Really controversial tactic. I am just thinking that it's a great time to launch new a new "artstation" without photos and ai generated things. Or go back to devianart etc.
Marcel Laverdet
maybe I'm totally off the mark, but, well... I'm an illustrator, I work "full time", and I can't see ANY of the images I've made in the last year being manageable on ai; I tried Stable diff and Dalle, with one of my simple, rather old image, and the results just sucked... isn't that a bit too much of a panic?
Marcel Laverdet
just forgot to add that, for example, the billboard I'm working on is going to be delivered in 16000 px; it doesn't seem to be a size available on ai, I suppose...
8857426048265
But you realise how fast this is evolving, give it a week and it'll be 16000px. Ok maybe a month:)
Marcel Laverdet
the AI proceeds, if I understand correctly, by taking images from the web; the best enhancer plug ins allow, actually, a x3, and that's already exciting;( x4 ou x6 is possible, but without professional results usables) going from 512 to 15000 will take maybe more than a month; in a way, if ai is capable of all our phantasms, in creation and format, then it is of course a lost battle; for the moment, if we try to make a non-random image, simply the reproduction of a photorealistic illustration, the result is pitiful;
Translated with www.DeepL.com/Translator (free version)
Hrach Gevorkyan
you alrdy can increase size with "ai" in photoshop, result is pretty solid
info@torquemod.com
Try mid journey. It’s the only ai that actually generates quality images and it’s got to be terrifying for an artist.
I think you need to go and research more on AI image generation and it's use in art and stop trying to lump everything together under just AI. Yes there are people writing some words and pushing a button but there are also a lot of artists using it as part of their workflow. There is low effort AI art just as there is low effort non AI art.There will be more of it of course as it's just much easier to produce low effort AI art.
You might be too young to remember when computers started to become part of the art and design world but people were having the same opinions and arguments back then. Many people considered using computers as cheating or so easy it wasn't art.
I spent 3 years studying traditional graphic desgn methods, we didn't use computers at all, just photography, photocopying, clear plastics and Letraset. I left collage just as computers were becoming prominient in the work place and every studio was switching to using software like Photoshop and Quark Express making a lot of what I had just learnt irrelevant.
3D now can be considerd easy compared to how it was a couple of decades ago too. We all have the equivelent of a Hollywood FX studio at home with an endless supply of free online tuition. Plus people like yourself making free and low cost tools. The same for music. I used to make music back in the early 90s. I had a whole room full of hardware worth 10s of thousands, now I have all that and more on an average laptop, most of which can be also be found for free or low cost.
Without advances in technology most of the people making a living from art and music today wouldn't be.
Through history you will see the same arguments about things like digital cameras from photographers, and analog cameras from painters. Also yes I am an artist and yes I have used AI a lot. The reason people use these kind of arguments is because they are good past examples of big changes in image creation and tools.
Technology advances and workflows change, things become obsolete, like it or not that's just how industry and the world is. New tech doesn't go away so people need to either adapt to it or fall behind. This has happened many times before and it will happen in the future too.
As for Artstation and other platforms I don't think they can or will remove AI art for the reasons I stated. There's no way of knowing just how long someone has spent on it or how much effort was put in. If you want them to follow the rule of removing AI that looks bad or low effort then to be fair they would also need to remove low effort bad non AI art too.
If using AI only involved people writing a few words and pusing a button then I would agree but it doesn't.
"No raw AI generated image is protected by copyright"
This is wrong, it's still yet to be seen if something is covered by copyright because copyright isn't bound by the tools used to create it but by how much human input has been used.
"Currently all popular AI image generators and everyone that’s using them are piggybacking on the backs of artists that haven’t given their consent"
You mean just like every artist alive is piggybacking from every other artist and artist that has come before them? Artist's don't need to give consent because it's called data scraping. No image has been copied anymore than Google copies an image when it data scrapes them. The AI is learning to create images by looking at images, just like a human can only it's far more efficient. It's not storing them in some huge database somewhere or scraping the web everytime someone generates an image.
/rant
Like your work btw, I hope you don't get affected too much by changes in the industry. Check out some of the AI stuff happening and see how you can use it to your advanatge is my advice.
5403654342180
A really thoughtful and helpful comment. Thank you.
dougieladd
Hmmm... you have to remember that "they" only get better because they are trained on copyrighted art. And considering there a numerous class actions being brought against companies producing these, most notably the one from Getty Images suing Stable Diffusion for scraping it's images.... It's a no from me. I'm out.
info@torquemod.com
You know what we should all do? We should find a legal article somewhere on the web outlining that a.i. generated art is not subject to copyright. Then everyone should buy a piece of a.i. Art from art station with a “license” and then every one of us send in a support request with Epic letting them know that we just bought a license that is legally untenable. This might actually make them take notice because I’m pretty sure it would be illegal for them to sell invalid licenses. It would make them look like snake oil salesmen.
