Post #4057863

@mjdxp@labyrinth.zone ruined feels like an exaggeration considering that my experience using distros that have been getting a lot of flak for AI use (Fedora) has been extremely smooth and largely issue free recently. Systemd also keeps on just working. And the desktop experience is getting better, which is the opposite of what I've experienced with commercial software that keeps degrading nowadays. The one time I'd agree would be rsync perhaps, of course the slop was the culmination of an overworked maintainer with no funding that saw a way to potentially lessen the load and took it. The software became untrustworthy, but I wouldn't give the maintainer flak for it due to the circumstances. A part of the solution to these issues will be recognizing why people use AI and hitting the problem at the source. And those issues go deep into the politics of open source and society at large. There's a significant practical difference between software where the maintainers use AI to review PRs or look for bugs and software where the code is slop that noone understands. They're still some level of morally in the wrong, considering the AI use contributes to the bad externalities, but the software they're developing is affected much less negatively than in the case of vibe coded software.