Michael Friesen

@advadnoun yknow, for a bunch of Monet stans, these folks are sure acting like a real buncha Louis Leroys heyooooo

You wouldn't photo a painting.

Really love stable diffusion deciding that the raft from Gericault's Raft of the Medusa is an inflatable one

I suppose Neo-Neo-Dadaism *was* a little bit past due...

i've been trying to keep my head out of the discourse tonight but man, this just hurts. :(

[whispering quietly] pssst AI art made without using any text-to-image stuff can still be cool tooooooooo

"Neon synthwave art by Hieronymus Bosch"

And so the #stablediffusion gates have been opened

made with #CogVideo

@_akhaliq @Gradio @huggingface

Image prompts are now a option for the @Gradio Demo for CogVideo on @huggingface Spaces demo: huggingface.co/spaces/THUDM/C… Get started with Gradio: gradio.app/getting_starte…

@_akhaliq @Gradio @huggingface Just like a real corgi

Cache Examples are now working in the @Gradio CogVideo demo on @huggingface Spaces "Pikachu on a skateboard" demo: huggingface.co/spaces/THUDM/C…

@advadnoun Art is whatever you can get away with, quoth McLuhan

What's common between art forms/mediums is a good indicator of what's actually necessary for something to be art.

"AI art is bad because it's derivative! AI art is bad because it's derivative!!", i continue to insist as i slowly shrink and transform into an iteration of The Kiss.

And there she goes

first experiments with CogVideo! send me prompt ideas

Outputs from ruDALL-e finetuned on M.C. Escher lithographs, used as #dalle2 image prompts

@beckerfuffle Oh the heads are actually the entire originals (it's rudalle finetuned on a few thousand facecropped portrait paintings)! So like these:

@MichaelFriese10 What do the originals of these look like?

Using #dalle2 to autocomplete rudalle portraits