https://news.ycombinator.com/item?id=48887942
> They can't know for sure whether what they're saying is true or false, and we can't know for sure how we should moderate it.
(Regarding people identifying text as genai).
You could moderate like another widespread but hard to prove problem: Propaganda and dis/misinformation campaigns. It's easy to see lots of people repeating well-crafted talking points (too well-crafted for an ordinary person), shouting down those who disagree. Yet users are forbidden from identifying it as such in comments; we're told to email hn@ycombinator.com. You could do that for genai suspicions too (assuming you have unlimited time to review such things).
It seems like a witch hunt to me; I have little reason or evidence to believe people know - because people speaking with no evidence is never reliable, because the angry mob is prone to witch hunts, because genai's are trained on human writing making it harder to distinguish, because some factors people target (e.g., em dashes) have been widely used, and because a highly disruptive technology at this stage of adoption is highly prone to strong emotional reaction and misunderstanding.
I'm not sure it matters: As a rule, we should hold the human who puts their name on it fully responsible. If their assistant, their genai software, or their dog writes it, it doesn't matter. Their name is on it. Speakers don't blame their speechwriters - the speaker said it.
I'm not sure it matters because right now it seems like a big reactionary over-response. I doubt we'll care much in a few years.
The automated spam potential, in posts and in comments, is a problem. One solution is somehow raising the standard of posts and comments: if it's that easy to generate middling crap, it makes better content more valuable and available and it may raise the standard naturally.
Thanks!