Really? I think HN has even fewer safeguards against AI-generated content than Google. No offense to dang and co., and I'm sure there's more going on there than I'm aware of. But still, I'm pretty sure it would be trivial to set up an account and use GPT-3 to produce the content. The only reason I'd suspect this isn't happening is because there isn't a strong financial incentive to do so. In other words, HN avoids the spam because it's still too small to matter.
While it is true that there are no good incentives for a bot to comment here, I believe that the best safeguard against AI-generated content is that here such content will be noticed easily.
I have seen a couple of times on some other even smaller technical forums what appeared to be sequences of comments generated by an AI, from a fake account (until the accounts were banned).
There could not have been any kind of financial gain from those actions, so the conclusion was that someone was testing the capability of their AI to carry a conversation on such subjects.
The messages posted were remarkably coherent, but in the end it was obvious that their source could have been only either an AI or a human with mental problems.
What made those messages stand-out was that even if they contained mostly correct information, the same that might be found e.g in technical magazines, there was always some weird disconnect between their answer and the message to which they replied.
Apparently, in most cases the AI failed to identify which exactly was the point-of-view of the previous poster. The reply referred to various items mentioned in the parent message, but instead of referring to the essential points it referred to some non-relevant things mentioned more or less accidentally, or it tried to argue for or against some points as if trying to contradict the previous poster, when they actually had argued in the same direction.
They will be downvoted into Oblidon. And then banned, shadow-banned, hell-banned and a few more creative ways of banning. The account, the IP, the site, and perhaps anyone that a bayesian filter put in the same bucket.
Yes, which will produce things that are stylistically similar to HN comments, but without any connection to external reality beyond the training data and prompt.
That might provide believable comments, but not things likely to be treated as high-quality ones, and not virtual posters that respond well to things like moderation warnings from dang.
I think it's possible to train the bot to write oneliners, but onliners are usually downvoted here.
Replies to comments are difficult, so I guess the bot must just write somewhat related top level comments.
It will accumulate a lot of "Did you RTFA?" and also polite request for clarifications, and it will be suspicious that the bot never replies or replies with nonsense.
If every commit has a link to http://www.example.com then people will notice and start flagging.
I guess it can fly under the radar posting a few top level bland comments per week in the megathreads with a few hundred comments, but I don't imagine how this can be used to gain anything.
But HN also adores tangents, and that gives plenty of room for GPT-3 to succeed. You could affect topical discussion merely by distracting it in several directions.
