AI Creators Ages 10–12

AI Game Designers

Gervigreindarleikjahönnuðir

The game is the lesson. The AI is the opponent.

Book a free trial

What they'll learn

  • AI that plays your game Train a model to control or compete in a game the student designs.
  • How game AI decides See the difference between hand-written rules and AI that learns its behaviour.
  • Designing fair fun Build challenges that are fair and genuinely fun to play against an AI.
  • Test and tune the opponent Playtest, measure and adjust the AI until the difficulty feels just right.
Final project

A playable game with an AI opponent or controller they trained, demoed at arcade night.

The 13-week journey

  1. Sprint 1 — Game AI basics

    How classic game enemies 'think' (rules) versus learned behaviour; building a simple game and a rule-based opponent. Ends with a playable prototype.

  2. Sprint 2 — Train the player

    Training a model to control the game (gestures, voice or moves) or to act as an opponent; accuracy tuning as difficulty tuning. Ends with an AI-controlled build.

  3. Sprint 3 — Fair and fun

    Designing challenge that's fair, not frustrating; testing the AI against real players; the 'when does AI make a game better?' discussion. Polishing for showcase.

  4. Week 13 — Showcase

    Arcade night: families play against the trained AIs and vote on the most fun.

What we cover

Every topic, unit by unit — so you know exactly what your child builds and learns.

01

Game AI basics

  • How classic enemies 'think' (rules)
  • Rules versus learned behaviour
  • Building a simple game
  • A rule-based opponent prototype
02

Train the player

  • Training a model to control the game
  • Gesture, voice or move-based control
  • An AI opponent that reacts
  • Accuracy tuning as difficulty tuning
03

Fair and fun

  • Designing fair, non-frustrating challenge
  • Testing the AI against real players
  • When does AI make a game better?
  • Polishing for showcase
04

Arcade night showcase

  • Final tuning and instructions
  • Presenting how the AI was trained
  • Families play the trained AIs and vote
What they show off

A playable game where the audience competes against (or plays with) a model the child trained.

For the kids who live in games, the most motivating way into real AI: design a game where the AI is the opponent or the controller — and train that AI yourself. The game is the hook; training, tuning and understanding model behaviour is the curriculum.

The hooks

Kid hook: “I trained the AI that you have to beat in my game.” Parent hook: “The hours they’d spend in games, turned into training and understanding real AI.”

Who it’s for

Gamers ready to switch sides of the screen; no coding required. Thrives: competitive, creative kids who want their AI to do something they can feel.

Outcomes — by the end, students can

Train a model to control or oppose a game; explain rules-based vs learned game behaviour; design fair difficulty; test and tune an AI opponent; judge when AI genuinely improves a game.

Tools & compliance

Block-based game tools plus account-free model training (Teachable Machine / Machine Learning for Kids); school laptops; no personal accounts under 13.

Where this course fits

Pairs with AI Trainers; a fun route toward AI Makers at 13–15.

Parent questions

Isn't this just a game-making class?

No — the game is the vehicle; the real learning is training and tuning an AI and understanding how game AI decides.

Do they need to code?

No prior coding needed — we use block-based tools and trained models; AI Trainers helps but isn't required.

Any chatbots?

No — under-13s train their own account-free models; no chatbot accounts.

The first lesson is a free trial.

Book a no-commitment trial — pay nothing if it's not a fit.

Book a free trial