Comment Re: I've tried (Score 1) 73
Which makes them an almost perfect simulation of human intelligence.
Humans aren't infallible, but even pretty stupid people clearly do things when they think that "AI" currently can't. There is at minimum some kind of filtering and going back to the well happening that the LLMs can't manage. I am not ruling out them becoming capable of it in the future, but they are clearly not there now.