Do you hate all amateur art, or just when it’s made with ai tools? Does a kid’s drawing, produced in scant seconds and with no training and remarkably little skill hold negative value to you, or is it worth something?
What about art produced with hours or days of effort and a specific goal in mind, but don’t so using primarily ai with perhaps a few finishing touches?
I love it when people get hyper defensive about this for no reason at all. Aesthetically, AI art is obviously better than a child’s scribbles, but the problem is that AI art is pure aesthetic, with no meaning behind it at all, and if you engage with art purely for the aesthetic, then you fundamentally miss the point of it. AI can’t mean anything when it produces art. It just spits out a series of 1s and 0s based on whatever nonsense you shout into it.
It doesn’t matter how many hours you spend working on a piece, if you use AI (Edit to clarify: if you use AI to generate the art in its entirety), then the AI made the art. An AI cannot answer questions about artistic decisions it made, because it made no decisions. It’s worse than tracing—at least an amateur artist can answer why they decided to copy another artist’s work.
Because charitable interpretation is dead, I have to clarify. I’m not saying that there is no valid use case AI generated art, nor am I saying that all human-made art is good. All I’m saying is that human-made art can have meaning behind it, while AI art cannot. It’s incapable of having meaning, so it isn’t really art.
It doesn’t matter how many hours you spend working on a piece, if you use AI, then the AI made the art.
Except that artists can use ai as a tool to make art. Sure, the ai can’t say why that pixel looks that way, but the artist can say why this is the output that was kept. They can tell you why they chose to prompt the ai the way they did, what outputs they expected and why the ones that were kept were special, let alone what changes they may have made after and why.
If Jackson Pollock can make art from randomness by flicking a brush, why can’t someone make art from randomness by promoting an ai? Is there a lone somewhere that makes it become art, in your opinion? I don’t think it would be uncharitable by interpreting the above quote to mean you don’t believe it is possible at all to use ai as a tool in the production of the art.
If ai is the only tool used, it never makes an image, let alone art, because there was never even a human using language to prompt the ai. But from that obviously ridiculous extreme there is certainly a long spectrum ranging through what I described above to something as far removed as a human generating landscapes for a storyboard before fully producing a movie that doesn’t include the air outputs in any physical way. I’m sure you would claim a line exists between there, and I’m curious where.
There’s a couple of orthogonal arguments here, and I’m going to try to address them both: are you an artist if you use AI generated art, and why do I hate AI generated art?
Telling a machine “car, sedan, neon lights, raining, shining asphalt, night time, city lights” is not creating art. To me, it’s equivalent to commissioning art. If I pay someone $25 to draw my D&D character, then I am not an artist, I’ve simply hired one to draw what I wanted to see. Now, if I make any meaningful changes to that artwork, I could be considered an artist. For example, if I commissioned someone else to do the line work, and then I fill in the colors, we’ve both made the artwork. Of course, this can be stretched to an extreme that challenges my descriptivism. If I put a single black pixel on the Mona Lisa, can I say I collaborated on the output? Technically, yes, but I can’t take credit for anything other than putting a black pixel on it. Similarly, I feel that prompt engineers can’t take any credit for the pictures that AI produces past the prompt that they provided and whatever post-processing they do.
As for why I hate AI art, I just hate effortless slop. It’s the exact same thing as YouTube shorts comprised of Family Guy clips and slime. I have a hard time really communicating this feeling to other people, but I know many other people feel the same way. Even aside from the ethical concerns of stealing people’s artwork to train image generators, we live in a capitalist society, and automating things like art generation and youtube shorts uploads harms the people who actually produce those things in the first place.
Telling a machine “car, sedan, neon lights, raining, shining asphalt, night time, city lights” is not creating art. To me, it’s equivalent to commissioning art.
When art is commissioned, art is produced. If no human produced it, an ai did. If ai cannot produce art, then a human must have.
Similarly, I feel that prompt engineers can’t take any credit for the pictures that AI produces past the prompt that they provided and whatever post-processing they do.
I suppose I don’t understand why engineering a prompt can’t count as an artistic skill, nor why selecting from a number of generated outputs can’t (albeit to probably a much lower degree). At what point does a patron making a commission become a collaborator? And if ai fills the role of the painter, why wouldn’t you expect that line to move?
As for why I hate AI art, I just hate effortless slop.
I’m with you there. And I would brook no issue with completing about the massive amount of terrible, low-effort ai art currently being produced. But broadening the claim to include all art in which the most efficacious tool used was ai pushes it over the line for me.
When art is commissioned, art is produced. If no human produced it, an ai did. If ai cannot produce art, then a human must have.
