[acdha]

This place hates laziness and imprecision. Using ChatGPT for editing or inspiration is okay as long as you personally review the results for accuracy and completeness, at which point people care about it as much as you announcing that you used a spell checker.

[delaminator]

Pasting chat GPT responses is against the site rules.

always has been even before GPT

https://news.ycombinator.com/item?id=46206457

[acdha]

Fair, because they’re not your words. I’ll edit my comment for what I had in mind: that it can be helpful for that like a spell checker - for example, I know non-native English speakers who find them useful for editing but they completely understand the topic intellectually.

[quietbritishjim]

True, but the original comment that we're talking about here (by sundarurfriend) just mentioned an LLM's output in passing as part of their (presumably) human-written comment. Nothing you've linked to prohibits that.

[delaminator]

Presaging your bot produced comment with "A bot said this" is not human written

[quietbritishjim]

That is still true and still irrelevant here. The comment we're talking about was not written by a bot with a disclaimer at the start. They just asked about its output. They didn't even quote its output - they paraphrased it and added their own commentary!

I know HN rules prohibit saying "did you even read it?" but you surely can't have read the comment to have come to this view, or at least significantly misread it. Have another look.

Most of all, HN guidelines are about encouraging thoughtful discussion. sundarurfriend's comment asked a genuinely interesting question and inspired interesting discussion. This subthread of "but AI!" did not.

[acdha]

Except in that case they were summarizing it, which I read as closer to “I found this on Stack Overflow but don’t know if it’s right”. I think that’s less offensive than having the post be LLM output or, especially, pretending to be authoritative.