I kinda agree. AI is indeed a reductive technology that very rarely (if ever) returns factual content. But that's tangential to the real problem, in my opinion;

I kinda agree. AI is indeed a reductive technology that very rarely (if ever) returns factual content. But that's tangential to the real problem, in my opinion; how much we trust people who are entirely capable of disseminating lies. Researchers and government employees know that they're doing the wrong thing when they delegate important work to language models. They are liable to be audited, questioned under-oath and punished if they deliberately publish wrong information. AI doesn't really change that, besides enabling them to fabricate lies faster (which was never something people struggled with in the first place). > It will get to the point where when asked to show proof of something, you'll realize there is no way to actually find it. This I think is just blatantly wrong. For all the hubbub around AI, I've seen pretty much nothing AI-generated persist past maybe 2 days after it was uploaded. AI-generated news articles and photos simply aren't held in high-regard, relative to the prevalence of factual and cross-referenced knowledge. People remember Pixar films, nobody remembers that stupid Pope Balenciaga trend from Twitter a few months back. Nobody likely ever will.