AI slop is ripping up the social contract between maintainers and contributors essential to open source development. Practitioners have been repeatedly assured that AI would supercharge their communities, but so far that hasn’t been the case. Just look at what happened last month. Mitchell Hashimoto’s Ghostty implemented a zero-tolerance policy where submitting bad AI-generated code
Really my biggest frustration is that it isn’t transparent what’s generated and what’s not. If a human wrote the code, then I want to teach them, especially if there’s glaring logic issues in the code.
But at the same time, the most likely cause for glaring logic issues, is if they generated the code. And then it’s just a complete fucking waste of my time to try to teach them.