[tivert]

> Ofc it's hard to prove any of this. But it almost feels like there's an algorithm with a bias at play

It doesn't pass the smell test that there would be an algorithm involved in actions like that. The development timeline and effort involved makes absolutely no sense vs. alternative explanations like manual action. Sometimes even things you'd expect to be algorithmic are not and are really AI (Actually, Indians).

I think the typical explanation for stuff like that on HN is user flagging, which does pass the smell test as an explanation for what you describe. Altman and OpenAI still seem to have a fanclub on HN (real and/or astroturfed), though it's perhaps less dominant than half a year ago.