AI Ground School: play a game, actually learn to use AI.
A browser game that teaches you to actually use AI, not just talk about it. Ranks run Recruit to Architect — first two tiers are free, no signup, no install. Paid tiers with more advanced lessons.
Teaching yourself how to use AI effectively.
6 ratings
AI video with native sound, paid per clip. Nothing stored.
A calmer Mac menu bar, one click away.
Enterprise Software. Intelligent Automation. Global Impact.
Comments
3Hey — I'm the one who built this. Not a developer by trade, just spent months teaching myself how to actually get useful work out of AI tools, mostly through trial and error and more closed tabs than I'd like to admit. The idea came from noticing how many people — me included, for a long stretch — "use" AI constantly without ever getting better at it. Same shallow prompts, same okay answer, no real progress from one session to the next. Every explainer I found was either a 40-minute video or a tips list I'd forget by the next day. I wanted something you actually do, not something you read. So AI Ground School is a rank-based game: you play through Recruit, Cadet, Operator, Specialist, Builder, Architect, and each rank teaches one specific, hands-on skill. Every lesson is also labeled for whether it works in any AI assistant or is specific to Claude's own tools — I didn't want anyone buying it expecting one thing and getting another. First two ranks are free to try right now, no signup. Genuinely curious where it breaks or where you get bored — that's more useful to me than a compliment.
Turning AI practice into a game with visible progression is a smart way to encourage hands-on learning; the no-signup entry point should make it easy for curious beginners to try.
Solo builder making small, focused web tools. Currently working on Spacebar CPS Test — a free spacebar clicking speed test with a live leaderboard.
The strictly sequential unlocking is a smart call. Most "learn AI" content lets you skip around, which usually means people bounce between surface-level concepts without ever building real fluency. Forcing each lesson to unlock only after the last one is done feels closer to how skill actually gets built. The tagging system (universal vs transferable vs Claude-specific) is also a nice touch — a lot of AI tutorials quietly assume everything you learn transfers 1:1 to other tools, which isn't really true. Being upfront about that seems like it'd build more trust with people who use multiple assistants.