I wonder about the implications for recruiting, starting now. I guess there's no way around art tests for these kinds of concept art positions anymore - and lengthy ones at that, designed to weed out the cleverest of the prompt users as well as test the usual skills. Surely terribly time consuming for both sides and probably leading to way less jobs being given to strangers from half the world away with a cool portfolio who you've just found on artstation.
Fingers crossed this sailing on actual artist's coat tails will indeed be put to a stop in the courts like NikhilR suggested. The cat is out of the bag though.
I am not sure where you try to head to with this question. There is none in Photoshop. Photoshop works with mouse input. This is where it differs. Well, it has hotkeys too. But text input is nothing new really. The first text adventures worked with text. And text adventures are also games.
@Tiles I presume the translation of text prompts into images is the primary function being focused on in these discussions. I think one would be mistaken to compare Photoshop as equal to AI generators, is my point.
Ah i see. I never said it is a substitute for Photoshop. Just that it is a graphics software, like photoshop. Of course there are differences. Like there are differences between Max, ZBrush, Substances and Photoshop too. But you can also use AI to manipulate images, not just to create it. And you can use Photoshop to create images, not only manipulate it. Both are tools to create and to manipulate images.
@Tiles Okay. Noting this is a game art forum, I think it's safe to assume a large majority of the members in the discussion here understand that Photoshop, Maya, Zbrush, etc. are graphics software. I think what's central to the topic is the impact of AI's standout feature of rapidly converting text prompts into images.
Sure that will get rid of the people who can't actually draw anything and are now busy googling basic Photoshop tutorials.
But what about those who are half-decent (but below the threshold for talent you'd seek out) and use AI and as such the work of others to boost their own output? Seems like it would be possible to feed a rough sketch or mockup to an AI and have it rerender your image with shapes and brushwork, shading and lighting you yourself couldn't have achieved. Constructing some layers should be possible then (are people actually this organised in concep art though?).
Recruiting shouldn't be affected, save for having to deal with applicants that don't understand that prompting an AI to make art isn't equal to actually knowing how to make digital art. Anyone with half a brain wouldn't apply for a concept art position if all they knew was prompting.
Its possible that they might still get an interview which they may bullshit through and then fail miserably on the job.
Where this gets concerning is commissioned art, where an AI user does not mention that they are using AI art and then lacks the skill to achieve a specific result for a commission, eventually either being sued or ghosting their client for part payment.
Doing ones due diligence is paramount in every case as usual.
Unless larger companies endorse using it as it is, I don't really think it will make a difference where employment is a concern. There is so much saturation already the most its getting is likes and followers on instagram and maybe the sale of the odd print or two.
If fact I don't really know the market for prints, its very niche unless it pertains to fan art and that has its own regulations. Sure the volume of this will increase but its more a hustle than anything if one attempts to monetise it.
"It's completely normal for people to want their piece of mind back.
Next false claim. You cannot want back what you never lost."
- - - - -
Being a contrarian/taking the position of the devil's advocate/being in disagreement is one thing ; showing a pure lack of empathy is another altogether.
@Tiles I get that you like AI, but every single one of your posts in this thread are just you saying "false claim" without giving a good reason why it's a false claim, and ignoring all the thoughtful stuff between the "false claims".
This WHOLE thread basically amounts to: AI uses art made by artists to build its database; many artists don't like this and didn't sign up for it, and would prefer if AI tools STOPPED using the data mined from their work anymore.
That's all the argument is. Not that AI is evil, not that AI can't be art, not that AI is explicitly or definitely illegal within existing laws, nothing like that. Just that we need to take ethics of data mining artists work into consideration, and either get permission or only use stuff that's in the public domain. Because the truth is, while "the images are not used" by the AI, the data was mined, and thus was used, and artists didn't have a chance to say no. Whether or not pixels are explicitly copied is besides the point, although there are some compelling examples of near exact copies of famous works of art generated from keywords.
You're the one going around in circles here most of all.
@Tiles Okay. Noting this is a game art forum, I think it's safe to assume a large majority of the members in the discussion here understand that Photoshop, Maya, Zbrush, etc. are graphics software. I think what's central to the topic is the impact of AI's standout feature of rapidly converting text prompts into images.
Sure. I know. That AI works by text input is indeed what makes it unique. But this kind of user interaction is nothing new as my text adventure example shows. And AI remains a tool to generate and modify graphics. AI is also such a graphics software.
Game art is not longer unaffected by AI. Tools are at their way to create 3d art directly by text prompt. 2D AI art can be used as concept art. Krita, Photoshop, Blender and quite a few other software has it as an addon meanwhile, and allows you to use it for texture generation for example. And i've seen tutorials to convert a Midjourney image to 3d too already.
@Tiles I get that you like AI, but every single one of your posts in this thread are just you saying "false claim" without giving a good reason why it's a false claim, and ignoring all the thoughtful stuff between the "false claims".
Makes me scratch my head. I always provided argument and facts why i think the claim is false.
AI uses art made by artists to build its database;
Again, this is false claim. AI does not store art in its files or databases. It learns from the files. Like a little kid does. It does not use this data. LAION does not ship with AI. It's the training data. And the training data is collected legally.
When yout think something illegal is going on then please report it. This would be the easiest and most effective way to defend AI. But AI is in the wild since months, and no lawsuit in sight. Since there is nothing illegal going on. And just claiming that it does something illegal is simply not enough.
The ethical part is another chapter. And as told, i fully understand where the hate is coming from. One of your images might be used to train your competitor. Wouldn't make me happy neither. And this is the culprit, that AI can make your jobs obsolete. Not that it "mines" artists data.
many artists don't like this and didn't sign up for it, and would prefer if AI tools STOPPED using the data mined from their work anymore.
I long understood this part. But it is not illegal. It is not even unethical imho. AI does nothing else than another artist would do. Unfair maybe, but not unethical.
Diretcly affected are in real just a handful people. Greg Rutkowsky for example, who is used in 90% of all prompts that tries to achieve concept art look. He has on the contrary made it into history for that. And still works as a concept artist.
You're the one going around in circles here most of all.
You force me to. It's the false claims, plain lies (see the image to image result) and the witchhunt against AI that keeps me in the thread. Without me you would still think that AI stores images in its files, that it breaks copyright by copying over image parts, and that the Image to Image example from above was generated instead of modified, to proof what a coypright breaking monster AI is. And when your repeat the false claim then i will repeat to correct the false claim. That's how i am. For good and for bad. Sorry for annoying you.
I know AI, most of you not. I simply have to jump in when false claims are made. Like you would jump in when somebody explains that a mesh is made of bones or a texture is made of tris.
When your assumptions are already wrong, how can then something good and useful come out?
Being a contrarian/taking the position of the devil's advocate/being in disagreement is one thing ; showing a pure lack of empathy is another altogether.
I said it before, it is not the lack of empathy. It is showing common sense. Horse riders have my full empathy. But my common sense tells me that cars are simply better suited for the job. And same goes for AI. Would horses be better for our environment? For sure. Would you want to go back to riding horses, now that cars exists? How far away from your job do you live?
AI raises productivity for sure. I can finally do what was impossible before. I create my videos for my music by myself now. Without stock photo / stock video content, and completely unique. Not because as an traditional artist i am not able to, but because the day has just 24 hours. I cannot spend half a year at a video when i have to visualize a whole album.
"Again, this is false claim. AI does not store art in its files or databases. It learns from the files. LAION does not ship with AI. It's the training data. And the training data is collected legally."
I didn't say it stores art in its files or databases, nor did I say it was illegal; I said "AI uses art made by artists to build its database"; training an AI has to use the data at some point, that's what I'm saying. Not that the images are stored or outright distributed. Just that, at some point, the AI is using the art to become what the AI has become. This is not a false claim, and feels like you shifting the goal posts by using a phrase like "False claim" the same way a famous man once used "Fake News".
I think many of us understand it's not necessarily illegal right now; I think many of us want aspects of it to be fundamentally changed or made illegal, for instance the nonconsensual usage of images for its training data. I understand LAION "is collected legally", but that doesn't mean its approach to collection should be legal for, say, commercial use; and the databases ought to be audited, considering the things that have been found in them.