Right, so this is what I mean when I say that charitable interpretation is dead. Taking my earlier assertion that AI generated art isn’t real art, along with my assertion that providing a prompt to an AI is essentially equivalent to providing a description to a human artist for a commission, should not have read as an argument for or against AI generated art being real art. Taking those statements together, the only reasonable conclusion you can make about my position is that prompt engineers aren’t artists.
I suppose I don’t understand why engineering a prompt can’t count as an artistic skill, nor why selecting from a number of generated outputs can’t (albeit to probably a much lower degree). At what point does a patron making a commission become a collaborator?
Never. It’s not an artistic skill in the same way that providing a description to an actual artist is not an artistic skill, which was the point of that paragraph. They become a collaborator the moment they make changes to the work, and the level to which they can say they’re an artist depends on what changes they make, and how well they make them.
Whataboutism and JAQing off. AI models are trained off blatant mass theft; as long as the originators of the training material (1) haven’t given consent to their being scraped and (2) aren’t getting paid for said already-done scraping, then the generator is unethical and deserving of hatred. You can’t have it both ways-- if capitalism is the game that must be played, then the originators of the training data need to give their consent and they need to be paid for every byte of training data that’s been stolen from them.
Hours of effort to create prompts to maneuver the models output until it looks closer to what you wanted, possibly with the addition of touch-up or addition steps at the end likely needed for certain kinds of image to clean up things the ai struggles with (like, say, hands) or to add something in particular the ai didn’t understand (like, say, a monster of your own invention or something).
It’s easy to say that doesn’t count, that the prompt engineer could have just come up with their final prompt in the first place, but then does it count when a digital painter sketches an outline a dozen times before deciding it’s where they want it? After all, the digital artist could have just drawn it the way they wanted at first blush. But I’d bet you’ll say the time the digital artist spent “counts” as time spent working on an art piece, even if you might be inclined to say the prompt engineer’s time doesn’t. I’d be interested to hear your take.
Dude, I don’t care how many iterations a person goes through. I care that the piece contains a bit of their soul.
The argument you’re making fails to appreciate why two images, one made by gen AI, one by a real human person, both exactly identical pixel by pixel, could possibly be valued differently.
If you want to know why I seem to lack respect for the prompt artist who spends a 3-month chunk of their life toiling over their latest piece, making everything just so, because some part of them desperately needs to say something and this piece is the only way they can—I would ask you to show me one.
But further, the prompt artist doesn’t even make it. Even if they did spend the time, credit goes to the AI. The prompt artist is welcome to claim their prompt, I guess, but I don’t often see them sharing those around. Would that even be entertaining?
Dude, I don’t care how many iterations a person goes through. I care that the piece contains a bit of their soul.
the prompt artist who spends a 3-month chunk of their life toiling over their latest piece,
I’m curious what could possibly convince you that someone put their soul into their work? Or why the assumption is always that ai is the only tool being used.
Here’s a list of artists using ai tools in their work.
But further, the prompt artist doesn’t even make it.
Again, ai is a tool. That’s like saying digital artists didn’t make their paintings, the printer did. Or maybe it’s like saying the director didn’t make the movie, the actors and cameras did. Actually, I really like the director analogy. They give directions to the actors as many times as they need to get the take they want, and then they finalize it later with post production.
Yeah sure except now to make deep fake porn you just need to go ‘famous star naked riding an old man’s cock’ set 8 images for each seed and set a job of 100 images, turn the air con to antarctic and make misogynistic videos about why movies are woke while the job slowly cooks your studio
Then when you finish you probably have some good images of whatever famous star you like getting railed by an old man and you can hop on YouTube and complain that people don’t think you are an artist.
It requires almost no effort or talent to make a boatload of deep fake material. If you put any effort in you can orchestrate an image that looks pretty good.
Add to that the fact that before ai, unless you’re already pretty famous, no one cares enough to make nonconsensual porn of you. After, anyone vaguely attracted to you can snap or find a few pictures and do a decent job of it without any skill or practice.
Ease of creation shouldn’t have a bearing on whether or not the final result is illegal. A handmade vs AI generated fake nude should be treated the same way.
I didn’t argue that it shouldn’t. The difference is the ease of creation. It now requires no skill or talent to produce it so the game has changed and it needs to be addressed and not dismissed
Well, the word deep fake is literally from the ai boom, but I understand you to mean doctored images to make it look like someone was doing a porn when they didn’t was already a thing.
And yeah, it very much was. But unless you were already a high profile individual like a popular celebrity, or mayyybe if you happened to be attractive to the one guy making them, they didn’t tend to get made of you, and certainly not well. Now, anyone with a crush and a photo of you can make your face and a pretty decent approximation of your naked body move around and make noises while doing the nasty. And they can do it many orders of magnitude faster and with less skill than before.
So no, you don’t need ai for it to exist and be somewhat problematic, but ai makes it much more problematic.
One ethics quandary is AI child porn. It at least provides a non-harmful outlet for an otherwise harmful act, but it could also feed addictions and feel insufficient.
