Post

Folks seem upset about the genAI involvement almost more than the quality of the project being bad, but don't most of the good recomps/FPGA cores of retro games involve genAI to some extent?

I never really looked into it, but from what I understand, you take a rom, decompile it into assembly code, convert the assembly code into something higher level + human readable, then extend it as needed for various QoL features, using the original rom for all non-code assets.

Used correctly, I feel like genAI would do pretty well for at least some of those steps. And probably has, but "it's AI slop" seems to get overshadowed by "the recomp is peak" when it's good...

If the rise of platforms like Netflix has taught us anything, it’s that people appreciate curated hubs that simplify choice. A launcher featuring only the best recompilation projects could thrive and let the mediocre ones quietly fade into oblivion.

My tin foil hat theory is that some of its intentional, sow distrust in a community that's been making HUGE strides and outdoing their originals, because now there's an added layer of time trying to figure out if a decomp is actually worth anyones time or not.

Why people keep doing it? Clout chasing. When you're the first to do it you get the most recognition, even if someone comes later and does it better you were still first and deeply ingrained in the culture around it. They don't care if it's good just that they did it and now it's a thing

the community is niche, but so extremely dense and supportive. i'm sure this type looks in from the outside, sees how genuine and compassionate folks are, and wishes they could contribute.. ofc blissfully unaware of the fact that 80% of the people involved are learning as they go

i only hope that some petty ass motherfuckers spend the time to do this stuff simply out of spite for AI, even with the diminished returns. decompilation, to me, is an incredible art, so it only makes sense that AI dipshits would come for it too