Source (Bluesky)
Alt text from source
01: A girl labeled ‘Artists’ is holding a drawing tablet and pen under her arm. She is wearing round glasses and a blue sweatshirt. She has messy dark-brown hair and a brown skin-tone. She says “We don’t really wanna MAKE A.I. art.”
02: A guy labeled ‘Art Enjoyers’ holds his hand out while speaking. She is wearing a purple and yellow hawaiian shirt with a floral pattern over a white t-shirt, and has red hair and a light skin-tone. He says “We don’t really wanna SEE A.I. art.”
03: Behind them both, there is a cute girl in a business suit with a pink tie. She is blushing a bit and has pink eyeshadow, and looks upset. Her messy shoulder-length hair is parted in the middle, and held by two hairclips: one that looks like a red arrow pointing down, and one that looks like a green arrow pointing up. She says “Um … I-Isn’t there someone you forgot to ask?” She is labeled ‘Shareholders’.
04: She puffs up her cheeks and pouts, a tear is on the verge of falling from one of her eyes. ‘Artist Enjoyers’ Guy is now in front of her yelling “AH!! SHAREHOLDERS-CHAN!!” while ‘Artists’ Girl is in front of her yelling “WE’RE SORRY WE HURT YOU!!!”
I am not artistic, but I want to make my own art. Why should I not use AI?
You are still not making your own art.
I am.
I’m not athletic, but I want to run a marathon. Why should I not use a motorcycle?
If your goal is to get from A to B then the motorcycle is a fine way to do so.
And if you want to generate soulless slop then using AI is a fine way to do so.
Agreed. But an artist could also use AI to generate art.
And an artist could use nuclear waste to ‘generate’ art, as you say. Doesn’t mean we need nuclear waste.
They do
But the goal is to run a marathon, which I’m not sure but I have a bunch that using a motorcycle doesn’t count as running, although I should triple check because who knows, the earth devs could’ve pushed an update where riding a motorbike increases your athleticism stats.
When you paint the goal isn’t to do 50 brush stokes per minute. The goal is to produce a painting. Likewise, in your analogy the goal is to get to Athens from Marathon.
Turns out, I’ve been producing paintings for around 3 years and digital art for around 2 years, and I’ve gotta say, it’s rarely about getting the product out there, it’s the journey, the iterations you make, the the inspiration you have, the feedback you get, the materials/programs you use, etc.
If you’ve ever created artistic pieces for more than just the small amount of paintings you’d produce in primary and high school, you’d realise that actually finishing the piece is only an implied goal, not an “i need to get this done by tomorrow or else my boss will fire me” situation.
Oh yes. From the artists perspective a piece is rarely finished. There always seems more to add or change. I very much respect artists who can create minimal pieces that still seem finished.
Using AI within these iterations does not seem to be cheating to me, but many people in this thread seem to disagree.
But from a viewer perspective they can only imagine the work that has gone into a piece. The end result is usually all they have.
Art is not about the product. And your lack of artistic literacy displays this attitude. Art is about meaning and process. Here, try to use your empathy to learn something for a change.
True, some art is more about the process than the result, but the painting of the mona lisa is certainly the product, not the process.
See, that’s how I know you know nothing about art. You mention mona lisa because it is widely popular. But don’t know that RX of the painting show that davinci painted other two portraits underneath before the final result. Or the fact that it became famous because it was stolen for a while and when restored, the Louvre placed a security glass which, along with the burglary story, piked the interest of audiences for a painting that was thus far mostly ignored.
Art is more than the product, it’s its history, context and the value humans place of them. Without realizing this, art will never be more than a product to you. In which case I would suggest to stick your art criticism to marketing and ads.
Art history is very interesting, but let’s not kid ourselves. The tourists are packed into a small part of the Louvre only for the finished product.
You are not achieving your goal though. You are not making art with AI, something else is doing it.
In many art forms, something else is doing it. The paintbrush and gravity do most of the work in a Jackson Pollark.
You could learn to be an artist. Just like how a musician learns…or a teacher learns…holy shit people are so lazy now!
The definition of an artist doesn’t mean you have to use a skill. This artist created an oak tree from a glass of water.
There is no correlation between effort and art.
Then why aren’t you doing it?
I want food but I am not cooky, doesn’t mean that ordering pizza makes you a chef.
I want to go but I am not a car person, doesn’t mean that calling a uber makes you the driver.
I am doing. And I’m using AI. I don’t see why this is a problem.
