The fediverse used to feel pretty anti-ai, but over the past month or two I’ve noticed a LOT of generated memes and images, and they tend to have positive votes.

Has there been a sudden culture shift here? Or is there a substantial percentage of people just unable to tell the difference anymore?

  • TheLeadenSea@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    26
    arrow-down
    8
    ·
    1 day ago

    As long as it’s intentionally made by a human and the end result is high quality, I personally don’t really care what AI or other tools they did or didn’t use to create that result

        • Scubus@sh.itjust.works
          cake
          link
          fedilink
          arrow-up
          1
          ·
          19 hours ago

          I think they were making a joke. Its the same vein as if you said you dont care about dad jokes then they made a really bad one and said “are you sure?”

        • Riskable@programming.dev
          link
          fedilink
          English
          arrow-up
          9
          ·
          1 day ago

          Because they probably just took the first image that was generated without even checking it. For today’s generative image AI you usually want to generate great big batches then pick the best ones.

          Then if you’re looking for high quality in your meme you’ll also use inpainting and pull the result into Krita or the GIMP to make further improvements (e.g. add text with a proper outline/drop shadow). Then when you’re done you’ll use AI to upscale the image to a great big size so that it’ll look nice when you print it out and hang it on the wall in your cubicle or the stick it to the office fridge or some filing cabinet 😁

    • pleasestopasking@reddthat.com
      link
      fedilink
      arrow-up
      8
      ·
      1 day ago

      I feel like the problem with AI generated content is it’s so easy for anyone to generate stuff, so there can be a huge amount created with little effort. There is high quality AI-generated content, but whew there’s a lot that’s total slop.

      I don’t know what the best response is, though. Requiring disclosure of AI-generated comment doesn’t seem like it would help because that’s going to be mostly honor system. User-flagged could be used to brigade/suppress posts. Really it’s probably just a matter of blocking users and communities where you see consistent slop.

      • moonlight@fedia.ioOP
        link
        fedilink
        arrow-up
        4
        ·
        1 day ago

        Yeah I agree. The issue is that image generation tends to result in maximally bland outputs, and the people who post it tend to put minimal effort in.

        I’m not categorically anti-ai, but I feel like I am in practice.