• WhatAmLemmy@lemmy.world
    link
    fedilink
    English
    arrow-up
    45
    ·
    11 days ago

    TBF, the veracity of the information is relatively field dependent. Structural engineering? Yeah, probably still as relevant as the day it was published… Quantum computing or astrobiology theory? Far more likely to be superseded or debunked.

    • InverseParallax@lemmy.world
      link
      fedilink
      English
      arrow-up
      14
      arrow-down
      3
      ·
      11 days ago

      I want to see AI papers compared to today, we basically tore the guts out of everything, I don’t even think most of Minsky is applicable anymore (perceptrons particularly have been replaced with vector meshes from word2vec).

      • WhatAmLemmy@lemmy.world
        link
        fedilink
        English
        arrow-up
        8
        ·
        11 days ago

        Statistical modeling and machine learning theory goes back several decades. I’m not sure LLM’s even use new algorithms. They may just apply various techniques that improve the performance and accuracy of pre-existing algorithms.