Comment by octoberfranklin

2 months ago

Humans can generalize.

LLMs just predict the statistically-most-likely token.

human brains are just chemical reactions and electrical transmission between neurons too. You're comparing completely different layers of abstraction in your arguments.

  • You're comparing pie-in-the-sky scifi AIs with humans.

    LLMs are what we have today. They can't generalize. Humans can.

    That might change someday, but it certainly hasn't yet.