Once again, I don't think the argument here is "THIS IS ILLEGAL", I think the argument is basically, we should talk about this and maybe think about making aspects of it illegal, or at least way less sketchy (age old statement of the database containing everything from photos of beheadings, to nonconsensually uploading private nude/sexual photos, to medical photos,)— Laws change, approaches to things change, and many artists are clearly unhappy with how this has been approached. The fact that no lawsuit has happened says nothing about whether lawsuits would be justified or whether lawsuits would win. It just says there haven't been any yet (which, by the way, I'm taking you at your word— I haven't checked, and wouldn't know how).
You also say the AI is doing nothing another artist wouldn't do; but again, you miss the point: the AI is not another artist. The AI is a "machine" (I know, I know, it's not literally a machine). You seem overly pragmatic, equating AI and human artists. There's a core and fundamental difference, some aspects measurable, some not.
I understand you'll probably respond to this calling something a false claim; but the truth is I don't even care. The important thing is, artists are talking about an issue that has the potential to affect them, and that's a very good thing. We're hashing out the potential approaches, detailing our current thoughts and understandings, and adjusting them when they don't hold up— or sticking to them when we think they do.
This is undoubtedly my last word here; it's not really worth my time, as like you say, things are going around in circles.
"Horse riders have my full empathy. But my common sense tells me that cars are simply better suited for the job. And same goes for AI."
I will say though, you have made a patently false claim, by saying the AI is better at art than humans. Art is an intrinsically human act; no one and nothing can do it better. Quicker, sure. Easier, sure. But never better.
"Diretcly affected are in real just a handful people. Greg Rutkowsky for example, who is used in 90% of all prompts that tries to achieve concept art look. He has on the contrary made it into history for that. And still works as a concept artist."
You are in no position to claim who is affected and who isn't, or to decide that "being affected" only means losing a job. The fact that English isn't your first language doesn't excuse it.
I don't need excuses, you are not my boss. And i have my own view at the world. But i see it is more than time to end this discussion. The fronts are too hardened.
I look forward to let's say a year from now. And how the world has changed then. I predict that we will still have human artists.
If the AI could learn the way a person did, then why can't it remember what.. Shrek looks like without making hideous morphed monstrosities of him over and over after getting it right a few times. Why doesn't it just spurt out a 3d model of him since it has tons of angles, it should be able to illustrate full turn tables and very specific camera angles because it "understands" what it's drawing. Why can't I zoom in on his skins surface, it is "AI" it knows what skin is. I really cannot understand how people can defend it learning like a person does, even the 3d "ai" look like very basic morph targets on some low quality model library. To me its an extreme glorified combination of content aware fill and an image search combined with the "select all of the images of a crosswalk".
I think these opportunities should go to those deserving of the role. I have genuine concerns about hiring teams being unable to distinguish talent from tools. Partly because these companies deserve to spend their resources well and better position their companies to be successful. Bad actors that infiltrate the hiring process, sully the relationship between employer and employee by deceiving them and taking advantage of their offer. It would encourage an executive mindset of mistrust, which lends to a guarded relationship where owners are less inclined to go beyond for the wellbeing of their employees.
Ironically, these photo-based Captchas (or rather, ReCaptchas) were designed to train the Google AI model for self driving cars, and barely anyone was aware of it when using the service. Sounds familiar ...
Same goes for the earlier version showing lowres house street numbers (also car AI), and the previous one with the text that looked like faulty scans which was designed to train(or help, not sure) their OCR for the Google Books project.
Alright, sorry if this is just morbid curiosity but here is the tone of discussions on the Midjourney general chat (all 3 from today) :
I am just ... genuinely amazed by the absurdity of it all. I find this to be way more violent than any braindead tiktok dance going viral, or any other online stupidity I've ever witnessed. The effect of instant gratification on people is absolutely scary.
I also understand that probably no one here wants to be subjected/be aware of such hatred and envy, so I believe that I won't post any more of that and I should probably stop peaking at that server myself. Still, I find this to be incredibly shocking and worrying to a point where I find it important to at least be aware of what's brewing.
In other news Kickstarter suspended the "Unstable Diffusion campaign", so that's that.
Oh and, High on Life was in such obligation to cut corners (because, well, they just HAD to have a certain amount of detail in their game, no way around it !) that they used MJ for some assets.
Many people supporting AI art really are clueless. Instant gratification is all there is here. Though I can see it being used for some meaningful work, its just where the images were sourced from makes it disturbing to use.
That and the endless saturation, any iota or creativity even in the prompts is totally lost on me.
What about all the stuff that is prohibitively expensive to make right now but could be doable in the future with AI? We might see new avenues of creativity pop up that just weren't feasible before. An example would be how epic sci-fi movies are usually made by giant teams in large studios. Same goes for most hand drawn animated films. Because of the high cost, scripts and subject matter tend to stick with what's a safe bet. A small indie studio might have a great idea, but just not able to fund it. This might lead to new jobs that focus more on working out ideas rather than executing on them.
Well yes, that's what happens every time we invent new things.
Unfortunately what also happens every time is that legislation lags behind a bit and people get screwed over until it catches up. (music sampling in the 1980s is probably a decent parallel to draw)
There are a lot of noisy idiots jumping on the 'robot bad' bandwagon but at it's core the basic complaint around unauthorised use of people's work is I think a valid one. The problem (as always) is that the loud, reactionary voices drown out the more reasoned complainants and do nothing but harm their case.
Today, my son (same infant I posted his birth announcement of here on PC) expressed his sincere desire to continue to draw and become an artist because that's how he feels most connected to others and regulates his own well being. He's got skill and talent by all counts for his age. He could easily continue to develop his skillset and do very well anywhere he wanted to go. How in the actual expletivedeleted do I even begin to nurture his own natural evolution as a human being, knowing what I know with all of my experience in this current AI situation? This cuts to a strong point that Pior has brought up a lot throughout this topic. To what end would another human being even desire to continue searching their own soul to discover more than they dreamed possible through their own creativity when AI is the immediate answer to future capitalist driven generations? How do I tell my son, 'you are genuinely wasting your time'? This is such a screwed up timeline. Peace and love my dudes!
You said it yourself, "his sincere desire to continue to draw and become an artist because that's how he feels most connected to others and regulates his own well being." — he sounds like a true artist, someone who needs to do art for his own human reasons. This should never be discouraged, even if it becomes unlikely as a career. Art has its own value, separate from the capitalist rigamarole.
Thats like showing your son a photograph and saying that Monet ought not to have painted.
There is more to art than the final product. If AI is popular now its more among people that want to cash in on it, even if just an ego boost they are getting from likes on instagram and facebook.
Not sure how this would be monetized at a more corporate level, or even if it gains you commissions to make more art through prompts.
The most it can do is help tie a style (sourced unethically from a real artist) to your brand through generated images that you cement through volume which creates engagement.
Maybe Midjourney might throw a few dollars to support this since it directly benefits them but besides a print or a tshirt, can't think of anywhere the art itself would be worth anything.
That's difficult. I believe a parent knows what words to pass on. More so than a bachelor would anyways.
I would encourage a child to pursue their interests, their dreams. It makes life better. I think it's of benefit to also affirm a sense of reality, of society, the expectations that they ask in order to avoid difficulty in life, penalty, misery, these things. And when met, you've a base, a foundation to stand upon.
This thread has been a pretty good read all around. Sad, funny, ironic, depressing, and uplifting all in one. How fun. Some random thoughts to toss into the fire. Some of this will be rehashed shit in slightly different words, I just feel like writing again:
I think the whole "AI doesn't store data - nothing is stored, so nothing is stolen!" argument is hilarious. How would we define storing data? A piece of technology that holds information in a form that allows it to display, deliver, or have a user access that information at a later time? If AI can "learn" something well enough that it can reproduce it - then it's fucking storing the data. Maybe not as a .jpg, maybe not in a way that can be easily searched in a traditional "files" sense, but it's a technological thing that can retrieve data and allow you to access it (afghan girl being a solid example). That's fucking storage - might be non-traditional storage, might not be wholly reliable storage, but it is storage nonetheless. Next you're going to tell me my brain doesn't store data, because you can't access my memories with a Windows OS.
