I Built a Dice Wars Mobile Game in Three Days With AI

The game in the video is mine — a mobile version of Dice Wars, from zero to playable in about three days.

This is the thing I keep talking about: use AI to produce a lot of stuff in a domain you actually understand.

I know games. I’ve played enough of them to recognize, on sight, what kind of mechanic is going to hold up. Dice Wars — simple rules, real strategic depth, short matches — is exactly the kind of thing I can look at and immediately say “yes, that’s worth building.” Everything else — writing the code, tuning the balance, generating assets — AI compresses from weeks into days for me.

So what does “three days per game” actually mean?

It means I can ship ten in a month. A hundred in a year. Most of them will flop. But if one or two break out, the whole ledger balances. This isn’t betting on “which one will hit.” It’s trading output volume for probability. I wrote about exactly this idea a few days ago in Indie Hacking Is Just a Math Problem.

The old fear of an indie dev was “I worked on this for three months and nobody played it.” The math has changed. Now it’s: worked on it for three days, nobody played it, next.

I’m going to keep doing this — build a game, ship it, write a post about it. This is the first one.