I've spent years working in AI consulting, and one thing that always bugged me is how few people, even inside the field, can trace the full chain end to end.
Where does the money actually go? Why is compute the bottleneck? How does a pile of data turn into a model, and a model into revenue?
So I built a game about it. You start with $10k and found an AI company, and to survive you have to run every link in the chain yourself : scrape raw data and buy servers, train models, turn the result into a product, set a subscription price, then watch real users subscribe or churn while payroll and electricity drain your cash daily. The revenue funds more compute and research, and round you go.
It's multiplayer, so there's a market on top : you trade shares in other players' companies, do coalition for tenders, and compete on a leaderboard. Persistent and real-time, one real day is one game day.
Free, runs in the browser, English and French. I'm the dev.
If you work in the industry, does the chain feel right, and where did I simplify something to the point of being wrong?
[link] [comments]