4-5:06:AI Training Games – Simulating How Machines Learn

This week the students are the AI. Through training games they experience how a machine learns from examples, makes mistakes and improves – no code required.

Playing the role of the AI – guessing, getting feedback, improving – makes the training process concrete. Learners finish able to explain why an AI needs many examples and why it still makes mistakes.

What your child will do
• See a simple machine-learning game in action and ask how it “learned”
• “You Are the AI!”: predict cat or dog from features like ears, tail and size, using only the training data you’re given
• How It Works: Step-by-Step, plus a pattern challenge – what comes next in the sequence?
• Start the Data Diary: which data helped you predict well, and which was confusing?

By the end, your child can
• The train → find patterns → predict cycle
• Why AI makes mistakes and how it improves
• How to keep a data diary

Learning focus (Fink’s Taxonomy of Significant Learning): Application · Learning How to Learn.

Active, game-based lesson; works for individuals or groups. Part of Grades 4–5: Mission: AI Explorers, Module 2 – How Does AI Learn? – Feeding the Machine Mind, Week 2. Next lesson: Good Data, Good Decisions – Fairness and Accuracy.