me_irl

7998 readers

831 users here now

All posts need to have the same title: me_irl it is allowed to use an emoji instead of the underscore _

founded 3 years ago

MODERATORS

herrwoland@lemmy.world

486

me_irl (feddit.org)

submitted 2 months ago by NichEherVielleicht@feddit.org to c/me_irl@lemmy.world

31 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] Get_Off_My_WLAN@fedia.io 28 points 2 months ago (1 children)

I'd say many artists can at least probably tell because he starts out drawing with a light sketch first. People who don't draw wouldn't think to generate the steps in between like that.

Also, the rock is pretty clearly the same. And the drawing is consistent.

[–] MightEnlightenYou@lemmy.world 6 points 2 months ago* (last edited 2 months ago) (1 children)

I agree that AI slop wouldn't be able to produce this. But for someone who knows what they're doing with image generation and open source tools, these images are easy to generate through sequential generation where: first an overview picture with all the objects is produced and then the first image is used as the input for the second image and so on.

But the way, I don't think that the images are generated. I'm just saying that we wouldn't be able to tell if they were

[–] bunnyBoy@pawb.social 7 points 2 months ago

I'm not doubting what you're saying, if anything I'm trying to make sure I understand how that kind of workflow would create something like this. My understanding of image to image generation wouldn't get results anywhere near this consistent.