• nimpnin@sopuli.xyz
    link
    fedilink
    arrow-up
    59
    ·
    edit-2
    23 hours ago

    The use of LLM had a measurable impact on participants, and while the benefits were initially apparent, as we demonstrated over the course of 4 months, the LLM group’s participants performed worse than their counterparts in the Brain-only group at all levels: neural, linguistic, scoring.

    https://arxiv.org/pdf/2506.08872

    • wildncrazyguy138@fedia.io
      link
      fedilink
      arrow-up
      18
      ·
      22 hours ago

      I equate it with doing those old formulas by hand in math class. If you don’t know what the formula does or how to use it, how do you expect to recall the right tool for the job?

      Or in DND speak, it’s like trying to shoehorn intelligence into a wisdom roll.

      • misk@sopuli.xyz
        link
        fedilink
        arrow-up
        27
        arrow-down
        2
        ·
        22 hours ago

        That would be fine if LLM was a precise tool like a calculator. My calculator doesn’t pretend to know answers to questions it doesn’t understand.

        • Swedneck@discuss.tchncs.de
          link
          fedilink
          arrow-up
          4
          ·
          14 hours ago

          the irony is that LLMs are basically just calculators, horrendously complex calculators that operate purely on statistics…

    • kate@lemmy.uhhoh.com
      link
      fedilink
      English
      arrow-up
      2
      ·
      21 hours ago

      Participants were restricted to using ChatGPT? So I am smart because I use Claude and there’s no science to tell me I’m wrong 😎👍