Post #2330845

2026-05-09 12:32 UTC

I think at this point the only viable ai policy for open source projects is: &quot;If we can tell you&#39;re using ai we ban you immediately.&quot; Disclosure doesn&#39;t do much, people just lie about what they used ai for. And I&#39;m so fucking tired of talking to Claude via PR comment. If we can&#39;t stop people from using it, and it &quot;democratizes&quot; software development then let&#39;s just get rid of everyone who can&#39;t even be arsed to use it right. I&#39;m so tired.

Replies (2)

@hp@mastodon.tmm.cx 2026-05-09 12:39

If I read one more hopeful first contributor who will accept our verdict on the pr, or who has an "honest" suggestion or who wants to "gently push back" I'm gonna scream. Not one of them willing to even just admit they used Claude to write the comments. Why lie about it, it's not like anyone has any doubt about it.

Open ##2500447
@hp@mastodon.tmm.cx 2026-05-09 17:53

The point is to be left with contributions who don't use ai, and those who use it in such a way that doesn't lead to slop. Arguing with people who are OBVIOUSLY using ai to generate all communication they send to the GitHub is a huge waste of time. I don't want to operate someone else's Claude code session and have them get a commit credit. That is what reviewing an AI generated PR is. It's pointless. AIs don't learn. It's ALWAYS faster to do it yourself. #aislop

Open ##2500448