• crunchy@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    24
    ·
    8時間前

    You can tell because AI tends to copy and paste one face when it’s generating groups of people. And, it can only do front view, profile, and ¾ view.

    • einlander@lemmy.world
      link
      fedilink
      English
      arrow-up
      9
      ·
      8時間前

      Manga/manwa tends to be like that though. There is a reason their are memes for “mob” background characters. It’s the 3 fingered people that make it complete ai slop.

    • rockSlayer@lemmy.blahaj.zone
      link
      fedilink
      English
      arrow-up
      4
      ·
      8時間前

      You’re right, but I’m not good at checking those details, I was checking the background mostly. AI still tends to struggle with things outside of the image focus. If you look at the faces of the people in the background, they’re all over the place with random eye and mouth lines that don’t match anything an artist would do.

      Also the random hand with a pen in the foreground that isn’t connected to anything.

      • TowardsTheFuture@lemmy.zip
        link
        fedilink
        English
        arrow-up
        5
        ·
        8時間前

        For me it was… him leaning over and writing on a non-existent podium…. But yeah also the pen hand on stage going nowhere

        • TheTechnician27@lemmy.world
          link
          fedilink
          English
          arrow-up
          2
          ·
          edit-2
          2時間前

          Oh, wow, that’s a real fucky optical illusion. I saw him holding a couple pieces of paper (like a speech script) sideways, but it also looks exactly like he’s writing on an invisible podium.


          Edit: Wait, you meant the kid scribbling onto the stand. What the absolute fuck. lmao