All courses
Beginner · Ages 13–17

AI & Python

Your child already uses AI every day. Now let them build one.

Build the engine behind ChatGPT from scratch — in your own Python code.

12 lessons Online or in-person · 1.5 hours per lesson Max 12 students
Book a Free Trial Class

// overview

About This Course

Almost every AI course teaches teens how to use AI — better prompts, clever tricks, someone else's chatbot. This one doesn't. Over 12 lessons, students write their own text generation engine in Python, line by line, until they understand exactly what a language model is doing when it writes.

They start in lesson one with a program that produces nonsense, then spend the course making it smarter: teaching it to learn from real text, to make choices, to remember context, to break words into pieces, to turn meaning into numbers. Each step is a little more Python and one big idea from how real AI works.

No coding experience needed. Here, the engine is the reason to learn Python — not the reward for already knowing it. Throughout the course, students also get access to DotCode AI, our own platform for exploring and comparing the latest AI models from leading providers.

// skills

What Your Child Will Learn

Prediction & Probability

Every AI model is really answering one question over and over: what comes next? Your child builds that machinery themselves and learns to think in likelihoods rather than certainties — the maths behind weather forecasts, medicine, and risk.

Data & Structure

A model is only as good as what it learned from. Students load real text, clean it, and organise it into structures the program can use — the unglamorous work sitting underneath every AI system in the world.

Patterns & Loops

Generating a sentence means repeating one small decision hundreds of times. Loops turn a single rule into an entire behaviour — and spotting that pattern is the shift from typing code to thinking like a programmer.

Testing & Debugging

Their engine will repeat itself, crash on unseen words, and write gibberish. Rather than giving up, your child learns to form a theory about why, test it, and fix it — the patience every builder needs.

Knowing When Not to Trust It

The part most AI courses skip. Once your child has built a model that confidently produces nonsense, they understand in their bones why a chatbot can sound certain and still be wrong. That's not a coding skill — it's judgement, and it's the one they'll use every day for the rest of their lives.

// curriculum

The Learning Journey

12 lessons · 6 modules · every module ends with a working project

1

Python Foundations

Set up a real development environment and learn the building blocks: variables, lists, loops, and printing. Everything is taught on text from day one, so students finish this module with a program that strings words into sentences — badly. That failure sets up the rest of the course.

2

Learning What Follows What

Students load a real book, song, or script and count which word follows which across thousands of words, then walk that table to generate text. This is the first thing their program has learned rather than been told. Python: dictionaries, reading files, and cleaning up messy text.

3

Chance & Creativity

Always picking the likeliest word makes the program loop forever. Students switch to weighted random choice and the text starts reading like broken English — then add a temperature dial controlling how safe or wild the output gets. Python: functions, parameters, and organising code that has grown too big.

4

Context & Tokens

Look back at three words instead of one and the writing sharpens, then starts copying the source verbatim. Students meet context windows and memorisation, then rebuild the model on letters instead of words — and finally get a real answer to why AI miscounts the letters in "strawberry". Python: tuples, nested loops, and slicing strings.

5

Meaning as Numbers

Words become vectors. Students measure how close two words are and see how a model can work out that "king" and "queen" are related without being told — then use DotCode AI to give the same prompt to real frontier models and compare how differently they respond. Python: working with lists of numbers and pulling data in from the internet.

6

Build Your Own & Demo Day

Students pick their own training text and build something with it — a lyric generator, a story machine, a bot that writes like their favourite author — then present it, explaining what it gets right, what it gets hilariously wrong, and why. Python: splitting code into modules and debugging a project they designed.

// toolbox

Tools They'll Use

Python 3

The language of AI — the same one used by professional ML engineers.

VS Code

The same professional code editor used by developers everywhere.

Project Gutenberg

A free library of thousands of books — the training data students feed their engine.

DotCode AI

Our student AI platform — explore and compare the latest models from leading AI providers, and put the concepts learned in class to the test.

// the course

The Course at a Glance

12 lessons
1.5 hours per lesson
13–17 years old, beginners welcome

Try it free first

Book a free trial lesson — your child writes their first text generator and watches it produce nonsense, and you see exactly how we teach before committing to anything.

Book a Free Trial Lesson

// outcomes

By the End of the Course

By the end, your child walks away with an AI they built themselves — and the rare ability to explain exactly how it works.

  • A working text generation engine, written entirely by your child
  • A personal AI project trained on text they chose themselves
  • Solid Python foundations — enough to keep building independently
  • A genuine mental model of how ChatGPT and similar tools work
  • Code published on GitHub as a portfolio piece

Ready to Start Building?

Book a free 45-minute trial class. No commitment, no credit card — just great learning.

Book a Free Trial Class Browse Courses