Re:Ask HN: Should "I asked $AI, and it said" replies be forb...
by fwip on Tuesday December 09, 2025 @01:55PM
(#46208966)
I'd love to see that for article submissions, as well.
Re:Ask HN: Should "I asked $AI, and it said" replies be forb...
by manmal on Tuesday December 09, 2025 @01:58PM
(#46209016)
What LLM generate is an amalgamation of human content they have been trained on. I get that you want what actual humans think, but that’s also basically a weighted amalgamation. Real, actual insight, is incredibly rare and I doubt you see much of it on HN (sorry guys; I’ll live with the downvotes).
Re:Ask HN: Should "I asked $AI, and it said" replies be forb...
by dinkleberg on Tuesday December 09, 2025 @03:14PM
(#46210021)
Why do you suppose we come to HN if not for actual insight? There are other sites much better for getting an endless stream of weighted amalgamations of human content.
Re:Ask HN: Should "I asked $AI, and it said" replies be forb...
by ergonaught on Tuesday December 09, 2025 @03:38PM
(#46210320)
Coming here for insight does not in any way demonstrate that genuine insight is actually widely available here.
It's not worth polluting human-only spaces, particularly top tier ones like HN, with generated content--even when it's accurate.
Luckily I've not found a lot of that here. That which I do has usually been downvoted plenty.
Maybe we could have a new flag option, which became visible to everyone with enough "AI" votes so you could skip reading it.