Heh, getting down voted for pointing out that HN detects 200 out of 10 AI written articles.
I have AI blind hate fatigue.
Edit:
Heck, a year or two ago, someone accused me of being an LLM here on HN, sir, English is just my second language, I sometimes fumble it organically, thank you very much, I also have a touch of ADD.
I don't use LLMs for my comments, I write the dumbest things I think of and eat the downvotes, I make it a point that if I'm posting comments or writing a blog post, an LLM is only going to tell me suggestions on what to tweak, but for the most part I post raw if its just a comment on a forum.
AI is great at saying nothing specific really flamboyantly or juicing up very little substance with a whole bunch of that sort of puffery that's endemic to corporate communication. It is very bad at the subtle pairing of tone and diction with subject matter and audience that humans just do naturally.
That's the kind of language style LLMs picked on (recent attempts to make their writing more common-people, particularly Claude, notwithstanding), and the jarring/sloppy part is mostly just that you don't expect to see this kind of language in a bug report or commit description or design draft. If you got such response from a person in such context, you'd accuse them of showing off and making communication difficult. But it's not bad English - it's too good English for context.