//=time() ?>
@SeekerStreams DARE ACCEPTED. (DALL-E 2 totally ignored my "yellow" instruction)
"Close-up of the corner of a comic book panel outlined in white depicting the face of an ape with flat yellow color and black shadow, with white thick tamarin whiskers, background is orange"
DALL-E 2 prompt: "Google DeepDream"
we are two layers of neural network indirection deep here
If you have read about the Void Hog on the Exe Wiki you know he is a creature from a dimension known as "The Void".
I used an AI (like Dall E) to generate images from this dimension, here are some interpretations.
I honestly liked the result and it can be canon.
DALL·E 2 extended this scene of The Ghost and Molly McGee #TGAMMfanart #tgamm #theghostandmollymcgee
That moment when you've poured yourself a bowl of late night cereal and taken a full bite before realizing the milk is sour.
This real life moment interpreted by @OpenAI DALL·E 2
"What DALL·E would look like if DALL·E was an anime girl"
special thanks to @gnostiquette for helping us use actual DALL-E to capture the vibe of Mercenary Planet!
catch up on the future of "lesbian space atrocities" before Chapter 3 next week:
https://t.co/lZW9MwZXUL
playlist out now:
https://t.co/eoAoVgzYaZ
Drew some characters I generated using a friend's name in Dall-E mini
#dalleMini #dalle #ArtistOnTwitter
i have just learned what DALL·E thinks Golden-Dawn style Enochian magic looks like
Just realised https://t.co/zt3wiVeTOu – the tiny prompt generator @daveforall and I built – could generate inputs for DALL-E & co.
But turns out they can't do abstract representation of concepts yet.
So "{noun} to illustrate {concept}" doesn't work. Perhaps a future benchmark?