People can tell. The premise is false. It’s sometimes hard to tell, obviously it’s hard to ascertain false negatives and false positives, but it’s usually pretty obvious.
I genuinely don’t understand AI people anymore. Like the cognitive gap is so huge that I feel like I’m from another planet now. Im not religious, but automating religion is so absolutely meaningless that it boggles my mind. You could have a machine emit million of prayers up to heaven per second, but why would you?
And despite what you think, most of us can tell apart AI generated content from the genuine thing. I am, however, starting to believe AI bros are being sincere when they tell us that they can’t. Every time someone gives me that tired “well how do you know we’re not just stochastic parrots too!” crap, I’m getting a little closer to taking their word for it. Maybe they are just that.
I used to worry that the problem was that LLMs allowed you to be stupid, but I recently realized the actual problem is that they reward you for being stupid.