Because using GenAI to call yourself an artist is like hitting play on Spotify to call yourself a musician. You’re not doing the thing, thus you are not the thing doer. A kid making turd shaped dogs with play-doh is closer to being an artist than you inputting prompt into a generator.
DJs are considered artists, and they just press play.
Granted, a one prompt AI generated image is not very impressive and is unlikely to be “good”, but it doesn’t mean it isn’t art.
Read about what DJing is, it’s more about playing the crowd, and less about playing the music. Really, do you never pay attention to anything beyond the superficial level. And DJs don’t say they are the musician, they are the DJ. Even though most DJs are also musicians.
Any DJ with a light show is not reading the crowd at all. They are just pressing play and dancing to a prearranged set.
If you want to use AI to express yourself through art. Go ahead. You are free to express yourself.
It’s not your own art, that’s the whole point. Learn to draw or paint or even use graphic design software like every other artist.
“I want to be artistic without putting in any actual effort.”
If you can only “create” AI art, you’re not an artist. I don’t even care if you use it, just don’t act like you’re anywhere near the same as someone who spent time actually learning and honing a skill, or that it’s your own work. It’s the AI’s “work” that was typically stolen from actual artists online.
I’ve never had a problem with using assistants.” Damien Hirst
If he and Warhol are allowed skilled assistants, why not me? Are these artists stealing?
OK, two things:
Skilled assistants are actually skilled assistants. LLMs are not.
You are specifically saying that you don’t want to develop skill. Once you do, then you can take on assistants and train them. You’ve explicitly stated you don’t have the skill where an assistant would be useful. Sure, you can hire someone to make art for you, but they aren’t an assistant. They’re doing the work and you aren’t. They’re the artist, you’re just paying them.
Stop complaining and just learn how to make art. No one was born with the skill or talent. They just practiced and learned. You want to skip this and act like you’re still doing anything interesting or useful. You aren’t. You can, you just refuse to and still want to be treated like you are.
LMMs/diffusion models can do something I cannot, that takes lots of training to perfect. That’s a skill.
The artist is the idea generator, not the labourer. Architects are the artists. Construction workers not.
I can learn using AI.
Yeah, this isn’t even worth a response. Lol.
lol
Well those artists consented to being assistants and having their works used. Not so much with these image training databases
We are discussing if I’m an artist by using diffusion models, not if training data has been fairly compensated.
The answer is no. You’re not an artist.
Then how am I different from Warhol and Hirst who are artists?
They’re not artists, they’re grifters, you’re no different.
Artists are grifters.
To avoid paying restaurant bills, Dalí would draw on the backs of cheques.
Michelangelo faked an ancient Roman sculpture, buried it to age it, then sold it as a rare antique.
Yves Klein sold “invisible art"
Elmyr de Hory forged Matisse, Picasso, Modigliani etc.
Fuck Andy Warhol and Damien Hurst, I would definitely say the only kind of artists they were is con-artists. They ripped off talent and took credit for what others did. Fuck them.
Ok. Your opinion is entirely consistent with hating “AI artists”. But it does fall in the “That’s not art” argument that has been continual throughout time.
There is no such thing as an AI artist. Point blank.
The art was art, in regards to those two fucks, it just wasn’t their art. They were con men putting their names on the work of others. AI “art”, however, simply is not art, it’s not even AI, it’s just theft processed through algorithms.
If you want to make art then you have to actually make something. LLMs and dispersion models are just plagiarism and procedural generation, you are literally not creating a fucking thing with it, you’re just using software.
Is collage plagiarism? Are found objects not art?
LLMs and dispersion models don’t necessarily create good art, but they do create art.
I disagree.
About collage and found objects?
Because telling something to make a picture is not you making it. It’s that thing making it.
I’m baffled by how this is so fucking hard to comprehend?
Anish Kapoor doesn’t build his sculptures. Warhol or Hirst telling their assistants to make something is not them making it, but they are still the artists.
Lots of artists use it to make interesting compositions, to paint after.
Are you an artist? I mean it’s not art obviously if the only thing you do is to prompt some slop.
If you want to use image generation to generate an image, go ahead. I sometimes do the same for placeholder images.
It’s not art though, it’s just plagiarism that’s been abstracted several times over. I could come up with a really cool search string for Google Images that has an awesome first result. But by downloading the first result and passing it off as my work isn’t art, it’s plagiarism.
If you want real art for your project, commission an artist to make it; or keep practicing until you find yourself good enough to make it.
I’m not gatekeeping art; the creativity, the effort, and the skills of the craft are the gatekeeper.
