I have to disagree with you. I don´t think that truth or righteousness is in numbers. See how many morons would support AI just because they don´t need to know sh*t? I wouldn´t call him a moron because of the numbers. Don´t get me wrong - you are right. But for different reasons.
Hey let's keep it civil please. No need for name calling. Otherwise I'll be forced to close the thread. But I'd much rather see polite debate, and keep it open.
Websites can pass the http headers X-Robots-Tag: noai, X-Robots-Tag: noindex , X-Robots-Tag: noimageai and X-Robots-Tag: noimageindex By default img2dataset will ignore images with such headers.
*(This one is an official option for those AI systems;
To disable this behavior and download all images, you may pass --disallowed_header_directives '[]'
tl:dr
Whether going the route of building your own website,or hosting it at big(ger) platforms, it doesn´t matter, you can´t protect
your data against AI with any noai tag since it is at best optional.*
Been had this one on ignore, and by the looks of it rightly so, seems to have stirred up the rest. ANYONE for it is against your well being if this is the only source of income and for many of us it seems it is, the poor argument of you shouldn't put all your eggs in one basket is fighting for the same people that put us in our current situation... The scheme being played out everywhere is, *terminate all sources of income across all networks*. When you realize it then you'll get this post. continued on my sketchbook...
While I personally disagreed with using premade assets in 3D for a beginner I totally understand why it's useful for more experienced 3D artists simply because it saves time filling a scene. I would be lying if I said it wasn't useful to use premade assets and assemble them in a scene, it's working effectively. Whether it's pulled from previous projects or paid for by another artist to be used. Something I don't see is artist's, who have had their work run through a training data set, haven't been compensated for their work in that regard... something far removed from what you're referencing in terms of premade assets. I don't feel the sting of guilt when I load in premade assets I had purchased for a scene because the effort someone else put in has been compensated for it. Can Ai say the same?
The thing about Daz3D that I remember is how I slowly saw it fade out when I was just starting to get into 3D. Probably because you could easily tell a Daz work over something more personalized, but that's just my opinion and even then Daz gave off that low effort vibe to me. However, if people want to use Daz I'm not going to stop them, I just don't care because it doesn't hurt anyone.
While I see where you're coming from, I have much more respect for kitbasher's over prompters... anybody can load in premade assets into a scene and call it a day, doesn't mean it's going to look good. Kitbashing still involves good lighting, presentation, positioning of objects in a scene, what's the story? what's is the purpose of what I'm seeing? I feel like you get that sort of control. What I don't get is how is prompting an image tells me about your creative decisions and why you did the things you did to get there, because if a pretty image was your goal then you didn't have much to say to begin with.
Just my two cents. Sure wish AI could do retopo and UV mapping... that'd be pretty cash money though.
I am not against AI per se, but I am deeply concerned about the way it is being used. The way it is being used also sparks much bigger concerns about the future. The people creating these systems seem to have utopian views about how this will turn out. They promise us a world in which we will never have to work again and we can have everything we want, whenever we want. Ignoring the fact that not everyone thinks this is desirable, and that arguably, a measure of struggle arguably makes someone better, and that cultivating a skill over many years ennobles them, they promise us a theoretical future good, at the expense of current suffering. They argue that everything we enjoy right now was made by sacrificing the livelihoods of people in similar situation such as medieval book copiers, artisans but just because I enjoy technology and some advancements were made by a kind of symbolic human sacrifice, does not mean this this necessarily has to happen every time a technological revolution happens and that we can't do things in a better way! The present is real but the future is theoretical and they are causing real pain right NOW in favor or some murky, theoretical future.
This COULD have been done in a way that does not spit on the work of people who devoted their lives to cultivating a skill! If the utopia is inevitable, then why not do it in such a way that it does not cause tremendous short term suffering to people? Short term can mean anything from 5 to 15 years, time in which huge swaths of people can be in extreme vulnerability. IF you are a good person who wants to help people and humanity, why not use public domain art out of PRINCIPLE and not just because there is a loophole in the law that lets you do it just because? Why not do good all the way instead of waiting for this sacrifice, that no one asked for, to pay off in the future. Your god might accept your sacrifice or be offended by your hubris and presumptuousness! Even if you don't believe in such "silly fairy tales", because you are men of science after all, you can't deny that these silly stories show a pattern of life and that pattern has a habit of punishing hubris! Besides, this so-called, imminent "utopia" seems to have a creepy religious undertone to it, like it's a messiah whose coming is inevitable.
Limiting the model to public domain might slow technological progress in the short term but In terms of history that can mean 5, 15 or 50 years and what does such paltry time mean if the utopia is inevitable? If an artist loses his job and is forced to flip burgers for the next 15 years, how is he supposed to "adapt or die"? The AI can just shit out 5 results in 5 seconds. How can you compete or adapt to that? If you were a monk in the 15th century who lost his job copying bibles, you could still have an illustrious career writing books about theology, and philosophy! OOPS, YOU CAN'T DO THAT ANYMORE! Your papers will just be fuel for the AI! Even IF you invent a new art style (which is a monumental task in itself) your art will just become fuel for the AI. You will never be compensated and no one will EVER be grateful for your work!
All this is done in the name of so-called "progress". You keep hearing the same mantra online: "you can't stop progress". But this word is very murky, ill-defined and problematic. Progress according to whom? What IS progress? And what are we progressing towards? So far, no one has given me a coherent answer of WHAT "progress" is and WHERE we are "progressing" to. All I get are answers that point to some kind of techno-utopia that they just presuppose is good because "progress".
