A lot of replies to this post (and to fears about AI in general) are of the form "But humans are as untrustworthy, if not more. So the AI can't be any worse". I

A lot of replies to this post (and to fears about AI in general) are of the form "But humans are as untrustworthy, if not more. So the AI can't be any worse". I have to say many of you have a very dim view of humans. When I read an answer off of stackoverflow that is not highly voted, I know that the answer could be incorrect. But there is a trust that the actor on the other side was also facing the same problem, and is not malicious. After all, why would a malicious actor care to give a wrong answer to a "Rust borrow checker question"? If the answer was AI generated and posted by a bot operated by a karma farmer, out goes that trust. This is obviously not a threat to society, and chances are the asker will eventually figure out the right answer, but you can see how this sort of thing reduces the signal-to-noise ratio of almost anything on the internet by orders of magnitude. Trust underpins everything. Hacker news IS a social network in my opinion. I come here to read comments knowing that there is a like-minded person on the other side who is sharing their thoughts. If I was looking for information or something to learn, I would read a newspaper, a textbook or a research paper. The discussion with a real human from some corner of the world is what makes this forum tick. This tech is not going away and I presume some steady state will be reached eventually, but in the meantime, there is no doubt in my mind that the internet as we know it is in jeopardy.