Comment by aiven

But LLMs still can't reason... in a reasonable sense. No matter how you look at it, it is still a statistical model that guesses next word, it doesn't think/reason per se.