A 60-minute lesson on how machine learning really works, and why AI can be unfair. Pupils become the AI in an unplugged activity, then find out how missing or wrong examples lead to mistakes and bias.

Lesson outline

• Do now: what is training data?
• How AI learns: collect, label, find patterns, predict, test
• Be the AI: what is a flurb? Pupils learn a rule from examples, then hit a case the examples never covered
• What is bias? From green apples to face recognition
• Making AI fairer: variety, correct labels, testing on everyone
• 17 minutes on the interactive drill
• Exit ticket: who is responsible for checking AI is fair?

Included: slides.html and slides.pdf · lesson-plan.pdf with teacher notes and sensitive-topic guidance · drill.html (10 questions with an explanation for every wrong answer) · worksheet.pdf with answers · how-to-use.pdf

Curriculum links: KS2 Computing (logical reasoning about how algorithms work), KS3 Computing, and Education for a Connected World. Works offline. No logins.