Another thought - I am just sick to fucking death of the whole "AI is good, it lets everyone express their ideas!" Spoilers, ideas are useless. Dime a dozen. Nobody gives a shit about ideas. Ideas are not impressive, ideas are not uncommon, and a significant majority of ideas are not unique. "Boy wizard with dead parents goes to wizard school, fights bad wizard who is like the baddest bad wizard". "Hot guy with cool voice and silver hair and like, doesn't age fights demons and monsters and stuff and also is in love with powerful magic lady who also basically gets to stay young and hot forever and they run into djinn and witches and elves and wraiths and all the supernatural fantasy stuff". Don't give a shit. The execution of the idea is what matters and being reliant on AI means you're not executing it (or at least, not a substantial part of it) - and consequently, you don't get credit for it.
I think that's something only actual artists can really appreciate - what it means to actually know your idea. To actually know what you want it to look like, or at least to have to consciously make every decision in its creation. Most people don't actually realize how unfamiliar they are with their own idea, their own "design". Until they try to actually create it. Then they realize, "Oh shit, I don't actually know what this guy looks like, in detail. I don't know how tall he is. I don't know where the seams on his clothes are. I don't know how long his arms are compared to his body. I don't know what kinds of buttons are on his shirt. I don't know what the lip of his boots looks like exactly. In what...style, should this be depicted in? What actually makes that style work?"
But once you do know - once you have designed it, once we can see Geralt of Rivia or whoever it is in detail - how that was executed, then it becomes something of value. The execution is the value. The journey of that execution, and how that impacted the final result - that is its identity. When you make it, you own it. When you don't do that - it's not yours, and I don't give a shit about it. You don't get artist points for picking the right prompts and personally selecting the results anymore than I get points when I type some shit into google images. Even if I spent a really long time on google images.
Final thought:
Funnily enough, if you actually know what you want - if you actually have an idea, down to the details - I can't tell that AI's really any faster at all. In the bit of testing I've done - I've never been able to get AI to actually show me what I want. And by the time I get even remotely, sort-of-kinda-eh-I-guess-it's-similar, I could've just done it myself. I've seen many pro-AI folk actually use this exact same argument, "It's really hard to get AI to show you exactly what you want - it took days to get it to generate _______!"
And that's something that actually strengthens my distaste for AI works - because it is so incredibly difficult to get AI to generate EXACTLY what you had in mind (assuming you had something very specific in mind)...
Then, chances are - if they're an AI "artist" - well...we're not actually seeing their designs - what they had in mind, are we? So they want credit for... not executing the work, AND credit for showing us something that's only sort of close to what they had in mind?
Hahaha.
No.
Actual final thought:
It does warm my heart to see the discussion here on Polycount. I feel quite proud to know this community, and to have been so molded by it. I appreciate everyone sharing their thoughts on the subject, even though it's kind of a shitty subject.
"3D software is only for lazy people who need a computer to paint for them, it's not real art because the 'artist' doesn't do anything but push buttons"
Somebody old enough to remember this war? :)
I've never been able to get AI to actually show me what I want.
While these ai generators have opened up a can of worms and some moral gray area...isn't the idea of copyright infringement (as it pertains to commercial enterprises i.e selling reference packs, paid apps, etc..) still black and white?
So if an original piece of concept art is used as direct input for training an ai model to mimic that as output without express permission and compensation...that is outright theft, plain and simple, no?
Anyone else remember a long while ago when Monsanto was able to copyright/patent soybean seeds (yup, seeds!) that prevented farmers from planting offspring seeds without paying hefty license fees? Ok, maybe not exactly the same problem, but similar idea 😅
Obviously i'm no expert on these things, but it seems like the artists that are directly and most affected (like the group from the gofundme above) just want some simple and sensible actions.
I totally agree that the data created from training should be considered exactly the same as directly copying images and applying a filter to them in photoshop. You are still deriving a product from the original work, doesn't matter how convoluted the process is. The fair use doctrine needs to be updated to account for this.
I´m mostly writing here to express my support for the people protesting against AI. Unlike many others I don´t share the optimistic point of view on Artstation though. I don´t think that a platform that behaves like this is reformable and I do believe that there should be consequences. Ideally it would become AI art swamp and become irrelevant eventually.
Personally, I see an hierarchy in art. I will always admire traditional artists for their ability to capture realism without undo button or mirroring. Digital art (painting or sculpting) comes with its advantages + we never run out of resources (canvas, paint, clay - all infinite, no limits, no waste). The tools are different, technique is more forgiving. But it still requires years of practice building your skills to reach quality. Perhaps that´s why none of these new tools turned digital artists into assholes that spit on the skills of masters from the past. Unlike what we can see today from many of the AI "artists". To me, these people are equivalent of the pretentious canvas smearers that "see things that way" (paint your girlfriend or mom and tell her she looks like that to you, see what she tells you) or "capture their deepest feelings" (would you mind capturing the same feeling again for me? I´d love to see if the splashes fit the previous time you "captured" it...) and have the audacity to call themselves artists. These people were always here, they just got their chance to flood digital art as well. And it´s easier than ever, so why wouldn´t these participation trophy winners take this opportunity to outnumber the real champions? For no cost at all.
AI is learning..? Great! Now give the f*cker Lascaux cave paintings and see how it evolves. It´s super fast, so we should see what other path art could have taken in no time. And I´d be really curious to see it. Except it won´t evolve. Because it´s not learning. It´s deriving. Do you want to derive art? Derive from your own. Grow - don´t stand on the shoulders of others and look down on them.
Could AI be useful in certain circumstances? Maybe if a studio/artist used their own artwork to generate some variations for base sketches to build up from. Would that be satisfactory from artistic point of view? Doubt it... Also, real artists could use the AI as a base and do some touch-ups to finalize things. Sort of leveling up from whatever AI gives you, add value to it and thus overshadowing these "google search" artists once again - because they don´t have any painting skill really and once they get beat again, they´ll disappear. Unfortunately, doing things this way would become a new standard, and then there´s probably no going back. Industry will require you to deliver dozens of variations because "that other guy using AI can"... So AI is probably here to stay and become the new standard for speed of work.
And no. My feelings aren´t hurt. I´m not concept artist or digital painter. Sure, my time will come as well, I´m not naive. But please understand that being disgusted by these Synthol artists is not a matter of hurt feelings. People who never tried getting better at something will never understand that path can be the goal. I wouldn´t do 3d art if it was just a random art generator. These people disgust me just as the tiktok garbage and other talentless shameless losers that are celebrated/valued by masses rather than talented musical instrument players, woodcarvers, knitters... none of which I aspire to become, so it´s not necessarily about being involved personally. Call it taste, call it morals... Hell, Leo. I ain´t even embarrassed to use that word - ethics... :D
I must say that I am now genuinely torn between two very conflicting states :
• On one hand the combative spirit is a powerful driver, and the rollercoaster of the last few days/weeks put a lot of incredibly important things in sharp perspective : how one expresses oneself through their craft as imperfect as it may be, and how it is possible to retrace the whole life of a person through the marks they left. This is imho truly beautiful stuff, especially in contrast to incredibly polished "/beautiful woman, /big breasts, /sunset, /4k, /8k, /realistic, /superrealistic, /in the style of xxx, /trending on yyy" AI vomit. Even the small mistakes sometimes made by master craftsmen become suddenly more beautiful than ever in my eyes. Seriously : Frazzetta posters, Bilal and Moebius comics pages, Miyazaki storyboards, and of course Kim Jung Gi madlad murals are real tear jerkers now because of how beautifully imperfect their art is in many ways.
• However ... on the other hand the complete disregard to the love that artists put into their craft (as expressed by the incredible toxicity of AI bros, the deafening silence of Artstation/Epic, and the way the masses splurge on LensaAi (including tech "influencers" who should really know better) is just an incredibly heavy weight to carry and that's really new. I of course understand that people who never had the desire (and from there the opportunity) to be involved in a creative field are unlikely to know the feeling of rolling up their sleeves to tackle a seemingly huge task as a team and eventually getting it done ; so I can't blame them for not knowing that the biggest motivator is knowing that someone out there will appreciate the result. But now that there is an audience absolutely ready to consume (and generate) bot-made content at breakneck speeds thanks to an unethically trained model, I feel like the incentive for artists to work on entertainment properties will take a nosedive. Why put in the hard work, if a massive part of the audience doesn't actually care that someone did it ? This is now impossible to blissfully Ignore.
