I dunno. One of my pet peeves of the last few months is that somebody posts an AI generated blog post and it's not good enough that we can click on it to read it so they also post an AI summary as a comment. Like... we can click on the link and read your post. Haven't you ever used HN before... Oh, I see, you created this account six hours ago.

I think they might have tuned up the heuristics to suppress that kind of thing but that's not what is going on there... like these are real replies in a real conversation. I don't like it that I can't upvote or vouch for them.

I take it is due to automation, mostly. Every single website that has some sort of user-contributed content, forums, source code repository, social media, anything that has comment or upload capability of any kind whatsoever is being bombarded by a massive torrent of crap, and hn is no exception. The internet has no good defense against it, therefore people have to resort to captchas, geoblocking, shadow banning, proof-of-work portal solutions, cloudflare (eek!) and adjusting detection thresholds "just right" so things feel usable.

Hat's off to the mod staff, things feel "alright" even with the current situation.

Self promotion posts are one thing, not sure if it applies to the examples you gave, good luck if your hn user name and your github user name is same when you have low karma.

HN’s mods have said they have huge amounts of code fighting spam alongside human workers, and they deliberately don’t discuss their tactics in order to avoid an arms race between themselves and the spammers

Everytime you make a comment about how Grok is led by a pro-fascist team, you get downvoted and your comment gets locked