News for hackers. Stuff that matters.
Re:LLMs can't self-correct in reasoning tasks, DeepMind stud... by ben_w on Monday October 09, 2023 @03:52PM (#37824656)
It may be "plainly and obviously" different for you, but for me? I seem to have two components to my thinking process, one of which has ideas without needing words, another of which turns those ideas into words and gets (for lack of a better description) "annoyed" if I try to skip the word-generation part.

There's no reason to presume we all think the same way; rather the opposite, given how many ways in which humans are already known to think unalike to each other.

There are no comments on this comment yet.