- - - - -
I also feel like it will be harder than ever for younger artists to gather the strength to self-improve and attempt to reach the heights of the masters they look up to when knowing full well that someone, somewhere is gleefully typing up prompts to generate pictures straight up emulating that one fragile thing they cherish. Of course that doesn't prevent anyone from picking up a pencil, but the mental background noise to filter out is pretty damn massive. This simply wasn't there before and this is very invasive in many ways.
Same for seasoned pros : I just can't imagine Miyazaki gathering up strength to design/direct another movie after witnessing people gleefully generating facsimiles of his art at the click of a button (because of course "in the style of Ghibli" is a popular prompt).
On the plus side, perhaps the sometimes challenging and tedious aspects of "human" art will become something to cherish more than ever. But for it to be even possible, one still needs to put a roof over ones head first ...
- - - - -
Lastly another somewhat related shower thought : Now that AS/Epic crapped the bed, artists from all over the entertainment industry have a permanent grudge against them. How can this possibly be reconciled with the massive market penetration of Unreal Engine in games and now TV and movies as well ? Sure enough there's a irreconcilable personal conflict there for anyone working at a studio commercially tied to UE. Or (sigh ...) producing marketplace assets.
It´s not "hate" when you use common sense to call something what it is. Call us pencil-grabbers, paper scratchers, tablet tip worners (personally I prefer this one, because I have to put a new tip on my wacom anyway...). Find some name you would call us that would be accurate and would make us feel uncomfortable at the same time. If you can´t, maybe THEN you will see what the issue is here.
He is one artist of many. Of many others that share the sentiment and create the masses that create the industry. Go see what´s trending on Artstation for quite some time now. Meet the masses. "AI artists" will be able to replace digital artists in the future, but right now at this point their voice in the "industry" is irrelevant (the only relevance lays in the fact that they are a threat for the future. FUTURE. but at this point, they are incompetent to successfully do the tasks that could be assigned by the industry). Yet nobody tells them to shut up just because they´re not at the grown ups table. I see people concerned not only about their jobs but art as well. I see people speaking their mind on what future might bring with this new trend that breaks the line between consumers and creative artists (regarding a result. not the path, skill or worth of such result). Imagine two children that grow up together. One draws a drawing and the other one prints a clipart picture (granted - he had to search for it. kudos...). Both pictures end up on the same fridge. At some point both kids will just print pictures. And the parents will never understand how come none of their children is the creative being that they thought they raised.
Right now there are kids happy that their picture got on a fridge with no hard work invested and parents that see what this might lead to in the future. And just as parents we´re not cool, we´re not hip and we don´t get the hype of this new trend, but when you get bored of this new toy, there will be a lot of mess left in the room and you won´t be around to clean it for those who actually wanted to achieve something on their own and already invested their lives into it.
AI Artists wouldn´t be able to do anything if it wasn´t for the direct input of the real artists that put their time and sweat in their works. If real artists lose motivation, there will be no new input for their "tools" to generate new stuff. It will all become repetitive because the AI does not evolve. It replicates, derivates, whatever... How is a statement like this hateful to anybody? Grab your camera and go to the theatre get some footage of a movie. Tell them that it´s not "camcording" if you just want to feed the AI. We´ll see how that goes...
It's about the hate that pior spreads here. Not about that he dislikes AI. How shall we discuss with each other when we call us with names? You can dislike or like AI without to become personal for disliking the others opinion. That's at least my opinion.
I don't want to lead the discussion from scratch again since we hopped through the very same loop more than once now in this tread. But copyright in its usual form is simply not touched by AI. This is what makes the matter so complicated. And when it comes to the masses, go have a look how many users the AI systems meanwhile have. And not all of them are clueless children that plays with a toy. That's the smallest part. You need to have a connection to art to make images. No matter with what tool. I for example have nearly 25 years of experience in the field. And the very most people who uses AI that i know are also skilled artists. The dislikers at Artstation is not the voice of everybody, not even of the artists at Artstation. You can see quite a few pro AI comments meanwile. And the very most artists at Artstation simply doesn't care, and continues to post their usual images. The week is over, and as predicted, the protest is getting fewer and fewer.
Hey, at the end of the day a buch of people learned a brand new shiny word that feels smart (LUDDITES ! LUDDITES ! LUDDITES !), so that's that.
@ghost-d : Thanks for sharing your thoughts, you Tablet Tip Worner you :D Even as a mere thought experiment the example of the kids is chilling because it just rings so true.
Perhaps this chasm is just a matter of some poeple being able/used to projecting themselves into the future by logically extrapolating the curve of current events, seing the consequences that instant gratification can have ; and some others living in the moment and the perceived urgency/FOMO of it all, having no choice but to use the tech because [insert justification]. Who knows really.
This new ignore feature is awesome. You should try it out pior. Every time I see you invest your energy and effort into certain replies, I grimace at the thought you're being sidetrack from more fruitful endeavors.
I've been seeing this sort of thing a lot; person shares objectively cool image, gets a compliment, doesn't mention it was AI. A lot of "I made" "I did" "my" language.
I understand liking AI art, but at least be up front about it. He just calls it "digital art".
I appreciate that Verge article, Tiles. It's a good read.
Although I found it odd that while they talk about the amount of "human input" as a key criteria for copyright for images that ai spits out...It completely skips over that same criteria in the following section about the input side/training data.
I sort of wished the author would've continued that train of thought and posed the question:
"What would this ai software be, or be capable to produce without the scrapped data that it is trained on?"
Because one could easily argue that those images represent all or (99%) of the "human input".
And when the article mentions the fact that the groups that are most adversely affected are the ones that don't have the financial resources to launch legal actions...I can't help but feel bad for them.
"What would this ai software be, or be capable to produce without the scrapped data that it is trained on?"
What kind of images can a blind child draw? AI analyzes what it sees. No input means no output in both cases. What you basically would do is to blind the AI by taking away the training data.
It's definitely no easy matter, no matter at which side you are.
I've been seeing this sort of thing a lot; person shares objectively cool image, gets a compliment, doesn't mention it was AI. A lot of "I made" "I did" "my" language.
In a 3d software you can load some premade assets and click a button to render the image. Did you now create and render the image? Or was it the software?
Can somebody who just posed a few assets around claim to be an artist? DAZ for example lives from this. And most Archviz artists are using premade content too. And somebody remembers Poser? In all these cases it is still the artist who decides what he wants to produce and how the outcome should be. It's the artists input that leads to the image.
AI is not different here. When somebody types in a man with a hat, then AI will try to create a man with a hat. It's the artists input that leads to the image.
Replies
@Tiles Since they are equivalent tools, how does one convert a text prompt into images in Photoshop?
I wonder about the implications for recruiting, starting now. I guess there's no way around art tests for these kinds of concept art positions anymore - and lengthy ones at that, designed to weed out the cleverest of the prompt users as well as test the usual skills. Surely terribly time consuming for both sides and probably leading to way less jobs being given to strangers from half the world away with a cool portfolio who you've just found on artstation.
Fingers crossed this sailing on actual artist's coat tails will indeed be put to a stop in the courts like NikhilR suggested. The cat is out of the bag though.
I am not sure where you try to head to with this question. There is none in Photoshop. Photoshop works with mouse input. This is where it differs. Well, it has hotkeys too. But text input is nothing new really. The first text adventures worked with text. And text adventures are also games.
@Tiles I presume the translation of text prompts into images is the primary function being focused on in these discussions. I think one would be mistaken to compare Photoshop as equal to AI generators, is my point.
Ah i see. I never said it is a substitute for Photoshop. Just that it is a graphics software, like photoshop. Of course there are differences. Like there are differences between Max, ZBrush, Substances and Photoshop too. But you can also use AI to manipulate images, not just to create it. And you can use Photoshop to create images, not only manipulate it. Both are tools to create and to manipulate images.
