Mostly it’s a process of sketching, prompting Stable Diffusion in image-to-image and then digital painting over outputs and going back and iterating. Anything I’ve put out here has been through several iterative transformations. It’s still a lot of work to get something good. I’m ready to share about 1% of the images I make.

Replies (4)

Default avatar
nobody 2 years ago
My Midjurney Results. Nice to look at but I wanted a seahorse and not a pregnant plant:) Not that easy to get what you want from AI.
💜 these are gorgeous, but yeah, getting control of your own vision takes some input from you beyond the text prompts for sure. Does Midjourney provide a way to do that? I haven’t really dug into it.