I made Vibe Inc., a free browser game built with Three.js, React, and a lot of Cursor/AI assistance over about a month for Vibe Jam.
Vibe Inc. is basically my attempt to turn what AI founders are already doing in real life into a tycoon game.
You pick a questionable product idea, throw AI agents at it, call the result “engineering,” and try to survive long enough to ship. The game treats every cursed app, dashboard, landing page, tiny game, or chat product like a serious launch.
Each project goes through the full startup ritual:
idea → build → chaos → launch → showcase → reviews → cash/reputation.
You can:
- run a 3D AI studio
- hire and assign AI agents
- ship weird internet products
- manage launch events, economy, and reputation
- open the products you shipped as 20 different showcase artifacts
- grow the studio after each launch
You never write a line of code. Your agents barely do either. Somehow, it still ships.
That was the inspiration: the vibe-coding/startup scene already feels like a tycoon game, so I made it one.
The game is fully 3D and runs in the browser. Performance is pretty solid overall; some first loads can have a short hitch, but once loaded the main flow is smooth.
I also didn’t just let AI randomly decide the game rules. I used AI heavily for iteration and balancing, but the project logic, review outcomes, and event effects were shaped around real vibe-coding patterns and what different product types actually need to succeed.
I ran a lot of longer balance tests/simulations to make the economy and launch scores feel fair, but it’s still a 1-month build, so I’d love to hear how the balance feels.