I see a few options outside of art tests to prove work is human-
Having a good work history with work in shipped titles.
Having a time-lapse of a painting.
Having a piece with good layers
@Tiles Okay. Noting this is a game art forum, I think it's safe to assume a large majority of the members in the discussion here understand that Photoshop, Maya, Zbrush, etc. are graphics software. I think what's central to the topic is the impact of AI's standout feature of rapidly converting text prompts into images.
Sure that will get rid of the people who can't actually draw anything and are now busy googling basic Photoshop tutorials.
But what about those who are half-decent (but below the threshold for talent you'd seek out) and use AI and as such the work of others to boost their own output? Seems like it would be possible to feed a rough sketch or mockup to an AI and have it rerender your image with shapes and brushwork, shading and lighting you yourself couldn't have achieved. Constructing some layers should be possible then (are people actually this organised in concep art though?).
Recruiting shouldn't be affected, save for having to deal with applicants that don't understand that prompting an AI to make art isn't equal to actually knowing how to make digital art. Anyone with half a brain wouldn't apply for a concept art position if all they knew was prompting.
Its possible that they might still get an interview which they may bullshit through and then fail miserably on the job.
Where this gets concerning is commissioned art, where an AI user does not mention that they are using AI art and then lacks the skill to achieve a specific result for a commission, eventually either being sued or ghosting their client for part payment.
Doing ones due diligence is paramount in every case as usual.
Unless larger companies endorse using it as it is, I don't really think it will make a difference where employment is a concern. There is so much saturation already the most its getting is likes and followers on instagram and maybe the sale of the odd print or two.
If fact I don't really know the market for prints, its very niche unless it pertains to fan art and that has its own regulations. Sure the volume of this will increase but its more a hustle than anything if one attempts to monetise it.
"It's completely normal for people to want their piece of mind back.
Next false claim. You cannot want back what you never lost."
- - - - -
Being a contrarian/taking the position of the devil's advocate/being in disagreement is one thing ; showing a pure lack of empathy is another altogether.
@Tiles I get that you like AI, but every single one of your posts in this thread are just you saying "false claim" without giving a good reason why it's a false claim, and ignoring all the thoughtful stuff between the "false claims".
This WHOLE thread basically amounts to: AI uses art made by artists to build its database; many artists don't like this and didn't sign up for it, and would prefer if AI tools STOPPED using the data mined from their work anymore.
That's all the argument is. Not that AI is evil, not that AI can't be art, not that AI is explicitly or definitely illegal within existing laws, nothing like that. Just that we need to take ethics of data mining artists work into consideration, and either get permission or only use stuff that's in the public domain. Because the truth is, while "the images are not used" by the AI, the data was mined, and thus was used, and artists didn't have a chance to say no. Whether or not pixels are explicitly copied is besides the point, although there are some compelling examples of near exact copies of famous works of art generated from keywords.
You're the one going around in circles here most of all.
@Tiles Okay. Noting this is a game art forum, I think it's safe to assume a large majority of the members in the discussion here understand that Photoshop, Maya, Zbrush, etc. are graphics software. I think what's central to the topic is the impact of AI's standout feature of rapidly converting text prompts into images.
Sure. I know. That AI works by text input is indeed what makes it unique. But this kind of user interaction is nothing new as my text adventure example shows. And AI remains a tool to generate and modify graphics. AI is also such a graphics software.
Game art is not longer unaffected by AI. Tools are at their way to create 3d art directly by text prompt. 2D AI art can be used as concept art. Krita, Photoshop, Blender and quite a few other software has it as an addon meanwhile, and allows you to use it for texture generation for example. And i've seen tutorials to convert a Midjourney image to 3d too already.
@Tiles I get that you like AI, but every single one of your posts in this thread are just you saying "false claim" without giving a good reason why it's a false claim, and ignoring all the thoughtful stuff between the "false claims".
Makes me scratch my head. I always provided argument and facts why i think the claim is false.
AI uses art made by artists to build its database;
Again, this is false claim. AI does not store art in its files or databases. It learns from the files. Like a little kid does. It does not use this data. LAION does not ship with AI. It's the training data. And the training data is collected legally.
When yout think something illegal is going on then please report it. This would be the easiest and most effective way to defend AI. But AI is in the wild since months, and no lawsuit in sight. Since there is nothing illegal going on. And just claiming that it does something illegal is simply not enough.
The ethical part is another chapter. And as told, i fully understand where the hate is coming from. One of your images might be used to train your competitor. Wouldn't make me happy neither. And this is the culprit, that AI can make your jobs obsolete. Not that it "mines" artists data.
many artists don't like this and didn't sign up for it, and would prefer if AI tools STOPPED using the data mined from their work anymore.
I long understood this part. But it is not illegal. It is not even unethical imho. AI does nothing else than another artist would do. Unfair maybe, but not unethical.
Diretcly affected are in real just a handful people. Greg Rutkowsky for example, who is used in 90% of all prompts that tries to achieve concept art look. He has on the contrary made it into history for that. And still works as a concept artist.
You're the one going around in circles here most of all.
You force me to. It's the false claims, plain lies (see the image to image result) and the witchhunt against AI that keeps me in the thread. Without me you would still think that AI stores images in its files, that it breaks copyright by copying over image parts, and that the Image to Image example from above was generated instead of modified, to proof what a coypright breaking monster AI is. And when your repeat the false claim then i will repeat to correct the false claim. That's how i am. For good and for bad. Sorry for annoying you.
I know AI, most of you not. I simply have to jump in when false claims are made. Like you would jump in when somebody explains that a mesh is made of bones or a texture is made of tris.
When your assumptions are already wrong, how can then something good and useful come out?
Being a contrarian/taking the position of the devil's advocate/being in disagreement is one thing ; showing a pure lack of empathy is another altogether.
I said it before, it is not the lack of empathy. It is showing common sense. Horse riders have my full empathy. But my common sense tells me that cars are simply better suited for the job. And same goes for AI. Would horses be better for our environment? For sure. Would you want to go back to riding horses, now that cars exists? How far away from your job do you live?
AI raises productivity for sure. I can finally do what was impossible before. I create my videos for my music by myself now. Without stock photo / stock video content, and completely unique. Not because as an traditional artist i am not able to, but because the day has just 24 hours. I cannot spend half a year at a video when i have to visualize a whole album.
Ok.
"Again, this is false claim. AI does not store art in its files or databases. It learns from the files. LAION does not ship with AI. It's the training data. And the training data is collected legally."
I didn't say it stores art in its files or databases, nor did I say it was illegal; I said "AI uses art made by artists to build its database"; training an AI has to use the data at some point, that's what I'm saying. Not that the images are stored or outright distributed. Just that, at some point, the AI is using the art to become what the AI has become. This is not a false claim, and feels like you shifting the goal posts by using a phrase like "False claim" the same way a famous man once used "Fake News".
I think many of us understand it's not necessarily illegal right now; I think many of us want aspects of it to be fundamentally changed or made illegal, for instance the nonconsensual usage of images for its training data. I understand LAION "is collected legally", but that doesn't mean its approach to collection should be legal for, say, commercial use; and the databases ought to be audited, considering the things that have been found in them.
Once again, I don't think the argument here is "THIS IS ILLEGAL", I think the argument is basically, we should talk about this and maybe think about making aspects of it illegal, or at least way less sketchy (age old statement of the database containing everything from photos of beheadings, to nonconsensually uploading private nude/sexual photos, to medical photos,)— Laws change, approaches to things change, and many artists are clearly unhappy with how this has been approached. The fact that no lawsuit has happened says nothing about whether lawsuits would be justified or whether lawsuits would win. It just says there haven't been any yet (which, by the way, I'm taking you at your word— I haven't checked, and wouldn't know how).
You also say the AI is doing nothing another artist wouldn't do; but again, you miss the point: the AI is not another artist. The AI is a "machine" (I know, I know, it's not literally a machine). You seem overly pragmatic, equating AI and human artists. There's a core and fundamental difference, some aspects measurable, some not.
