1.
2. 3. 4. 5. 6.

  • tsonfeir@lemm.ee
    link
    fedilink
    English
    arrow-up
    7
    ·
    9 months ago

    Assuming some are real and some are fake, my bet is 1 and 3 are fake.

    • DominusOfMegadeusOP
      link
      fedilink
      English
      arrow-up
      8
      ·
      9 months ago

      I am, but I was going to wait until at least the middle of the afternoon.

  • lad@programming.dev
    link
    fedilink
    English
    arrow-up
    4
    ·
    9 months ago

    I agree with many others, 1, 3, 5 look most suspicious (assuming numbers precede pictures)

    Others seem to have heavy upscaling but it doesn’t tend to become too patterned. In 1, 3, 5 grass tends to become something like a fractal rather than grass, plus reflections are sometimes off (that is why 2 is a bit suspicious because of a small tree reflection, but maybe it’s just hard to see the original tree), and the branches tend to grow from multiple trees at the same time

  • Jaybob32@lemmy.ca
    link
    fedilink
    English
    arrow-up
    4
    ·
    edit-2
    9 months ago

    1,3 and 5. They have that weird pattern in the grass and trees. However, it’s getting close. I think soon most people would never be able to tell.

  • mindbleach
    link
    fedilink
    English
    arrow-up
    4
    ·
    9 months ago

    Nature shots have been cheating for a while, because most errors are still plausible.

    The major tell is how screen-space anything is. In real life, there’s very few angles where the top of a close thing stops at the bottom of a far thing… but neural networks aren’t modeling depth. Probably. So things are tangent or coincident all the dang time. Even in the patterns of grass and brush and whatnot, where the network does T-junction patterns like brickwork or cracked pottery, when it should be closer to woven or thatched.

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

    With the HDR effect added to these it’s honestly difficult to tell which are generations and which are just heavily post processed