word piece) at a time by predicting what is likely to come next. This tiny model learns which word follows which in the training text. Change the training text and see how its predictions change.

    Most likely next words

    Press "Write a sentence".

    Notice: the model can write sentences that never appeared in its training text, and some of them are false. It only knows which words tend to follow each other, not what is true. Big chatbots look at many earlier words, not just one, and learn from billions of sentences, but the core idea is the same.

    Activity

    Compare two AI answers and spot unsupported claims

    Two AI chatbots answered the same question. Read the trusted source first. Then tap every sentence in the answers that the source does not back up: claims that disagree with it, or extra "facts" that it never mentions. Then press Check.

    Both answers are simulated, written for practice. In real life, the "source" is your NCERT textbook, your teacher, or an official website.

    Responsible use

    Using AI the right way at school

    AI can be a great study helper, but it should help you think, not think for you. Always follow your school's and teacher's rules first. Sort each situation: green means fine, yellow means okay with care, red means don't do it.

    The CHECK habit for every AI answer

    1. Compare with your textbook or class notes.
    2. Hunt for numbers, dates and names. These are the claims AI gets wrong most often. Verify each one.
    3. Evidence: ask "where does this come from?" A claim with no source is only a guess.
    4. Cross-check anything important with a second trusted source, such as ncert.nic.in or your teacher.
    5. Keep it yours. Write the final answer in your own words, and say when you used AI if your school asks.
    Glossary

    AI words in simple language

    Artificial intelligence (AI)
    Computer programs that do tasks that usually need human thinking, like recognising speech or answering questions.
    Machine learning
    A way of building AI where the computer learns patterns from examples instead of being given every rule.
    Training data
    The examples a machine learns from, such as labelled photos or lots of text.
    Model
    What the machine has learned from its data. It is used to make new guesses.
    Prompt
    The instruction or question you type to an AI tool.
    Context
    The extra information around your question that helps the AI understand what you mean.
    Hallucination
    When an AI states something false or made up as if it were true.
    Bias
    When an AI works better for some people or groups than others, often because its data was unbalanced.
    Token
    A word or piece of a word. Chatbots read and write text as tokens.
    Large language model (LLM)
    A very big model trained on huge amounts of text to predict the next token. Chatbots are built on them.
    Neural network
    A model made of layers of simple connected units, whose connection strengths (weights) are learned from data.
    Overfitting
    When a model memorises its training examples, including their mistakes, and does badly on new data.

    Keep learning

    Practise your board subjects, or test your skills with our free learning games.

    Exam Corner Test Yourself
      error:
      ✕