I understand you'll probably respond to this calling something a false claim; but the truth is I don't even care. The important thing is, artists are talking about an issue that has the potential to affect them, and that's a very good thing. We're hashing out the potential approaches, detailing our current thoughts and understandings, and adjusting them when they don't hold up— or sticking to them when we think they do.
This is undoubtedly my last word here; it's not really worth my time, as like you say, things are going around in circles.
"Horse riders have my full empathy. But my common sense tells me that cars are simply better suited for the job. And same goes for AI."
I will say though, you have made a patently false claim, by saying the AI is better at art than humans. Art is an intrinsically human act; no one and nothing can do it better. Quicker, sure. Easier, sure. But never better.
I didn't say it stores art in its files or databases, nor did I say it was illegal;
Then i did misunderstand you. Thanks for clearing up.
I will say though, you have made a patently false claim, by saying the AI is better at art than humans.
Huh? I didn't talk about the rider, i talked about the vehicle. AI is the vehicle. It is a tool.
There goes my incognito :'3
"Diretcly affected are in real just a handful people. Greg Rutkowsky for example, who is used in 90% of all prompts that tries to achieve concept art look. He has on the contrary made it into history for that. And still works as a concept artist."
You are in no position to claim who is affected and who isn't, or to decide that "being affected" only means losing a job. The fact that English isn't your first language doesn't excuse it.
I don't need excuses, you are not my boss. And i have my own view at the world. But i see it is more than time to end this discussion. The fronts are too hardened.
I look forward to let's say a year from now. And how the world has changed then. I predict that we will still have human artists.
The fix is physical IRL, art, just keep moving to something new every-time they "upgrade their b.s.".
oh no, don't stop now!
If the AI could learn the way a person did, then why can't it remember what.. Shrek looks like without making hideous morphed monstrosities of him over and over after getting it right a few times. Why doesn't it just spurt out a 3d model of him since it has tons of angles, it should be able to illustrate full turn tables and very specific camera angles because it "understands" what it's drawing. Why can't I zoom in on his skins surface, it is "AI" it knows what skin is. I really cannot understand how people can defend it learning like a person does, even the 3d "ai" look like very basic morph targets on some low quality model library. To me its an extreme glorified combination of content aware fill and an image search combined with the "select all of the images of a crosswalk".
I think these opportunities should go to those deserving of the role. I have genuine concerns about hiring teams being unable to distinguish talent from tools. Partly because these companies deserve to spend their resources well and better position their companies to be successful. Bad actors that infiltrate the hiring process, sully the relationship between employer and employee by deceiving them and taking advantage of their offer. It would encourage an executive mindset of mistrust, which lends to a guarded relationship where owners are less inclined to go beyond for the wellbeing of their employees.
"select all of the images of a crosswalk"
Ironically, these photo-based Captchas (or rather, ReCaptchas) were designed to train the Google AI model for self driving cars, and barely anyone was aware of it when using the service. Sounds familiar ...
Same goes for the earlier version showing lowres house street numbers (also car AI), and the previous one with the text that looked like faulty scans which was designed to train(or help, not sure) their OCR for the Google Books project.
Alright, sorry if this is just morbid curiosity but here is the tone of discussions on the Midjourney general chat (all 3 from today) :
I am just ... genuinely amazed by the absurdity of it all. I find this to be way more violent than any braindead tiktok dance going viral, or any other online stupidity I've ever witnessed. The effect of instant gratification on people is absolutely scary.
I also understand that probably no one here wants to be subjected/be aware of such hatred and envy, so I believe that I won't post any more of that and I should probably stop peaking at that server myself. Still, I find this to be incredibly shocking and worrying to a point where I find it important to at least be aware of what's brewing.
In other news Kickstarter suspended the "Unstable Diffusion campaign", so that's that.
Oh and, High on Life was in such obligation to cut corners (because, well, they just HAD to have a certain amount of detail in their game, no way around it !) that they used MJ for some assets.
https://www.thegamer.com/high-on-life-ai-generated-art/
https://www.vg247.com/high-on-life-uses-controversial-ai-art-and-voice-acting-technology
Many people supporting AI art really are clueless. Instant gratification is all there is here. Though I can see it being used for some meaningful work, its just where the images were sourced from makes it disturbing to use.
That and the endless saturation, any iota or creativity even in the prompts is totally lost on me.
What about all the stuff that is prohibitively expensive to make right now but could be doable in the future with AI? We might see new avenues of creativity pop up that just weren't feasible before. An example would be how epic sci-fi movies are usually made by giant teams in large studios. Same goes for most hand drawn animated films. Because of the high cost, scripts and subject matter tend to stick with what's a safe bet. A small indie studio might have a great idea, but just not able to fund it. This might lead to new jobs that focus more on working out ideas rather than executing on them.
Well yes, that's what happens every time we invent new things.
Unfortunately what also happens every time is that legislation lags behind a bit and people get screwed over until it catches up. (music sampling in the 1980s is probably a decent parallel to draw)
There are a lot of noisy idiots jumping on the 'robot bad' bandwagon but at it's core the basic complaint around unauthorised use of people's work is I think a valid one. The problem (as always) is that the loud, reactionary voices drown out the more reasoned complainants and do nothing but harm their case.
Today, my son (same infant I posted his birth announcement of here on PC) expressed his sincere desire to continue to draw and become an artist because that's how he feels most connected to others and regulates his own well being. He's got skill and talent by all counts for his age. He could easily continue to develop his skillset and do very well anywhere he wanted to go. How in the actual expletivedeleted do I even begin to nurture his own natural evolution as a human being, knowing what I know with all of my experience in this current AI situation? This cuts to a strong point that Pior has brought up a lot throughout this topic. To what end would another human being even desire to continue searching their own soul to discover more than they dreamed possible through their own creativity when AI is the immediate answer to future capitalist driven generations? How do I tell my son, 'you are genuinely wasting your time'? This is such a screwed up timeline. Peace and love my dudes!
You said it yourself, "his sincere desire to continue to draw and become an artist because that's how he feels most connected to others and regulates his own well being." — he sounds like a true artist, someone who needs to do art for his own human reasons. This should never be discouraged, even if it becomes unlikely as a career. Art has its own value, separate from the capitalist rigamarole.
Thats like showing your son a photograph and saying that Monet ought not to have painted.
There is more to art than the final product. If AI is popular now its more among people that want to cash in on it, even if just an ego boost they are getting from likes on instagram and facebook.
Not sure how this would be monetized at a more corporate level, or even if it gains you commissions to make more art through prompts.
The most it can do is help tie a style (sourced unethically from a real artist) to your brand through generated images that you cement through volume which creates engagement.
Maybe Midjourney might throw a few dollars to support this since it directly benefits them but besides a print or a tshirt, can't think of anywhere the art itself would be worth anything.
That's difficult. I believe a parent knows what words to pass on. More so than a bachelor would anyways.
I would encourage a child to pursue their interests, their dreams. It makes life better. I think it's of benefit to also affirm a sense of reality, of society, the expectations that they ask in order to avoid difficulty in life, penalty, misery, these things. And when met, you've a base, a foundation to stand upon.
This thread has been a pretty good read all around. Sad, funny, ironic, depressing, and uplifting all in one. How fun. Some random thoughts to toss into the fire. Some of this will be rehashed shit in slightly different words, I just feel like writing again:
I think the whole "AI doesn't store data - nothing is stored, so nothing is stolen!" argument is hilarious. How would we define storing data? A piece of technology that holds information in a form that allows it to display, deliver, or have a user access that information at a later time? If AI can "learn" something well enough that it can reproduce it - then it's fucking storing the data. Maybe not as a .jpg, maybe not in a way that can be easily searched in a traditional "files" sense, but it's a technological thing that can retrieve data and allow you to access it (afghan girl being a solid example). That's fucking storage - might be non-traditional storage, might not be wholly reliable storage, but it is storage nonetheless. Next you're going to tell me my brain doesn't store data, because you can't access my memories with a Windows OS.
