hey all, i shutdown starlight labs (@starlightlabsco), the game studio I made to create new games at the frontier of what's possible with current AI models. i open sourced all the work for it, although fair warning i didn't clean it up at all: github.com/orgs/StarlightLab… biggest takeaway is that in the game industry the *only* thing that matters is: is what you make fun? is it fun on the micro, moment to moment gameplay? is it fun on the macro, overarching learning curve, that keeps you coming back for game session after game session? i found that my game prototypes could never hit that sweet spot of being fun in the moment but also last more than a few minutes. they were cool demos to try rather than lasting games. the reality is that most games do not succeed in doing that!! you can have teams of AAA experienced devs make a game and have it flop, and a single dev make a simple game that is the next megahit. finding that gameloop is the jackpot, and it's really freaking hard. so if you're interested in doing something similar, focus on the fun. in regards to what's next for me, I've already been working on something that makes me so unbelievably excited about that future (for quite a while now). it's much much bigger scale and scope, with a kickass team going after a really hard problem, and also, it lets me get my hands on hardware again, but more on that later :)