Learn to AI-code — starting with games

Why games? Why now?

Working with AI is becoming the way most software gets made — and it is a learnable skill, not a party trick. The loop is always the same: cover your ideas (say everything that matters — the AI builds what you said, not what you meant), test it against reality (play it: if the jump feels wrong, it is wrong), iterate (describe the gap precisely), and persevere (the first build is never the good one). Games are the perfect training ground because reality answers back instantly — a boring game tells you your instructions were incomplete.

Along the way, learners pick up genuine software knowledge — code, architecture, debugging, how languages differ — on the job, because their own game needs it. The same loop later builds websites, tools, school projects and anything else they can describe.

What we cover

• Prompting fundamentals and how to stay safe.
• Writing a real game brief: verbs, decisions, difficulty, unlockables.
• Building a browser game with AI assistance — then playtesting it honestly.
• Iterating on visuals, controls and difficulty until it is genuinely fun.
• Reading the code you made: what changed, why it works, how it fits together.
• Publishing a playable game on this site, with your name on it.