• Nepenthe@kbin.social
    23·
    1 year ago

    The thing about these is whenever I see them the first time, I can read them because scrolling quickly leaves the picture just unfocused enough.

    And then I scroll past, back up because I’m very sure of what I just read, and it’s just a photo of cats or something. Amazing. My mind knows what it saw, but it has no proof.

    • Fubarberry@sopuli.xyzEnglish
      22·
      1 year ago

      AI generated art. To make something like this you need stable diffusion (either running locally if you have a good GPU, or a website that will let you use it).

      In stable diffusion there are plugins called control net, where you can add details you want in the final image. Usually people use it to force specific poses or scene arrangement, but you can also use it to make sure that the final image has a clear pattern in it such as words or a geometric pattern. A recent popular example was a town with a clear spiral effect.