Another thought - I am just sick to fucking death of the whole "AI is good, it lets everyone express their ideas!" Spoilers, ideas are useless. Dime a dozen. Nobody gives a shit about ideas. Ideas are not impressive, ideas are not uncommon, and a significant majority of ideas are not unique. "Boy wizard with dead parents goes to wizard school, fights bad wizard who is like the baddest bad wizard". "Hot guy with cool voice and silver hair and like, doesn't age fights demons and monsters and stuff and also is in love with powerful magic lady who also basically gets to stay young and hot forever and they run into djinn and witches and elves and wraiths and all the supernatural fantasy stuff". Don't give a shit. The execution of the idea is what matters and being reliant on AI means you're not executing it (or at least, not a substantial part of it) - and consequently, you don't get credit for it.
I think that's something only actual artists can really appreciate - what it means to actually know your idea. To actually know what you want it to look like, or at least to have to consciously make every decision in its creation. Most people don't actually realize how unfamiliar they are with their own idea, their own "design". Until they try to actually create it. Then they realize, "Oh shit, I don't actually know what this guy looks like, in detail. I don't know how tall he is. I don't know where the seams on his clothes are. I don't know how long his arms are compared to his body. I don't know what kinds of buttons are on his shirt. I don't know what the lip of his boots looks like exactly. In what...style, should this be depicted in? What actually makes that style work?"
But once you do know - once you have designed it, once we can see Geralt of Rivia or whoever it is in detail - how that was executed, then it becomes something of value. The execution is the value. The journey of that execution, and how that impacted the final result - that is its identity. When you make it, you own it. When you don't do that - it's not yours, and I don't give a shit about it. You don't get artist points for picking the right prompts and personally selecting the results anymore than I get points when I type some shit into google images. Even if I spent a really long time on google images.
Final thought:
Funnily enough, if you actually know what you want - if you actually have an idea, down to the details - I can't tell that AI's really any faster at all. In the bit of testing I've done - I've never been able to get AI to actually show me what I want. And by the time I get even remotely, sort-of-kinda-eh-I-guess-it's-similar, I could've just done it myself. I've seen many pro-AI folk actually use this exact same argument, "It's really hard to get AI to show you exactly what you want - it took days to get it to generate _______!"
And that's something that actually strengthens my distaste for AI works - because it is so incredibly difficult to get AI to generate EXACTLY what you had in mind (assuming you had something very specific in mind)...
Then, chances are - if they're an AI "artist" - well...we're not actually seeing their designs - what they had in mind, are we? So they want credit for... not executing the work, AND credit for showing us something that's only sort of close to what they had in mind?
Hahaha.
No.
Actual final thought:
It does warm my heart to see the discussion here on Polycount. I feel quite proud to know this community, and to have been so molded by it. I appreciate everyone sharing their thoughts on the subject, even though it's kind of a shitty subject.
"3D software is only for lazy people who need a computer to paint for them, it's not real art because the 'artist' doesn't do anything but push buttons"
Somebody old enough to remember this war? :)
I've never been able to get AI to actually show me what I want.
Welcome to the art of text prompt ...
While these ai generators have opened up a can of worms and some moral gray area...isn't the idea of copyright infringement (as it pertains to commercial enterprises i.e selling reference packs, paid apps, etc..) still black and white?
So if an original piece of concept art is used as direct input for training an ai model to mimic that as output without express permission and compensation...that is outright theft, plain and simple, no?
Anyone else remember a long while ago when Monsanto was able to copyright/patent soybean seeds (yup, seeds!) that prevented farmers from planting offspring seeds without paying hefty license fees? Ok, maybe not exactly the same problem, but similar idea 😅
Obviously i'm no expert on these things, but it seems like the artists that are directly and most affected (like the group from the gofundme above) just want some simple and sensible actions.
My take on the situation in my open letters to Artstation:
Part 1: https://www.artstation.com/artwork/r9qDqJ
Part 2: https://www.artstation.com/artwork/Jex8E0
Part 3: https://www.artstation.com/artwork/B3Z5Km (My reply to the reaction from Artstation)
Thanks for reading!
I totally agree that the data created from training should be considered exactly the same as directly copying images and applying a filter to them in photoshop. You are still deriving a product from the original work, doesn't matter how convoluted the process is. The fair use doctrine needs to be updated to account for this.
Remember the "Make Art" button? Ironic, huh..? :D
I´m mostly writing here to express my support for the people protesting against AI. Unlike many others I don´t share the optimistic point of view on Artstation though. I don´t think that a platform that behaves like this is reformable and I do believe that there should be consequences. Ideally it would become AI art swamp and become irrelevant eventually.
Personally, I see an hierarchy in art. I will always admire traditional artists for their ability to capture realism without undo button or mirroring. Digital art (painting or sculpting) comes with its advantages + we never run out of resources (canvas, paint, clay - all infinite, no limits, no waste). The tools are different, technique is more forgiving. But it still requires years of practice building your skills to reach quality. Perhaps that´s why none of these new tools turned digital artists into assholes that spit on the skills of masters from the past. Unlike what we can see today from many of the AI "artists". To me, these people are equivalent of the pretentious canvas smearers that "see things that way" (paint your girlfriend or mom and tell her she looks like that to you, see what she tells you) or "capture their deepest feelings" (would you mind capturing the same feeling again for me? I´d love to see if the splashes fit the previous time you "captured" it...) and have the audacity to call themselves artists. These people were always here, they just got their chance to flood digital art as well. And it´s easier than ever, so why wouldn´t these participation trophy winners take this opportunity to outnumber the real champions? For no cost at all.
AI is learning..? Great! Now give the f*cker Lascaux cave paintings and see how it evolves. It´s super fast, so we should see what other path art could have taken in no time. And I´d be really curious to see it. Except it won´t evolve. Because it´s not learning. It´s deriving. Do you want to derive art? Derive from your own. Grow - don´t stand on the shoulders of others and look down on them.
Could AI be useful in certain circumstances? Maybe if a studio/artist used their own artwork to generate some variations for base sketches to build up from. Would that be satisfactory from artistic point of view? Doubt it... Also, real artists could use the AI as a base and do some touch-ups to finalize things. Sort of leveling up from whatever AI gives you, add value to it and thus overshadowing these "google search" artists once again - because they don´t have any painting skill really and once they get beat again, they´ll disappear. Unfortunately, doing things this way would become a new standard, and then there´s probably no going back. Industry will require you to deliver dozens of variations because "that other guy using AI can"... So AI is probably here to stay and become the new standard for speed of work.
And no. My feelings aren´t hurt. I´m not concept artist or digital painter. Sure, my time will come as well, I´m not naive. But please understand that being disgusted by these Synthol artists is not a matter of hurt feelings. People who never tried getting better at something will never understand that path can be the goal. I wouldn´t do 3d art if it was just a random art generator. These people disgust me just as the tiktok garbage and other talentless shameless losers that are celebrated/valued by masses rather than talented musical instrument players, woodcarvers, knitters... none of which I aspire to become, so it´s not necessarily about being involved personally. Call it taste, call it morals... Hell, Leo. I ain´t even embarrassed to use that word - ethics... :D
I must say that I am now genuinely torn between two very conflicting states :
• On one hand the combative spirit is a powerful driver, and the rollercoaster of the last few days/weeks put a lot of incredibly important things in sharp perspective : how one expresses oneself through their craft as imperfect as it may be, and how it is possible to retrace the whole life of a person through the marks they left. This is imho truly beautiful stuff, especially in contrast to incredibly polished "/beautiful woman, /big breasts, /sunset, /4k, /8k, /realistic, /superrealistic, /in the style of xxx, /trending on yyy" AI vomit. Even the small mistakes sometimes made by master craftsmen become suddenly more beautiful than ever in my eyes. Seriously : Frazzetta posters, Bilal and Moebius comics pages, Miyazaki storyboards, and of course Kim Jung Gi madlad murals are real tear jerkers now because of how beautifully imperfect their art is in many ways.
