Why you should say yes
A kid who makes games stops being only a consumer of them. The skills are unusually transferable: game logic is programming's friendliest on-ramp, difficulty tuning is empathy for an audience, and shipping something imperfect then improving it is the single most valuable work habit there is. The research-y way to say it: making games converts screen time from passive to constructive. The parent way to say it: they'll proudly show grandma the thing they built.
Under 10: Scratch, together
MIT's free block-based Scratch remains the right starting point for young kids — dragging logic blocks, instant results, a moderated community. Sit with them for the first projects; the concepts (loops, conditions, events) are the same ones real engines use later. Expect the ceiling to arrive in a year or two: games look Scratch-y, and ambitious ideas outgrow the blocks.
10–14: the idea era (this is where AI builders shine)
This age has enormous imagination and finite patience — the killer combination for an AI game builder. On Arcade Sandbox, your kid describes the game in a sentence and plays it two minutes later; making it better is describing changes, which quietly teaches the real skill of design: noticing what's wrong and articulating it. A few honest parent notes:
- Costs are visible upfront: a new account starts with ⚡ 5 free tokens (a first 2D game costs 4 — so game one is free); after that, token packs start at $9.99 for 10. No subscription, tokens never expire, and there's no in-game loot-box economy — the only thing money buys is making more games.
- Sign-in is a Google account, so it rides on the family account setup you already control.
- Published games are public in the arcade (family-friendly rules are enforced in generation, plus a report/moderation system). If they publish, anyone can play their game — that's usually the thrilling part, but it's your call to make together.
- Do the first one together. Type their idea verbatim — the moment their own imagination becomes playable is genuinely worth witnessing. (Fifty starter ideas here if they freeze at the blank box.)
14+: let ambition pick the tool
Teenagers split into two paths, both healthy: the designer path — keep shipping fast AI-built games, build a portfolio of finished, played things (some teens' games earn real tokens from player unlocks — a first taste of getting paid to make games); and the engineer path — graduate to Godot or GameMaker and learn actual programming, using AI-built games to prototype ideas cheaply first (every tool ranked here). A teen who's done both is, frankly, ahead of many game-design freshmen.
▶ FIRST FAMILY BUILD, FREE
Their idea, playable before dinner. New accounts start with ⚡ 5 free tokens — the first game costs nothing.
Build it together →2D builds 4 tokens · 3D builds 8 · revisions 3 · no subscription, no loot boxes · publishing is free and optional.
Keeping it healthy (the part you actually googled for)
- Making beats consuming — usually. The failure mode isn't hours spent, it's passivity. A kid iterating on their own game is in a different cognitive gear than one doomscrolling gameplay clips. Watch the gear, not just the clock.
- Celebrate shipped, not perfect. The instinct to endlessly restart projects is the thing to coach against; finishing small games is the skill (it's the through-line of every great solo developer's story).
- Let them fail in public, small. A published game that gets three plays teaches market reality gently. Resist fixing it for them; ask "what would make someone play it twice?"
- Their games are a conversation window. Kids put their inner world into what they make. The game where the hero outruns a storm is worth a curious question at dinner.
The one-afternoon test
Before buying courses or camps, run the cheap experiment: one afternoon, one AI-built game from their own idea, shipped and shared with the family. If they immediately want to make it better — congratulations, it's real; invest accordingly. If they shrug, you found out for free. Either way you spent an afternoon making something with your kid, which was never a loss.