‹ BackHN Continuity

Thread

The LLMentalist Effect (2023)

235 points · 309 comments · jalev

  1. rahidz · · focus · HN ↗
    Man I remember back when a psychic conned me by solving the Navier-Stokes problem.

    Also >July 4th, 2023

    1. Angostura · · focus · HN ↗
      But were you successfully conned into believing an AI had solved the Navier-Stokes problem?
      1. tescreal · · focus · HN ↗
        What is the consensus among experts?
        1. grey-area · · focus · HN ↗
          That OpenAI stole the discovery from them.
          1. daishi55 · · focus · HN ↗
            That is very much not the consensus of experts.
            1. pixl97 · · focus · HN ↗
              It's kinda funny how hard the anti-ai people bias themselves and in the same sentence will say the AI people are irrational.
              1. krupan · · focus · HN ↗
                The point of this article we are all commenting on is that we are all irrational and we all would do well to keep that in mind and be more skeptical
                1. pixl97 · · focus · HN ↗
                  Skepticism is a fine balance.

                  If you're not skeptical you invite idiocy to your bed to sleep with you.

                  If you're too skeptical you put yourself in a box of ignorance that puts you at a disadvantage.

                  Skepticism doesn't really work well without the dialectic. For dialectics to work well you need knowledge of your arguments and counter arguments.

                  The problem with most articles and online posts is we get "no no no" and "yes yes yes" and very few "well, maybe".

                  1. krupan · · focus · HN ↗
                    Absolutely true and the comments here are mostly in one of those two categories, sadly
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.