It’s about time I use this computer. I have Stable Diffusion (a fairly old version) generating images from this prompt:
two muscular men in jeans and tight white t-shirt full body hiking boots visible standing close outside dry hills background with a wolf sitting beside them
The results are “interesting”. The faces in all of them are hideous, so ignore that problem. A few samples that are close to what I’m looking for…
Thematically, I like this one – and it has the space needed for cover-izing. It does not “say” LitRPG, at all, though.
Poor wolf. This is one of the few where the AI realizes that “leg day” is a thing. Many of them have muscular upper bodies, but twigs for legs. Torn jeans are undesired (there are reasons).
I didn’t ask for a werewolf, but it’s not half bad. No room for cover-izing, but it does feel more LitRPG.
I think shirtless sends the wrong vibe. Yes, there is shirtless stuff going on, but that’s not the point. And there is something wrong with the wolf. I can’t tell quite what it is, but that’s just not right.
I like that one, although the wolf has a leg for a dick. And I just noticed that the leftmost foot is not standing on anything.
Next steps:
- Upgrade everything, particularly the models. This is over two years old. AI has vastly improved.
- Learn how to use Stable Diffusion, again. There are models for people (and porn), but I don’t remember how to tell it what model to use. One can also multi-phase it to take something close and tweak it. I don’t remember how to do that, either.
- Work with the prompt. Low camera angle and wolf between like the last one. Get something LitRPG-y in there. Right now, all of those say Camping Story at best and Bestiality Gay Romance at worst.
Note that I didn’t say “short hair” or “white men”, but got both. That’s appropriate for the people, but unexpected.
Marked Up 