Children are surrounded by AI and told almost nothing about how it works. This course fixes that by having them build the pieces themselves: matching patterns, reading features, classifying, scoring confidence and combining votes.
Nothing here is a black box and nothing is hype. By the end a child has built a small chatbot and a classifier by hand, and can explain why an AI that was trained on lopsided data gives lopsided answers - which is a more useful thing to understand than any single tool.
Start this course freeFull lessons from this course are readable here, with no account - the explanation, a worked example and the mistakes children usually make:
Each lesson opens with the child’s own coding buddy explaining the idea. The explanation is spoken aloud, typed on screen and performed by an animation at the same time - the same sentence in all three, so a child who reads slowly can listen, and a child who is not reading at all can watch.
After the explanation come the practice games: spot the bug, predict the output, order the code, and a written challenge that runs the child’s real code and checks it.
The core ideas, yes. Classification, pattern matching and weighted voting are all things a child can do by hand with cards and tally marks, and that is exactly how this course introduces them before any code appears.
It teaches the ideas underneath tools like that, including a lesson on writing clear prompts. The focus is on understanding the machinery rather than on driving any one product.