Real games, not exercises
A racing game, a pirate trading adventure, a dragon-raising sim, a Quidditch match — an actual playable project, not a practice drill.
AI-Powered Coding
Build a real game. Test it yourself. Make it better.
Squirrel Wisdom's coding course pairs each student with an AI programming partner that does the technical work — writes the code, runs it, fixes what breaks. The student's job is to describe what they want, play what got built, and decide what's next.
- One real game, built session by session
- The AI types; the student decides
- Nothing's "done" until it's been played
Not a tutorial, not a template
Real projects, one small working piece at a time.
No worksheet, no fill-in-the-blank starter file. A student describes what should happen next, checks the result themselves, and only then moves on.
The AI types, the student decides
The student describes what should happen and how they'll know it worked; the AI partner writes and runs the code. Direction and judgment stay the student's job, not syntax.
Every step ends in play
Nothing counts as finished until the student actually plays the result and checks it against what they expected — not because the AI said it was done.
Real debugging, real skill
When something's wrong, students learn to describe exactly what they saw instead of guessing at causes — the same habit a professional developer relies on.
A course, not a single lesson
The same game, built a little more each time.
One project runs across many sessions — a racing game that starts as a single car on screen and grows, session by session, into something with a track, opponents and a finish line. The same cycle carries every project in the course, from racing to pirate trading, dragon-raising, Quidditch, animal chess, solar-system exploration, even Minecraft mods.
Every session starts the same way: run what already exists and confirm nothing broke, before a single line of new work begins.
The course ends with a demo day. The student explains the game, the decisions they made, and how playtesting changed it — not just "here's my game," but "here's what I built and why."
What a student actually learns
Skills the AI can't do for them.
The AI handles the code. Everything below is the student's own job — and the actual point of the course.
Say exactly what you want
An unclear instruction gets an unexpected result, every time — that's the lesson, not a failure, and it's the habit the whole course is built around.
Decide how to check it
Before asking for a change, agree on the exact actions that would prove it worked — not a vague "does it work?"
Play it, don't just watch
The real test is a student's own hands on the keyboard, not someone else's word that it works.
Recover from a stuck spot
When a fix doesn't land, narrowing down the problem and rolling back to a known-good state — a real engineering habit, not a shortcut.
Curious what your child could build?
Every course runs on this same cycle — a real project, grown one working piece at a time.