//=time() ?>
Interpolating through stable diffusion prompt embedding space while holding the seed fixed.
#dalle2 has great coverage of the entire prompt. Instead of smushing all the nouns together, it includes most of the details.
"Dolphins with opposable thumbs discover tools"
Here's what a few years of progress on text-to-image generation looks like, one prompt at a time.
"Frank Sinatra as a purple alien in surrealist style"
1) AttnGAN (2018)
2) CLIP+VQGAN (2020)
3) CLIP+Diffusion (2021)
4) DALL-E 2 (2022)