The process is effectively just doing a Google image search.
It’s interpolation and collating. It is digital found objects
I’m not talking about the algorithm. I’m saying that the process of generating an image is the same as typing in some text to an image search engine.
It is clear you don’t understand how AI works.
Not a single person will pick the first image that is generated. All of them get tweaked and changed before they are seen as completed.
It’s clear you don’t understand how AI works.
It’s basically just creating image after image until you find one that looks cool. About as artistic as searching through a database.
And every time you create a new image, you change the prompt slightly until you get what you are after.
If you are looking for something specific, you never get it at the first prompt. No matter what kind of “prompt engineer” you are.
Same is true for looking through a database. So what’s your point?
Oh they absolutely do at approximately the same rate of choosing the first Google image search result.
Choosing the words. Choosing the type of engine. Choosing what part of the image to keep and what to discard. Choosing to inpaint and outpaint. It is digital Collage Art.
Because actually learning to make art is fun?
Agreed. So is learning how to make AI art.
Typing in a prompt isn’t “learning.”
You know that developing new skills is a good long term way to prevent dementia? Pencil and paper exercise your hand eye coordination in a way that’s healthy long term.
Take a look at comfyui and try to claim that no learning is required.
Art does not have to be pencil and paper.
Can you show something you’ve made?
Almost all AI art I see is anime girls and video game landscapes. Show me something that demonstrates creativity.
I’m no skilled artist, but I’ve spent dozens of hours just blocking. It’s really interesting to try to translate what I see in three dimensional space to two dimensions. How long do you typically spend preparing a project?
What I enjoy is making a rough sketch of what I’m trying to create, the using AI to convert it into something more presentable. Bears in woods. Cityscapes.
Inpainting and outpainting is OK for fixing errors and getting close, but you can’t one-shot an entire image. I’m playing around with building things in layers.
I spend a couple of hours of my time on an image, broken up into many small intervals. Local models, not cloud.
“you’re not a real artist if you use XYZ” has been with us since the invention of photography.
AI slop is not art, but who said you can’t use AI to do art at all? I know several artists (real ones, oil painters, selling their work) using AI.
Okay, I want to see it though. That’s kinda my question. I see a lot of slop, I haven’t seen anything else.
I’m a huge fan of Duchamp even, who seems to be massively misunderstood, as he is typically, in the thread. I want to see evidence of good art with AI.
Edit: it’s really interesting, I’ve had similar conversations multiple times before, but no one actually shares good AI art. Just argues that it is possible, never any proof…
deleted by creator
Ha ha the Dada movement was so… Bizarre. I don’t even know if they themselves even cared about it. It was really a watershed time for painting and art in general IMO. Like Fauvism or cubism or absurdism.
I think people misunderstands a lot about AI and also a lot about art, you see some slop supposedly taking a graphic designers job, and people are up in arms about “AI is making art and stealing artists jobs”. For me it’s not that black & white and say if an energy-efficient AI can help someone make a fantastic comics well then IMO that is good.
AI on itself will never make art, like a camera won’t, but I think lots of people will use them as tools in their art creation journey.
My AI art is better than one shot slop of anime wafu, but isn’t good enough to exhibit.
Partially because it’s a tool made by stealing other’s work. AI exists today because hundreds of thousands of artists and writers and real human people did not get compensated for their work. It’s like support Walmart despite it purposefully shutting down small businesses or buying blood diamonds (is that too far?). It’s something made by hurting other people.
I don’t know if I’d go so far as to say “There is no ethical consumption under capitalism” but like it’s not easy. We all are forced to make shortcuts and are forced to be a part of harmful systems. So I wouldn’t judge you if you used AI to generate art, but I would judge you if you pretended it wasn’t unethical and harmful to both the environment and fellow humans.
I think people are particularly up in arms about this topic because the output is, in some instances, so bad or so commercial or so lifeless that it feels like a really poor exchange for taking money away from artists and burning rare materials and energy. Like it’s so easy to NOT do this thing that it feels extra bad when people do it.
I get it, when midjourney first came out I was a huge fan. In fact I think it’s easy to say I love the idea of AI as a tool for quick creative creation. I think it’s incredibly empowering. I just think there isn’t a form of it today that wasn’t created via value extraction from the working class.
Yes. There is a whole discussion about misappropriation, compensation and rights which I weigh heavily on the artists’ side. But in 100 years that problem disappears (for current models and artists).
My main point is that if an artist uses AI in their workflow it doesn’t make their output not art. And as an extension, there is no reason I can’t be that artist.