I don’t get the hate for AI porn.
On its own, it’s just the same as hate for porn. But there’s also deep fake porn, ai porn of real people, and that’s potentially far more problematic.
In my case it’s just the same as hate for AI generated slop
Do you hate all amateur art, or just when it’s made with ai tools? Does a kid’s drawing, produced in scant seconds and with no training and remarkably little skill hold negative value to you, or is it worth something?
What about art produced with hours or days of effort and a specific goal in mind, but don’t so using primarily ai with perhaps a few finishing touches?
I love it when people get hyper defensive about this for no reason at all. Aesthetically, AI art is obviously better than a child’s scribbles, but the problem is that AI art is pure aesthetic, with no meaning behind it at all, and if you engage with art purely for the aesthetic, then you fundamentally miss the point of it. AI can’t mean anything when it produces art. It just spits out a series of 1s and 0s based on whatever nonsense you shout into it.
It doesn’t matter how many hours you spend working on a piece, if you use AI (Edit to clarify: if you use AI to generate the art in its entirety), then the AI made the art. An AI cannot answer questions about artistic decisions it made, because it made no decisions. It’s worse than tracing—at least an amateur artist can answer why they decided to copy another artist’s work.
Because charitable interpretation is dead, I have to clarify. I’m not saying that there is no valid use case AI generated art, nor am I saying that all human-made art is good. All I’m saying is that human-made art can have meaning behind it, while AI art cannot. It’s incapable of having meaning, so it isn’t really art.
Except that artists can use ai as a tool to make art. Sure, the ai can’t say why that pixel looks that way, but the artist can say why this is the output that was kept. They can tell you why they chose to prompt the ai the way they did, what outputs they expected and why the ones that were kept were special, let alone what changes they may have made after and why.
If Jackson Pollock can make art from randomness by flicking a brush, why can’t someone make art from randomness by promoting an ai? Is there a lone somewhere that makes it become art, in your opinion? I don’t think it would be uncharitable by interpreting the above quote to mean you don’t believe it is possible at all to use ai as a tool in the production of the art.
If ai is the only tool used, it never makes an image, let alone art, because there was never even a human using language to prompt the ai. But from that obviously ridiculous extreme there is certainly a long spectrum ranging through what I described above to something as far removed as a human generating landscapes for a storyboard before fully producing a movie that doesn’t include the air outputs in any physical way. I’m sure you would claim a line exists between there, and I’m curious where.
There’s a couple of orthogonal arguments here, and I’m going to try to address them both: are you an artist if you use AI generated art, and why do I hate AI generated art?
Telling a machine “car, sedan, neon lights, raining, shining asphalt, night time, city lights” is not creating art. To me, it’s equivalent to commissioning art. If I pay someone $25 to draw my D&D character, then I am not an artist, I’ve simply hired one to draw what I wanted to see. Now, if I make any meaningful changes to that artwork, I could be considered an artist. For example, if I commissioned someone else to do the line work, and then I fill in the colors, we’ve both made the artwork. Of course, this can be stretched to an extreme that challenges my descriptivism. If I put a single black pixel on the Mona Lisa, can I say I collaborated on the output? Technically, yes, but I can’t take credit for anything other than putting a black pixel on it. Similarly, I feel that prompt engineers can’t take any credit for the pictures that AI produces past the prompt that they provided and whatever post-processing they do.
As for why I hate AI art, I just hate effortless slop. It’s the exact same thing as YouTube shorts comprised of Family Guy clips and slime. I have a hard time really communicating this feeling to other people, but I know many other people feel the same way. Even aside from the ethical concerns of stealing people’s artwork to train image generators, we live in a capitalist society, and automating things like art generation and youtube shorts uploads harms the people who actually produce those things in the first place.
When art is commissioned, art is produced. If no human produced it, an ai did. If ai cannot produce art, then a human must have.
I suppose I don’t understand why engineering a prompt can’t count as an artistic skill, nor why selecting from a number of generated outputs can’t (albeit to probably a much lower degree). At what point does a patron making a commission become a collaborator? And if ai fills the role of the painter, why wouldn’t you expect that line to move?
I’m with you there. And I would brook no issue with completing about the massive amount of terrible, low-effort ai art currently being produced. But broadening the claim to include all art in which the most efficacious tool used was ai pushes it over the line for me.
Right, so this is what I mean when I say that charitable interpretation is dead. Taking my earlier assertion that AI generated art isn’t real art, along with my assertion that providing a prompt to an AI is essentially equivalent to providing a description to a human artist for a commission, should not have read as an argument for or against AI generated art being real art. Taking those statements together, the only reasonable conclusion you can make about my position is that prompt engineers aren’t artists.