6079194267338
I am computer scientist and I can say that this is indeed a real problem. Everything goes so fast and is so sotisfticate that an AI can be trained with all Van Gog works and generate new drawings with the van gogh style. So taking several works from real illustrators an AI can generate new illustrations. So in a few hours, an artist presenting a new and innovative style can be copied with no human skill involved in the process.
Dillon Russo
AI images can be fun and belong on the fridge. Some people have claimed that painters had the same worry about photography back in the day. To make a comparison here would be inaccurate. A better comparison to make would be clay-mation to digital sculpting. The only difference would be that Zbrush takes much more than a string of 6 descriptive words, and anyone previously in clay-mation before the switched had a massive advantage and rightfully so (being able to translate their sculpting skills). Ai is only going to get better and it won't stop at still
2D imaging..... not to mention AI is the "artist" behind these new images... not the people using the AI platforms. So to be able to monetize them is just plagiarism in every way (aside from the couple of words typed to prompt the image itself)
Marcel Laverdet
I understand that only entering a few words and waiting for the AI to "come up with" a great image is distorting the process... ok... but what characterizes for example my work is the total control (well, the most total possible) on the image: for example, I'm doing a poster for a park: two children and two happy adults are in a gondola on rails, we see under them a farm with different animals, a river with rafts, a big statue's head from which comes out a waterfall (and which is precise), and the rails of a roller coaster style attraction. ... all precisely laid out, and, of course, even from a helicopter, impossible to photograph; it's a classic 3d illustrator's job; what would be the words to put in the ai? and I totally agree with the fact that defending the brush against the computer, or the stagecoach against the train, was a useless fight,... ok; still, I don't understand for the moment what is usable in the ai... (for the kind of jobs I do)...
Translated with www.DeepL.com/Translator (free version)
sibience
You should start using AI and then you will realise it involves a lot more than just putting in a few words if you want to create something specific. Most of the AI art you see is from non artists that just let the AI plop something out close to the words they put in and then choose whatever they like the look of. There's a big difference between using AI as an art tool and as a random image generator. Don't fall into the trap that most people unfamiliar with AI are doing right now with confusing the two and just lumping everything under the same heading.
16 comments
Really controversial tactic. I am just thinking that it's a great time to launch new a new "artstation" without photos and ai generated things. Or go back to devianart etc.
maybe I'm totally off the mark, but, well... I'm an illustrator, I work "full time", and I can't see ANY of the images I've made in the last year being manageable on ai; I tried Stable diff and Dalle, with one of my simple, rather old image, and the results just sucked... isn't that a bit too much of a panic?
just forgot to add that, for example, the billboard I'm working on is going to be delivered in 16000 px; it doesn't seem to be a size available on ai, I suppose...
But you realise how fast this is evolving, give it a week and it'll be 16000px. Ok maybe a month:)
the AI proceeds, if I understand correctly, by taking images from the web; the best enhancer plug ins allow, actually, a x3, and that's already exciting;( x4 ou x6 is possible, but without professional results usables) going from 512 to 15000 will take maybe more than a month; in a way, if ai is capable of all our phantasms, in creation and format, then it is of course a lost battle; for the moment, if we try to make a non-random image, simply the reproduction of a photorealistic illustration, the result is pitiful; Translated with www.DeepL.com/Translator (free version)
you alrdy can increase size with "ai" in photoshop, result is pretty solid
Try mid journey. It’s the only ai that actually generates quality images and it’s got to be terrifying for an artist.