Because the end result just isn’t very good.
Usually, that is true. But sometimes it is good. Creating good AI art takes skill and time.
Absolute slop background, how this won anything is beyond me. Also the AI probably lifted the theme and style from actual artists’ artworks. How is this soulless bullshit “good”?
It might fool someone who wasn’t familiar with the stuff that AI can generate, but how does this lead to any actual creativity and progress, and not just repeating overall patterns many artist have drawn before, over and over again until everyone gets saturated with such AI “art”.
This is just an image generator that will be used to cut corpo costs and reduce work opportunities.
You can bet that if you didn’t know it was AI, you would call it a masterpiece.
I can bet, but I won’t. Seems pretty garbled and undefined to me to call it a “masterpiece” anyway. Vibe-AI-painting, at best.
Also, I browsed the artist’s other AI artworks, and lol does this guy have a high opinion of himself and his “works”. (And do the courts not have that, refusing him copyright…) Looks to more in it for the publicity and quick buck, than actual artistic motivation.
Like any skill, you have to do it badly for a while, then you do it kind of ok for a while, then you do it mediocre for a while, and you keep taking those steps, and after some time you will have the skill to craft what you see in your minds eye, that’s why art is appreciated, because it requires effort.
AI is not you creating anything, it’s not a skill, nothing was honed, earned, or done with earnest, you’re just commissioning a program to do statistical algorithms with colors, it’s soulless garbage. In order to create something you have to actually create something.
No. There is a huge amount of art that is unskilled and requires little effort. E.g. https://en.wikipedia.org/wiki/Fountain_(Duchamp)
I think that depends how many prompts are tried, how much inpainting is done and how your loras are created.
Get the fuck out of here with that take. The reason that piece works is because it didn’t take effort.
‘When explaining the purpose of his readymade sculpture, Duchamp stated they are “everyday objects raised to the dignity of a work of art by the artist’s act of choice.”’
It’s a well known artist using his name to put these pieces into a place where you expect “high art” but they’re not, which creates an discrepancy that makes you think about what it all means. It’s more than just a urinal, because of the context in which it’s set. If you can’t understand that then I get why you can’t understand what makes art interesting, and why you can’t do it. It’s not that you lack skill (which you do, because you need to practice it to have it) but rather you’re an uninteresting person who can’t ponder things that are different.
Luckily, this is also a muscle that can be exercised. Stop listening to the very stupid people who think they’re clever by saying certain art isn’t deserving attention. Art is anything that makes you think and has nothing to do with skill (though the ability to make you think, again, is a skill of its own). Just look at art and consider why it exists the way it exists. It’s all purposeful, whether you like it or not. Flex your brain by thinking about the meaning or message, or what you can get out of it that maybe the artist didn’t intend.
Edit: To give an example of how AI art could be used in art is to make it copy Picasso’s (or whoever’s) style, and put it in the context of real ones. Let people examine it and see how little intentionality there is in it compared to an authentic one. It’d allow it to bring more meaning out of the other pieces because the AI one is so meaningless.
The fountain is not the only Found object, but it is the most famous. All are considered art even though most of the effort is not produced by the artist. Similarly with Collage. That was why I chose this line of argument.
AI is capable of “creating an discrepancy that makes you think about what it all means” but the vision comes from the artist and is passed through multiple stages. It’s not just prompt -> algorithm -> art.
Comparing AI to Master is something I’ve not considered before but it makes total sense as a teaching aid.
A teaching aid doesn’t do the teachers job for them in the same way AI doesn’t make art.
These are tools, not artists, there are good uses for LLMs and dispersion models, but replacing art, artists, expression, thought process, and the human condition in general, are not good or properly utilized ways to try to use them.
Some things are difficult. Creating things is difficult, learning a new skill is difficult, self expression is difficult, it just is, it’s how we grow and how we learn. Handing over the difficult parts of something for a generic filler is literally giving up your effort and input for a tech bro insert, you’re surrendering your own personal identity, potential, growth, and expression every time you use any sort of generative AI and letting commercialism fill in the gaps. Don’t be a tool.
Im arguing that these are tools for artists, not to replace them.
If they’re doing the work for you then you are using them as a replacement. It’s hack shit for hacks. You’re a hack, not an artist, you’re being fed slop, not creating art.
Congratulations, you do not understand Duchamp in the slightest.
I’m disproving the statement that art requires effort.
You can’t claim that Duchamp provided the effort behind creating the fountain.
