Augmented Metacognition using LLMs
21 October 2024
I think this is something that is hella slept on.
I find the primary purpose of LLMs is as logic engines. I use them to help me think through problems, and to debug my own thinking process itself.
I find that half the battle is formulating what your problem even is in the first place; identifying this to me is analogous to switching on a light in a dark room, so that you are no longer fumbling around in the dark.