• Possibly linux@lemmy.zipOP
      link
      fedilink
      English
      arrow-up
      12
      arrow-down
      2
      ·
      9 months ago

      That’s a fair statement however I though it was relevant as this is 10 months old and we have AI now that can produce very convincing videos

  • apotheotic (she/her)@beehaw.org
    link
    fedilink
    English
    arrow-up
    6
    arrow-down
    1
    ·
    9 months ago

    It amuses me to no end that AI doesn’t yet know how to make people actually eat the food. I guess most stock photos are either someone about to, or has just, put the food in their mouth, but there’s not much for the interim. So we get horror like this

    • hydroptic@sopuli.xyz
      link
      fedilink
      arrow-up
      3
      ·
      9 months ago

      It’s a bias in the training data, not a lot of pics of people eating eating compared to all the stock photos of women smiling at salad etc.

        • hydroptic@sopuli.xyz
          link
          fedilink
          arrow-up
          1
          ·
          edit-2
          9 months ago

          Oh yes my point was that you’re exactly right. Current neural net architectures need absolutely ridiculous amounts of labeled data to “learn” anything, and they’ll just repeat the biases in their training data.

    • Ignotum@lemmy.world
      link
      fedilink
      arrow-up
      2
      ·
      9 months ago

      Updated version would just look like will smith eating spaghetti normally, that’s no fun