Duchamp was calling out the idiocy in claiming the creation of a toilet did not require (someone’s) skilled labor and therefore did not count as art. You have completely misunderstood the assignment.
But until he signed it, it wasn’t art.
No. When he signed it people payed attention. There’s a difference. It’s still art if no one notices it.
The same argument. AI is still producing art, even if you don’t appreciate it.
Effort doesn’t equate to skill, you still have to out in the physical effort of fucking trying to make something, again, especially in regards to unskilled art what is being appreciated is specifically the effort put into it. Someone tried to express something.
Prompts are not a fucking skill. AI is not a skill, it is not an artistic effort. You made nothing, you had a program procedurally generate colors. If you commission an artist you don’t get to take credit for their work, the same goes for AI color vomit. Art requires expression, intent, and effort, those aren’t things you can procedurally generate nor code.
Writing is a skill. But also inpainting areas with prompts and building different layers of the image with different prompts is certainly a skill.
A marketing skill, sure, it don’t make you an artist and it don’t make you creative or a creator. It’s coloring by numbers, but instead of actually painting anything you are just having printed. You’re a fucking hack, you fucking hack.
Here’s my sci-fi novel:
Who needs implementation when some techbro wants you to think art is all about coming up with an original idea?
The one about the fucking hairdresser, the space hairdresser and the cowboy. The guy, he’s got a tin foil pal and a pedal bin. His father’s a robot and he’s fucking fucked his sister. Lego! They’re all made of fucking Lego.
But why would you?
Because AI can enhance the piece I am trying to create.
Can a painter not use certain colours?
Can a sculptor not use certain tools (chainsaws don’t require enough effort)?
Can a potter not use a wheel?
Is Collage no longer art?
AI is just another tool for the artist to use.
In every situation you mentioned, the person behind the act has to do more than one thing and still need some skill to ensure there’s no fuckup.
Where is your effort in using AI?
There is no effort in using AI, just like there is no effort in using a paintbrush.
Even crows can do so.
There is however effort in using AI to create something good, just like there is effort in using a paintbrush to create something good.
Unless you think everything an AI creates is good?
You say there is effort in using AI to make something good, but failed to mention what that effort is. “Giving the proper input” is just telling someone else to do what you want, which is hardly any effort and needs almost no skill.
No, I think it’s just ok at best. There was no skill or effort involved, someone just told an algorithm to do something and got the result. Some of them even feel proud of “their” achievement, which makes as much sense as a boss claiming credit for their worker’s work.
Think of AI art as a collage piece. The likelihood of a one shot prompt creating a masterpiece is tiny, but taking parts of various prompts and combining them does take human skill and judgement.
Why don’t you consider yourself artistic? Why do you want to make your own art? By using AI you surrender ownership of your work to opaque algorithms and lose so much of what makes you special as someone unique and individual.
What is pictured in my minds eye is not what appears on paper. Nor is my first AI draft but inpainting and gimp get me much closer.
Practice more.
Why don’t you just make your own art. Then you can be an artist. You want to be an artist, but you don’t want to make art, so you don’t actually want to be an artist.
I don’t care how bad it is. Whatever you make is better than the soulless garbage LLMs make. Just start doing it if you actually want to.
not everyone want to put in effort to get a desirable result.
art is a product just like a tire or sheet of paper.
Sure, and you’ll be treated like someone who doesn’t want to spend any money or put any effort into it: useless. Whatever is generated won’t have any life to it because there wasn’t intent with the creation processes; only a statistical algorithm.
Art isn’t just a product. Sure, it can be, but how many people are making comics or memes online and aren’t being paid? There’s a ton. That proves it’s more than just a product, because a product is made to be sold. It can be a product, but it often isn’t, and even when it is that’s only a part of it.
No one is going to pay for something they can just generate themselves, so AI “art” is not even a product. It’s worthless, and a sink of resources to create.
if you aren’t being paid to do it then it’s a hobby and that’s ok, but papermaking is a hobby too and that doesn’t make paper any less of a commodity.
Art is a subjective expression. What is a tire or sheet of paper trying to express?
Also, desirable results require effort, that’s a simple fact, you’d be stupid to think or expect otherwise.
Could be any number of things. An extreme attention to detail, a production process with a political message, an innovative use of materials, etc… Putting artificial limits on which human creations are allowed to be fulfilling and engaging is foolish.
Nothing about the use of AI or any other technology/tool/process precludes effort. A vast majority of pictures taken in the world today are disposal trash, but that doesn’t mean you can’t make meaningful art with your iPhone.