Never. It’s not an artistic skill in the same way that providing a description to an actual artist is not an artistic skill, which was the point of that paragraph. They become a collaborator the moment they make changes to the work, and the level to which they can say they’re an artist depends on what changes they make, and how well they make them.
Whataboutism and JAQing off. AI models are trained off blatant mass theft; as long as the originators of the training material (1) haven’t given consent to their being scraped and (2) aren’t getting paid for said already-done scraping, then the generator is unethical and deserving of hatred. You can’t have it both ways-- if capitalism is the game that must be played, then the originators of the training data need to give their consent and they need to be paid for every byte of training data that’s been stolen from them.
No, because amateur art is interesting.
Hours of effort to what, exactly?
Hours of effort to create prompts to maneuver the models output until it looks closer to what you wanted, possibly with the addition of touch-up or addition steps at the end likely needed for certain kinds of image to clean up things the ai struggles with (like, say, hands) or to add something in particular the ai didn’t understand (like, say, a monster of your own invention or something).
It’s easy to say that doesn’t count, that the prompt engineer could have just come up with their final prompt in the first place, but then does it count when a digital painter sketches an outline a dozen times before deciding it’s where they want it? After all, the digital artist could have just drawn it the way they wanted at first blush. But I’d bet you’ll say the time the digital artist spent “counts” as time spent working on an art piece, even if you might be inclined to say the prompt engineer’s time doesn’t. I’d be interested to hear your take.
Dude, I don’t care how many iterations a person goes through. I care that the piece contains a bit of their soul.
The argument you’re making fails to appreciate why two images, one made by gen AI, one by a real human person, both exactly identical pixel by pixel, could possibly be valued differently.
If you want to know why I seem to lack respect for the prompt artist who spends a 3-month chunk of their life toiling over their latest piece, making everything just so, because some part of them desperately needs to say something and this piece is the only way they can—I would ask you to show me one.
But further, the prompt artist doesn’t even make it. Even if they did spend the time, credit goes to the AI. The prompt artist is welcome to claim their prompt, I guess, but I don’t often see them sharing those around. Would that even be entertaining?
I’m curious what could possibly convince you that someone put their soul into their work? Or why the assumption is always that ai is the only tool being used.
Here’s a list of artists using ai tools in their work.
Again, ai is a tool. That’s like saying digital artists didn’t make their paintings, the printer did. Or maybe it’s like saying the director didn’t make the movie, the actors and cameras did. Actually, I really like the director analogy. They give directions to the actors as many times as they need to get the take they want, and then they finalize it later with post production.
When it contains their soul, I already said this.
Yes, it’s very quaint.
Does the director take credit for their actor’s acting, though? Usually, the actors win the award for best acting.
But that’s the same issue of making fakes that we’ve had for 30+ years since digital manipulation became feasible.
Yeah sure except now to make deep fake porn you just need to go ‘famous star naked riding an old man’s cock’ set 8 images for each seed and set a job of 100 images, turn the air con to antarctic and make misogynistic videos about why movies are woke while the job slowly cooks your studio
Then when you finish you probably have some good images of whatever famous star you like getting railed by an old man and you can hop on YouTube and complain that people don’t think you are an artist.
It requires almost no effort or talent to make a boatload of deep fake material. If you put any effort in you can orchestrate an image that looks pretty good.
Add to that the fact that before ai, unless you’re already pretty famous, no one cares enough to make nonconsensual porn of you. After, anyone vaguely attracted to you can snap or find a few pictures and do a decent job of it without any skill or practice.
Ease of creation shouldn’t have a bearing on whether or not the final result is illegal. A handmade vs AI generated fake nude should be treated the same way.
I didn’t argue that it shouldn’t. The difference is the ease of creation. It now requires no skill or talent to produce it so the game has changed and it needs to be addressed and not dismissed
deepfakes predate the ai boom. you don’t need ai for deepfakes
Well, the word deep fake is literally from the ai boom, but I understand you to mean doctored images to make it look like someone was doing a porn when they didn’t was already a thing.
And yeah, it very much was. But unless you were already a high profile individual like a popular celebrity, or mayyybe if you happened to be attractive to the one guy making them, they didn’t tend to get made of you, and certainly not well. Now, anyone with a crush and a photo of you can make your face and a pretty decent approximation of your naked body move around and make noises while doing the nasty. And they can do it many orders of magnitude faster and with less skill than before.
So no, you don’t need ai for it to exist and be somewhat problematic, but ai makes it much more problematic.
One ethics quandary is AI child porn. It at least provides a non-harmful outlet for an otherwise harmful act, but it could also feed addictions and feel insufficient.
You clearly haven’t seen it, nor know anyone affected by it. It’s like 99% noncon shit from people who are too creepy for artists to work with.
EDIT: Sums it up https://www.youtube.com/watch?v=3aS97RKjEdI