My sincere hope is that LLMs break forever the common knowledge that computers are infallible. We've seen bad data since the dawn of vacuum tubes but because machines have no skin in the game they're held up as egalitarian and fair. ChatGPT is obviously straight-up bullshitting, and because the UI is so normie-parsable everyone can see it plainly. I think it's interesting that every culture that has spirits assumes they are wily and full of tricks. Jinn, Leprechauns, Tengu - if it isn't of flesh and blood it's trying to pull a fast one on you. Yet for the longest time our reaction to technology has been the opposite - it's a tool, tools don't lie. ChatGPT lies like a rug. I think/hope the result will be a revaluation of the unimpeachable algorithm.