I disagree.
My eyes are rolling out of my skull 🙄
They never claimed to want to be an artist, much less fit your definition of one. A movie producer might “make art” but they just care about the output and not using their job as a creative outlet.
I do. And one of the tools in my box is a diffusion model. Why shouldn’t it be in there?
An A.I.-Generated Picture Won an Art Prize.
You don’t make art then. You ask something else to make it for you, and it will be valueless. If I hire an artist online to create something for me, I don’t create the piece of art. They do. How does that logic differ if that is an LLM or an artist? You aren’t the one with the knowledge or skill, so you aren’t an artist. You’re a hack who wants to pass other’s work off as their own because you’re too lazy or lack the creativity to do it yourself. You’re pitiable, if you weren’t causing harm.
What are you on about? Who are you to tell people what they can and can’t find value in?
You’re making a hell of a leap here.
I’m going to try a different tactic here, in that I don’t expect I can convince you of anything on the other side of the argument that you’re engaging with, but something that might be useful to learn from is some constructive criticism of the works that the AI produces. In doing so I’m going to try to avoid what I think a lot of proponents of your style consider gatekeeping.
The piece that won this competition is technically sound, it has an aesthetic presence, we can agree here. If anything AI is kinda good at doing this because it’s trained to recognize the patterns that lead to this in the data and other artwork that was used to train this.
I’ll also even say that if someone spent 80 hours cumulatively taking 900 photographs of a particular scene to try to select just one that’s of particular quality, that we might even say they put in artistic effort into it. It’s just that that alone, the time and effort spent on curating a work, is not all there is that defines it as artistic. In this other case you keep posting about found art, a urinal that someone spent maybe 5 seconds on is also considered art.
It’s also true though that when that guy described the urinal and why he thought it deserved to be considered art, that he put thought into that. He made the selections he did for a reason.
So lets go back to our award winner. It’s competent, we agree, enough that a judge didn’t even realize it might be AI (I’ll get to that in a second). I’ll even say that the fact that the winner could add in an entire head on one of the characters in the pieces and have it blend well enough with the shading and lighting of the work suggests the guy who won with this piece probably knows at least some degree of traditional art concepts.
We can also see that it’s a blend of styles, the composition looks even renaissance a bit, and clearly also the person who submitted this knew enough about other art forms to recognize that. There remains one major question though: what story is this picture telling? The renaissance images that it’s an homage to are generally based on the moment in some story. A picture of Napoleon on a horse isn’t just Napoleon on a horse, there’s usually indications that he’s maybe returning from some battle. Images of Samson and Delilah focus on the moment she attempts the assassination.
That’s what people are trying to get at here. The artist says that his prompt was to make a mashup of styles (victorian and space themes, which does not seem even remotely like what he ended up with), and then refined it further with the lighting and shading, positioning and elements. I can agree this takes effort, it’s just also this is like saying the movie Wild Wild West should have won critically acclaimed awards for the cinematography. Even where he had intent, it seems like it was overwritten by the very tool he used, and it seems like he didn’t believe enough in his own original vision that he noticed that it wasn’t actually what he asked for.
So lets get to the judge then. I think there would be very real outcry if say, a judge for a oil painting competition was unable to tell the difference between an the works made in the requested medium and a photograph that was submitted for it. I could see an argument, much like for how photography was separated out from other artforms in competition, that the same should be done for AI. In addition, I think that the effort to make the AI emulate existing styles and convince people it was made with previously existing tools actually limits the possibility of what the style can create. Almost a decade ago, people were fascinated by google deep dreams because it was a machine creating a rendition of how the machine “thinks,” and showing all the strange symbolism and logical connections that were programmed into it and how they had developed, in a way very different than anything a human artist might think or do. I think that’s the envelope people and AI should be pushing here, I think it should be a completely different category, I think it should have it’s completely own style.
I also think the current AI programs suck for a variety of ethical reasons I won’t get into, but I don’t rule out that there couldn’t be versions made that lack those problems and aren’t just an gimme to corporations looking to streamline every process on the cheap. But if you want AI to be taken seriously, if you want anything you try to make with it to be taken seriously, these are all thoughts you should probably consider.
A well written and thoughtful response. Thank you. I disagree with nothing.
Current AI Diffusion models are absolutely nothing like thinking and the composition is very likely a mixture of creator choice and training on nicely composed pictures.
It does seem that an AI tool is useful for artists, but certainly everything it spits out isn’t art.
i sentence you to drawing human hands 1000 times as punishment