Our approach
The build is not a reward. It is evidence that learning happened.
Each session starts with a real question and ends with something that can run, be tested, and be explained. We assess the decisions, not only whether the final build worked.
01
The session loop
Predict, build, test, explain, then record what should change next.
- Frame the question
- Students state a judgement and a reason before following any procedure.
- Build and debug
- Errors are kept long enough to diagnose rather than being fixed by the teacher.
- Explain and transfer
- Students explain the mechanism and decide whether it still holds in a new context.
02
Responsible AI throughout
Bias, hallucination, misaligned goals, and human responsibility are checked whenever AI is used, not reserved for a final ethics lesson.
03
Different starting points
Beginners receive a complete entry path while experienced students use extension challenges. Both work on the same core idea without being labelled by speed.