[thunderbird120]

Except it is actually if you've been following the workflows which have popped up since stable diffusion's release. These models are not limited to just making a image which vaguely matches some description. They can also inpaint specific portions of an image, turn a vague sketch into a finished looking image, and blend together different images in convincing ways. A workflow for creating a high quality looking piece goes something like this:

1) Quick, low quality sketch

2) Generate detailed scene based on that sketch

3) Fix obvious errors by re-generating specific parts of the image

[joe_the_user]

I haven't heard of anything like this but I haven't following closely.

I also have questions, what is the "Generate detailed scene based on that sketch" process? Can an algorithm do this? Can a person easily do this? I'd be more impressed than with the raw pictures. Similarly with your #3.

[thunderbird120]

The Image2Image part of StableDiffusion lets you do that. For example https://old.reddit.com/r/restofthefuckingowl/comments/x4w3mn...