With Stable Diffusion I can now build reinterpretations of my work. They are far from perfect, sometimes it fails spetacularly but the good tends to outweight the bad. The output pictures can still be edited manually, of course, although these below are direct outputs.
I also tried making an “original” based on text2image and image2image and after a couple of hours of fiddling around and manual editing I got a solid hardcore image