• However ... on the other hand the complete disregard to the love that artists put into their craft (as expressed by the incredible toxicity of AI bros, the deafening silence of Artstation/Epic, and the way the masses splurge on LensaAi (including tech "influencers" who should really know better) is just an incredibly heavy weight to carry and that's really new. I of course understand that people who never had the desire (and from there the opportunity) to be involved in a creative field are unlikely to know the feeling of rolling up their sleeves to tackle a seemingly huge task as a team and eventually getting it done ; so I can't blame them for not knowing that the biggest motivator is knowing that someone out there will appreciate the result. But now that there is an audience absolutely ready to consume (and generate) bot-made content at breakneck speeds thanks to an unethically trained model, I feel like the incentive for artists to work on entertainment properties will take a nosedive. Why put in the hard work, if a massive part of the audience doesn't actually care that someone did it ? This is now impossible to blissfully Ignore.
- - - - -
I also feel like it will be harder than ever for younger artists to gather the strength to self-improve and attempt to reach the heights of the masters they look up to when knowing full well that someone, somewhere is gleefully typing up prompts to generate pictures straight up emulating that one fragile thing they cherish. Of course that doesn't prevent anyone from picking up a pencil, but the mental background noise to filter out is pretty damn massive. This simply wasn't there before and this is very invasive in many ways.
Same for seasoned pros : I just can't imagine Miyazaki gathering up strength to design/direct another movie after witnessing people gleefully generating facsimiles of his art at the click of a button (because of course "in the style of Ghibli" is a popular prompt).
On the plus side, perhaps the sometimes challenging and tedious aspects of "human" art will become something to cherish more than ever. But for it to be even possible, one still needs to put a roof over ones head first ...
- - - - -
Lastly another somewhat related shower thought : Now that AS/Epic crapped the bed, artists from all over the entertainment industry have a permanent grudge against them. How can this possibly be reconciled with the massive market penetration of Unreal Engine in games and now TV and movies as well ? Sure enough there's a irreconcilable personal conflict there for anyone working at a studio commercially tied to UE. Or (sigh ...) producing marketplace assets.
the incredible toxicity of AI bros
Have you noticed that only you call your "opponents" with names in this thread here? Hate should never replace common sense.
And you are not "the industry". You are just one artist of many. You don't speak for the masses.
It´s not "hate" when you use common sense to call something what it is. Call us pencil-grabbers, paper scratchers, tablet tip worners (personally I prefer this one, because I have to put a new tip on my wacom anyway...). Find some name you would call us that would be accurate and would make us feel uncomfortable at the same time. If you can´t, maybe THEN you will see what the issue is here.
He is one artist of many. Of many others that share the sentiment and create the masses that create the industry. Go see what´s trending on Artstation for quite some time now. Meet the masses. "AI artists" will be able to replace digital artists in the future, but right now at this point their voice in the "industry" is irrelevant (the only relevance lays in the fact that they are a threat for the future. FUTURE. but at this point, they are incompetent to successfully do the tasks that could be assigned by the industry). Yet nobody tells them to shut up just because they´re not at the grown ups table. I see people concerned not only about their jobs but art as well. I see people speaking their mind on what future might bring with this new trend that breaks the line between consumers and creative artists (regarding a result. not the path, skill or worth of such result). Imagine two children that grow up together. One draws a drawing and the other one prints a clipart picture (granted - he had to search for it. kudos...). Both pictures end up on the same fridge. At some point both kids will just print pictures. And the parents will never understand how come none of their children is the creative being that they thought they raised.
Right now there are kids happy that their picture got on a fridge with no hard work invested and parents that see what this might lead to in the future. And just as parents we´re not cool, we´re not hip and we don´t get the hype of this new trend, but when you get bored of this new toy, there will be a lot of mess left in the room and you won´t be around to clean it for those who actually wanted to achieve something on their own and already invested their lives into it.
AI Artists wouldn´t be able to do anything if it wasn´t for the direct input of the real artists that put their time and sweat in their works. If real artists lose motivation, there will be no new input for their "tools" to generate new stuff. It will all become repetitive because the AI does not evolve. It replicates, derivates, whatever... How is a statement like this hateful to anybody? Grab your camera and go to the theatre get some footage of a movie. Tell them that it´s not "camcording" if you just want to feed the AI. We´ll see how that goes...
It's about the hate that pior spreads here. Not about that he dislikes AI. How shall we discuss with each other when we call us with names? You can dislike or like AI without to become personal for disliking the others opinion. That's at least my opinion.
I don't want to lead the discussion from scratch again since we hopped through the very same loop more than once now in this tread. But copyright in its usual form is simply not touched by AI. This is what makes the matter so complicated. And when it comes to the masses, go have a look how many users the AI systems meanwhile have. And not all of them are clueless children that plays with a toy. That's the smallest part. You need to have a connection to art to make images. No matter with what tool. I for example have nearly 25 years of experience in the field. And the very most people who uses AI that i know are also skilled artists. The dislikers at Artstation is not the voice of everybody, not even of the artists at Artstation. You can see quite a few pro AI comments meanwile. And the very most artists at Artstation simply doesn't care, and continues to post their usual images. The week is over, and as predicted, the protest is getting fewer and fewer.
And when it comes to the concerns, copyright and stuff, i found this a very good read. The truth is nobody knows how it will continue and what the outcome will be. Maybe you are right, maybe i am right. Time will tell. https://www.theverge.com/23444685/generative-ai-copyright-infringement-legal-fair-use-training-data
Hey, at the end of the day a buch of people learned a brand new shiny word that feels smart (LUDDITES ! LUDDITES ! LUDDITES !), so that's that.
@ghost-d : Thanks for sharing your thoughts, you Tablet Tip Worner you :D Even as a mere thought experiment the example of the kids is chilling because it just rings so true.
Perhaps this chasm is just a matter of some poeple being able/used to projecting themselves into the future by logically extrapolating the curve of current events, seing the consequences that instant gratification can have ; and some others living in the moment and the perceived urgency/FOMO of it all, having no choice but to use the tech because [insert justification]. Who knows really.
This new ignore feature is awesome. You should try it out pior. Every time I see you invest your energy and effort into certain replies, I grimace at the thought you're being sidetrack from more fruitful endeavors.
Hehe no worries Cory, I am well aware of it. Thanks for the thought though !
Also, this made me laugh, litterally, out loud :
I've been seeing this sort of thing a lot; person shares objectively cool image, gets a compliment, doesn't mention it was AI. A lot of "I made" "I did" "my" language.
I understand liking AI art, but at least be up front about it. He just calls it "digital art".
I still don't get the point of all this.
I mean great if the guy gets an ego boost and some followers, besides posters and tshirts and maybe magnets, what other way is there to monetise this.
And if he gets a commission besides prompting to generate something random, what does his time cost?
Like I'd through a cent for the effort, maybe a dollar if he gets the image right since I can't own it due to copyright issues.
Its a lot of garbage to be honest that does have the potential to bury the art of real artists if allowed on platforms like Artstation.
I appreciate that Verge article, Tiles. It's a good read.
Although I found it odd that while they talk about the amount of "human input" as a key criteria for copyright for images that ai spits out...It completely skips over that same criteria in the following section about the input side/training data.
I sort of wished the author would've continued that train of thought and posed the question:
"What would this ai software be, or be capable to produce without the scrapped data that it is trained on?"
Because one could easily argue that those images represent all or (99%) of the "human input".
And when the article mentions the fact that the groups that are most adversely affected are the ones that don't have the financial resources to launch legal actions...I can't help but feel bad for them.
"What would this ai software be, or be capable to produce without the scrapped data that it is trained on?"
What kind of images can a blind child draw? AI analyzes what it sees. No input means no output in both cases. What you basically would do is to blind the AI by taking away the training data.
It's definitely no easy matter, no matter at which side you are.
I've been seeing this sort of thing a lot; person shares objectively cool image, gets a compliment, doesn't mention it was AI. A lot of "I made" "I did" "my" language.
In a 3d software you can load some premade assets and click a button to render the image. Did you now create and render the image? Or was it the software?
Can somebody who just posed a few assets around claim to be an artist? DAZ for example lives from this. And most Archviz artists are using premade content too. And somebody remembers Poser? In all these cases it is still the artist who decides what he wants to produce and how the outcome should be. It's the artists input that leads to the image.
AI is not different here. When somebody types in a man with a hat, then AI will try to create a man with a hat. It's the artists input that leads to the image.