it's huge problem in the foss world, endless slop projects that get abandoned immediately when the meat proxy gets bored and moves on to the next idea that solves no real problems, endless slop pull requests for real maintainers to waste time sifting through, etc etc etc. it's not something to just ignore because the fatigue of constantly seeing it is real
yes, there's plenty of legitimately useful cases even if you put aside how you feel about the World's Largest Theft Of Everything plagiarism machine (you cant use ai for projects like Wine since there's no guarantee 0 microsoft code was trained on, etc), but most people are not linus torvalds carefully debugging the linux kernel with claude's help, it's people who want to smugly feel like they've skipped the learning process straight to the internet glaze because they got claude to shit out software that worked long enough to farm twitter likes on a doomfist in elden ring video
something that has sort of worked in other communities has been to blanket ban project posts with less than x months of github activity to at least filter for projects that demonstrate the author has enough dedication to maintain something
to take this one step further, this conversation dovetails immediately into how ai use effects online communities and even ai use in general.
i believe the following are worthy reads:
https://deadsimpletech.com/blog/no-such-thing-as-just-a-tool
https://blog.fallible.net/it-breaks-a-village/