A thread for people to share any artworks they've made with DALL-E, Stable Diffusion, Midjourney, and similar programs of the like.
If you want to discuss the concept of Generative/AI art in general, please go to the OTC thread.
The thread you are in right now is for people wanting to share their works.
Putting this in Yack Fest instead of Visual Arts due to the V.A. forums pretty much being zombies- moving, but not really actively alive.
(Post-sticky edit: Feel free to post generative stuff that isn't strictly visual art, such as music or text, here.)
Edited by plakythebirb on Mar 27th 2024 at 1:45:10 PM
Plakis Morakis
Starting off with something pretty cool I've generated in my spare time, though it's pretty freaky so I'm foldering it:
Using Stable Diffusion, prompt was "dark nightclub slaughterhouse", and seed was 2724619.
Welcome to Ideal's World
a few from a while ago, made with Wonder AI; im not sure what training model it uses. prompt for these was "among us crewman"
i was astonished when it spit out what was clearly an attempt at a Magic: The Gathering card (same prompt):
šorange fursonaš§”
I prompted this from Craiyon for an Inktober piece
https://huggingface.co/spaces/stabilityai/stable-diffusion
Note to Self: So, image uploading lowercases everything and all weird characters become "_". Needed to recover a file location.
I wonder how many non-Magic: The Gathering words have obvious MTG connections, like Crew(man) and Artifact. Gonna try some Enchantment.
Edited by Malady on Dec 23rd 2022 at 12:20:38 PM
Disambig Needed: Help with those issues! tvtropes.org/pmwiki/posts.php?discussion=13324299140A37493800&page=24#comment-576
Figure of Hourai
Anyone set it up to run locally? I've seen a few guides (including some Docker ones), but the official documentation gets sidetracked with compiling facebook's compilers, and all the docker stuff is unaware of v2 coming out so the change in checkpoint names makes that an... interesting proposition.
Avatar Source
Plakis Morakis
For Stable Diffusion, I primarily use Stable Diffusion UI, which wasn't too hard to set up.
Figure of Hourai
Yeah, that definitely beats having to do it by CLI. I could have done that, but such a pain. ![[lol]](https://static.tvtropes.org/pmwiki/pub/images/lol2v12_4863.png "[lol]"){kind=small}
Been messing around with other checkpoints and stuff since I got it set up. The one trained on booru images can make for some pretty nice stuff, too.
◊
Edited by RainehDaze on Dec 24th 2022 at 9:28:48 AM
Avatar Source
That one I recently created. Though I forget what site I used. Just a quick humanoid reaper, cause there was nothing close online via some image site otherwise. The disguise was mostly shed quickly, but still.
Plakis Morakis
![[up]](https://static.tvtropes.org/pmwiki/pub/smiles/arrow_up.png "[up]"){kind=icon}
Question: Do you have the image in any higher resolution? The one you posted seems to be rather... crusty pixel wise and I'd like to be able to see more detail of it.
Oh, right, it was done on Picrew.
The Showrunner of Dzwiedz 24
The Showrunner of Dzwiedz 24
Plakis Morakis
Please keep discussion on the ethics of AI art to the On-Topic Conservations thread.
Welcome to Ideal's World
The Showrunner of Dzwiedz 24
I CAN NOT.
And I said The Extremist Was Right I was right i was right iw ass right...
I WAS!
Edited by NotSoBadassLongcoat on Dec 24th 2022 at 2:20:34 PM
"what the complete, unabridged, 4k ultra HD fuck with bonus features" - Mark Von Lewis
Chill out. AI art has no copyright anyways, so there's really no profit in using it. And stop being so quick to condemn it, it's probably gonna be sorted out legally and ethically soon. It's only a matter of time before people have to ask permission to use art to train an AI.
I'll admit, AI art can be a tool. The real fear comes from when it's used as a replacement.
Edit: Thought I just saw a damn skullfucker be used but two new posts were made actually
Edited by CardboardBot on Dec 24th 2022 at 4:21:17 PM
The Showrunner of Dzwiedz 24
It is a right tool, I agree. I wanted it to add a corridor to the execution room I asked it to scale up. It did. I am not going to creedit it. No no no no no.
I need a model dressed up as Ruth Ellis. I will use it then. I will dress in my imperial finery, clean shaven and with Aquila lid.
"what the complete, unabridged, 4k ultra HD fuck with bonus features" - Mark Von Lewis
Ordinary Corrupt Android Love
Uh, yeah, look, I loathe AI """"art"""" too for its more widescale repercussions (which is, as mentioned, more suited for OTC anyway) but I'm not gonna through some big multi-post fit over some thread using it for funsies, even if I have my personal qualms with doing even that.
Edited by AyyItsMidnight on Dec 24th 2022 at 8:31:12 AM
Self-serious autistic metalhead who goes by any pronouns. (avvie template source)
The Showrunner of Dzwiedz 24
I am still trying to make it work.
- Do you need an account to see that image?
The Showrunner of Dzwiedz 24
I don't know, it should display all my public images. There's a couple of them, mostly based on my old drawings, and two or three used to replace the face of the Big Brother with one of a hobo.
"what the complete, unabridged, 4k ultra HD fuck with bonus features" - Mark Von Lewis
Plakis Morakis
Trying to un-derail the thread, I've been messing around with the prompt "tree demon" (with neg. prompt of "cropped") and got this cool lad:
that looks really good. it does seem that AI is much better at intricate but nonspecific patterns than simpler ones with a definite layout.
Plakis Morakis
Frankly, "demon" tends to be one of my favorite words when it comes to casual prompting because it's really good for getting cool/creepy stuff.
Welcome to Ideal's World
Figure of Hourai
It's also better if you try to avoid naturalistic-looking people.
(amusingly, after setting up the UI thing yesterday, it promptly got a big update today. )
It will be interesting to see what it's like given another year or two for people to obsessively try to iron out weird faces and stuff.
Edited by RainehDaze on Dec 24th 2022 at 2:23:48 PM
Avatar Source
A thread for people to share any artworks they've made with DALL-E, Stable Diffusion, Midjourney, and similar programs of the like.
If you want to discuss the concept of Generative/AI art in general, please go to the OTC thread.
The thread you are in right now is for people wanting to share their works.
Putting this in Yack Fest instead of Visual Arts due to the V.A. forums pretty much being zombies- moving, but not really actively alive.
(Post-sticky edit: Feel free to post generative stuff that isn't strictly visual art, such as music or text, here.)
Edited by plakythebirb on Mar 27th 2024 at 1:45:10 PM
Welcome to Ideal's World