I would suggest you watch this video by Royal Skies on Youtube. https://www.youtube.com/watch?v=F-i09Nz96cw
I think you need to go and research more on AI image generation and it's use in art and stop trying to lump everything together under just AI. Yes there are people writing some words and pushing a button but there are also a lot of artists using it as part of their workflow. There is low effort AI art just as there is low effort non AI art.There will be more of it of course as it's just much easier to produce low effort AI art. You might be too young to remember when computers started to become part of the art and design world but people were having the same opinions and arguments back then. Many people considered using computers as cheating or so easy it wasn't art. I spent 3 years studying traditional graphic desgn methods, we didn't use computers at all, just photography, photocopying, clear plastics and Letraset. I left collage just as computers were becoming prominient in the work place and every studio was switching to using software like Photoshop and Quark Express making a lot of what I had just learnt irrelevant. 3D now can be considerd easy compared to how it was a couple of decades ago too. We all have the equivelent of a Hollywood FX studio at home with an endless supply of free online tuition. Plus people like yourself making free and low cost tools. The same for music. I used to make music back in the early 90s. I had a whole room full of hardware worth 10s of thousands, now I have all that and more on an average laptop, most of which can be also be found for free or low cost. Without advances in technology most of the people making a living from art and music today wouldn't be. Through history you will see the same arguments about things like digital cameras from photographers, and analog cameras from painters. Also yes I am an artist and yes I have used AI a lot. The reason people use these kind of arguments is because they are good past examples of big changes in image creation and tools. Technology advances and workflows change, things become obsolete, like it or not that's just how industry and the world is. New tech doesn't go away so people need to either adapt to it or fall behind. This has happened many times before and it will happen in the future too. As for Artstation and other platforms I don't think they can or will remove AI art for the reasons I stated. There's no way of knowing just how long someone has spent on it or how much effort was put in. If you want them to follow the rule of removing AI that looks bad or low effort then to be fair they would also need to remove low effort bad non AI art too. If using AI only involved people writing a few words and pusing a button then I would agree but it doesn't. "No raw AI generated image is protected by copyright" This is wrong, it's still yet to be seen if something is covered by copyright because copyright isn't bound by the tools used to create it but by how much human input has been used. "Currently all popular AI image generators and everyone that’s using them are piggybacking on the backs of artists that haven’t given their consent" You mean just like every artist alive is piggybacking from every other artist and artist that has come before them? Artist's don't need to give consent because it's called data scraping. No image has been copied anymore than Google copies an image when it data scrapes them. The AI is learning to create images by looking at images, just like a human can only it's far more efficient. It's not storing them in some huge database somewhere or scraping the web everytime someone generates an image. /rant Like your work btw, I hope you don't get affected too much by changes in the industry. Check out some of the AI stuff happening and see how you can use it to your advanatge is my advice.
A really thoughtful and helpful comment. Thank you.
Hmmm... you have to remember that "they" only get better because they are trained on copyrighted art. And considering there a numerous class actions being brought against companies producing these, most notably the one from Getty Images suing Stable Diffusion for scraping it's images.... It's a no from me. I'm out.
You know what we should all do? We should find a legal article somewhere on the web outlining that a.i. generated art is not subject to copyright. Then everyone should buy a piece of a.i. Art from art station with a “license” and then every one of us send in a support request with Epic letting them know that we just bought a license that is legally untenable. This might actually make them take notice because I’m pretty sure it would be illegal for them to sell invalid licenses. It would make them look like snake oil salesmen.
I am computer scientist and I can say that this is indeed a real problem. Everything goes so fast and is so sotisfticate that an AI can be trained with all Van Gog works and generate new drawings with the van gogh style. So taking several works from real illustrators an AI can generate new illustrations. So in a few hours, an artist presenting a new and innovative style can be copied with no human skill involved in the process.
AI images can be fun and belong on the fridge. Some people have claimed that painters had the same worry about photography back in the day. To make a comparison here would be inaccurate. A better comparison to make would be clay-mation to digital sculpting. The only difference would be that Zbrush takes much more than a string of 6 descriptive words, and anyone previously in clay-mation before the switched had a massive advantage and rightfully so (being able to translate their sculpting skills). Ai is only going to get better and it won't stop at still 2D imaging..... not to mention AI is the "artist" behind these new images... not the people using the AI platforms. So to be able to monetize them is just plagiarism in every way (aside from the couple of words typed to prompt the image itself)
I understand that only entering a few words and waiting for the AI to "come up with" a great image is distorting the process... ok... but what characterizes for example my work is the total control (well, the most total possible) on the image: for example, I'm doing a poster for a park: two children and two happy adults are in a gondola on rails, we see under them a farm with different animals, a river with rafts, a big statue's head from which comes out a waterfall (and which is precise), and the rails of a roller coaster style attraction. ... all precisely laid out, and, of course, even from a helicopter, impossible to photograph; it's a classic 3d illustrator's job; what would be the words to put in the ai? and I totally agree with the fact that defending the brush against the computer, or the stagecoach against the train, was a useless fight,... ok; still, I don't understand for the moment what is usable in the ai... (for the kind of jobs I do)... Translated with www.DeepL.com/Translator (free version)
You should start using AI and then you will realise it involves a lot more than just putting in a few words if you want to create something specific. Most of the AI art you see is from non artists that just let the AI plop something out close to the words they put in and then choose whatever they like the look of. There's a big difference between using AI as an art tool and as a random image generator. Don't fall into the trap that most people unfamiliar with AI are doing right now with confusing the two and just lumping everything under the same heading.