this post was submitted on 23 Jul 2025
79 points (95.4% liked)

chapotraphouse

13991 readers
811 users here now

Banned? DM Wmill to appeal.

No anti-nautilism posts. See: Eco-fascism Primer

Slop posts go in c/slop. Don't post low-hanging fruit here.

founded 4 years ago
MODERATORS
 

But they conveniently leave out that it costs money to do anything with AI. It's more like "open to anyone with a credit card." The vast majority of people don't have computers powerful enough to run generative AI models locally, and even then, server farms with a billion GPUs will always produce better results

This means that people have to rely on corporate platforms where you buy tokens that you use to get pulls at the various AI slop slot machines, hoping you get something decent. The mechanics more closely resemble a gacha game than any kind of artistic process

By contrast, learning how to draw, animate or make 3D models costs nothing. There's free tutorials and tools everywhere, and you can also just pirate commercial ones if you want

you are viewing a single comment's thread
view the rest of the comments
[–] NuraShiny@hexbear.net 23 points 1 month ago* (last edited 1 month ago) (3 children)

Ask any artist if they are satisfied with their art. They won't be. That is the nature of the beast. Doesn't matter how good they are at it.

I have done art on and off for years as well. I would estimate I did maybe a few thousand hours of art in my life. Other people have done literally 10 to 100 times that and they are better because of it. You hone your skills over time. Anyone like me dabbling in art has to accept that there are countless people who put in more time, who are better and who will stay being better because they will keep putting more time in then us.

Using AI, you steal those hours from those people, for a product that doesn't show what's in your head. It's stilly to pretend AI can do that and you know it. If that were true, all the art prompters wouldn't include the names of existing artists in their prompts.

So yea, put in the fucking time, or get off the field. Don't be fucking selfish and then try to defend it in flimsy ways.

[–] Le_Wokisme@hexbear.net 3 points 1 month ago (4 children)

Ask any artist if they are satisfied with their art. They won't be. That is the nature of the beast. Doesn't matter how good they are at it.

duh. but "good enough" or "ok this is fine i need to work on something else" are out of reach too.

Using AI, you steal those hours from those people,

lmfao

for a product that doesn't show what's in your head. It's stilly to pretend AI can do that and you know it. If that were true, all the art prompters wouldn't include the names of existing artists in their prompts.

idk lots of musicians try to copy the guitar tone or whatever else of their influences. what's in your head might be a copy of styles you've seen until you develop your own style... idk what "your own style" maps to in generative "ai", i'm not here to defend ai or say that the people using it are doing art per se, my whole thing is that I can't get even remotely close to an acceptable level of competence and it ticks me the fuck off when someone goes "aNy OnE cAn mAkE aRt" or upholds capitalist framings of intellectual property.

[–] PKMKII@hexbear.net 12 points 1 month ago

I can't get even remotely close to an acceptable level of competence and it ticks me the fuck off when someone goes "aNy OnE cAn mAkE aRt" or upholds capitalist framings of intellectual property.

In that regard, I do think there’s a distinct difference between sampling (whether a still image, moving pictures, or sound) and using the AI treat generator. There’s plenty of people who can’t play a musical instrument worth a damn, but give them some loops, one shot samples, and sequencers and they can transform the samples material into something new and fascinating. Likewise with visual media.

12tone on YouTube had this interesting argument that samples, rather than being “cheating,” are turning a particular performance in time into an instrument unto itself. Whereas AI is a smudging of all past art into an average. So while sampling celebrates the greatness of a performance, AI reduces it to a Borg-like state.

[–] NuraShiny@hexbear.net 8 points 1 month ago (1 children)

It's always sad when I gotta actually block someone from hexbear, but since you are just a larper who doesn't care about others as soon as you can have the slightest whiff of a treat for yourself, I am not really sad about it.

[–] Le_Wokisme@hexbear.net 2 points 1 month ago

??? comrade if you see this while logged out, i don't use gen ai to make visual "art".

this is only about the misunderstanding that people want to make stuff and can't because what they want to make is more specific than the head up ass maximum position that anything is art regardless of quality.

[–] sunshinesoul@hexbear.net 3 points 1 month ago* (last edited 1 month ago) (2 children)

i am a professional artist (and maybe this makes me a bit out of touch,) assuming that it's for personal use and not for profit at all, i would rather have someone take my work into photoshop/gimp/krita/whatever and trace it near directly or make edits to it to fit their vision than have that person go spend money to further refine The Slop Machine. other artists may have differing opinions on this but since generative AI has gotten popular i just simply do not care anymore as long as my work isn't being fed to train image models. hell, if you're tracing someone else's work using tools on paper, that's still building muscle memory and linework skill and while not the ideal scenario it's doing more for you than you might think. with generative AI you are paying to generate an image based off countless images that already existed from artists that were not paid for their work to be included in the model. is that...not capitalistic or not at the very least exploitative?

[–] Damarcusart@hexbear.net 1 points 1 month ago

Also a professional artist, and I agree with you. Someone making a collage or a trace from my art, or even just photoshopping it is still practicing creativity. Someone typing a prompt into the art slot machine has no creative process, they aren't learning anything, experiencing anything, struggling with anything, it's just empty output. It's just a "pretty picture" there's no sense of accomplishment or understanding. One of the most rewarding things in my life is when I draw something and recognise that I flat out wasn't skilled enough to do that 6-12 months ago.

And this can apply to any hobby or skill, are people so alienated from themselves that even the most basic concepts of satisfaction at self-improvement are seen as outright insults to them?

[–] Le_Wokisme@hexbear.net 1 points 1 month ago

end-users aren't necessarily paying into anything other than their own electricity bill. We universally have a problem with the companies profiting at your expense of course, and that might be the more common case.

Those parts of "ai" discourse are tangent to someone's aptitude or ability to have something that looks how they want it to look and i'm trying to limit myself here to being mad about the position that art is accessible already because everybody can make shitty art that isn't what they want to make.

[–] sudo_halt@lemmygrad.ml 1 points 1 month ago (1 children)

Stable diffusion works and learns on a conceptual level. If an AI model even understands artist tags, bias was trained into the model with a LoRA.

You looked at other people's art when you learned to draw you thief? Hand over the pencil and come out with your hands on your head

[–] NuraShiny@hexbear.net 7 points 1 month ago (1 children)

So you are seriously saying that AI has the ability to understand concepts? Waow.

Please educate yourself and don't spout bullshit, thanks.

[–] sudo_halt@lemmygrad.ml 3 points 1 month ago* (last edited 1 month ago)

Not concepts as we humans operate on, but concepts in the feature extraction sense. You have a gross misunderstanding of how these models work.

Imagine it as how an image detection model learns how to detect images. It extracts features (that are vectors of data that are completely meaningless to humans). Image generation can be somewhat imagines as the inverse of a feature detector, it creates features from noise.

These models are not plagiarizing your work. You teach it the shape a dog is expected to have, and you teach it the color black. Without such thing existing in the training set, it can generate a black dog.

This is fundamental to stable diffusion otherwise it could never work at all.