• Swedneck@discuss.tchncs.de
    link
    fedilink
    English
    arrow-up
    2
    ·
    15 hours ago

    i like to think of genAI stuff as a robot with a camera, especially with how things tend to have very very specific composition, the subject always has to go right in the middle.

    • Sv443@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      14 hours ago

      True, since the training datasets are of very specific known things, categorized in a machine-friendly way, it has a very objective, unambiguous view? If that makes sense?