Don't post generated/AI-edited comments. HN is for conversation between humans

Exactly when was this point added? It seems somehow not new, but on the other hand it was missing from an archive.today snapshot I found from last July. (I cannot get archive.org to give me anything useful here.)

Edit:

> Please don't complain that a submission is inappropriate. If a story is spam or off-topic, flag it.

> If you flag, please don't also comment that you did.

Perhaps these points (and the thing about trivial annoyances, etc.) should be rolled up into a general "please don't post meta commentary outside of explicit site meta discussion"?

Consider a much more cynical view where people are strictly self-interested and use these tools to garner engagement and self-promotion. Good chance the meaning did not originate from the person. And now these people have tools to outsource their parasitic intentions.

An LLM summarizing the contents of a blog post might be useful to you, but is a comment here the right place for something you could geneate on your own?

I would guess for most people here, real insight or opinions from others is the "useful" aspect of reading hackernews comments.

Using LLMs to generate or refine comments only moves things further away from that goal (in my opinion).

By all means make good use of LLMs and other AI. What counts as good use? The world is figuring that out, it will take years, and HN is no exception (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...). We just don't want it to interfere with the human conversation and connection that this site has always been for.

For example, it has always been a bad idea and against HN's rules when users post things that they didn't write themselves, or do bulk copy-pasting into the threads, or write bots to post things.

As I mentioned, the HN mods (who are also the HN devs) use AI extensively and will be doing so a lot more. The limits on that are not technical; they have to do with (1) how much work we still do manually—the classic "no time to do things that would make the things that take all our time take less of it"; and (2) the amount of psychic rewiring that's required—there's a limit to the RoA (rate of astonishment) that any human can absorb. (It's fascinating how technical people are suffering the most from that this time. Less technical people have longer experience being hit by disorienting changes, so for them the current moment is somewhat less skull-cracking.)

Getting this right doesn't mean replacing human-to-human interaction, it means we should have more time for that, and do a better job of supporting HN users generally, as well as YC founders who want to launch on HN, and so on. The goal is to enhance human relatedness, not diminish it.