It's sort of like social media. On paper, it sounds like a really good idea. It would bring people together, make friends across the whole globe, and it enables us to have this conversation but if you really look at it, social media has only made people more depressed and more lonely, exactly the opposite of what it was intended to do! I am not making an equivalence here, I am just saying that based on this precedent (and there are many many more), just because something is new, does not mean it is better. Technological progress does not automatically equate to social, spiritual or even human progress!
If you really want to make such extremely radical changes to society, then you need to find a universal definition to what progress is and not ONE of the tech bros seems to be interested in that! These scientists are all like a bunch of sailors on a ship and we are the passengers. They are all very good sailors but they have no idea where the fuck they are going! At best they have a general idea where they are going. But be careful, there might be sirens on the way or you might land on an island of cannibals! I would be extremely cautious of people who promote ideas of so-called utopia, they are capable of doing anything and stepping on anyone to reach their goals, with a smile on their faces too! Just because I like technology and technological progress does not mean I have to accept everything that's coming my way!
I repeat, I am not against AI, and I was actually very excited for this technology two years ago! I saw those upscaled 4k 60FPS videos from the 1900s, I saw AI upscaled textures and sprites to my favorite 90s video games and I thought to myself how these remasters would be better than human made ones because the remaster wouldn't be "tainted" by subjective impression of the artist. The original vision of the game would just be objectively better. I kept thinking how some tedious tasks like UVs, retopo, making non-tileable textures tileable would be automated. However, it seems like things are going in the opposite direction. Instead of automating the tedious boring, jobs, they are taking away the creative, meaningful jobs, leaving us with either scraps or forcing us to flip burgers because for now, constructing a robot that flips burgers instead of a human is more expensive than hiring a human. You can't eat a digital burger!
I am a 3D artist and I have a very good relationship with my employer and my client. I started 3D 5 years ago so the blow won't/wouldn't be so hard for me but this job is what turned my life around and what gave me a measure of confidence. How am I supposed to grow as a human and what am I supposed to aspire to if the AI does everything better than me, on the backs of people who never consented their works to be used in such a manner? I wanted to learn 2D art as well and since I am extremely picky about who I learn from, Steven Zapata's course is exactly what I need but it's going to take at least 5 years to get decent at it and right now AI is more than decent in just 1 year. Who knows how good this will get in 5 years? Why even bother? Even if I do it for free, it's going to be extremely disheartening to see an AI shit out two results in 5 seconds. No one will appreciate the hard work that I put into this. If I want to switch careers (again), don't be fooled that is coming just for art. If you can replace an artist, you can probably replace any other kind of highly skilled, intellectual job like an architect, lawyer or whatever. You'll just be flipping burgers until the so-called utopia comes.
A year ago I felt like I have the whole world ahead of me but now it seems like I need to plan an exit strategy. I can't even get away from it. I need to hone my craft and I am subscribed to many pages and at least half of them share AI stuff on my feed and the anxiety just starts again. But don't forget, it's all in the name of "PROGRESS"!!!
First off, it shows that believing that "it will never get better than humans" is beyond pointless really. It is (obviously) lacking the personal touch of an artist in terms of their own references and personal visual library, personal back story and taste ... since the machine is using an immensely bigger set of references and visual libraries (equivalent to millions of combined artists), and is relying on a sense of taste perfectly weighted by popularity. But in terms of execution it is absolutely equivalent to high-end production art.
Secondly and most importantly it IMHO 100% shows that these ML models would have not been able to create these outputs had they not been using copyrighted work without consent - since public domain datasets would have never included modern graphic design and layout styles, digital rendering techniques only widespread since the 2000s and available as highres images, and themes/prompts like "character turnaround sheet" and "cyberpunk".
Now all that said I have no doubt that such extra layers of training could have very well been added on top of an "ethically trained" model for sure, by injecting more recent imagery (with properly acquired rights) on top - but that's kindof besides the point for now.
(On a side note, the fact that this person-typing-words-to-get-AI-art finds the first batch of images to be "not great" says a lot about how the tech is already devaluing art ...)
- - - - -
All this leads me to believe that now that such tools are out there, and that there is a high chance that they'll keep being used (regardless of how regulation around scraping evolves), perhaps the only ways for illustrators and concept artists to keep being relevant is to :
- Develop a strong counter-culture always on step ahead of AI tools (if that is even possible).
- Seek longer-term contracts with studios that value human craft and can afford the cost of human labor (as, after all, concept art and promotional art still remains a very small fraction of production costs even if made by a slow meat-person).
- Perhaps shifting to a "post internet" way of looking at things, trying to find means of living away from the online crazyness of AI-vomit.
As far as I am concerned the hardest thing to do is indeed to manage/ignore this online craziness. It's like being in an abusive relationship : both the tech and the output are downright amazing, but the toxicity of it all really puts a huge weight on one's mind and ability to focus, and does hinder ones ability to deal with the art block knowing full-well that the tech is a threat and it only does what it does because it's been fed art without consent. It's infuriating.
Perhaps artists will need to find/build their own safe spaces (yuck) away from anything AI-related to be able to focus on their work. Maybe the key is to focus on team projects between like-minded individuals. After all, this guy developping his own Cyberpuk game by itself will not experience the pleasure of bouncing ideas with someone else, building a shared universe, and so on.
As a slow-meat-fellow traditional portraitist via another life, I'm utterly gobsmacked in just a brief period of time how far this algorithm contagion has spread.
Personally, I don't see those scraped 'remastered' pretty online pic's as visual creativity in the truest sense but rather nothing other than two bit knockoffs carpet bagged by ai-bros + cronies alike - so kinda needed a shower after seeing this
Which makes one both terribly angry and sad, all at once.
I CAN'T! I have to keep up with all the 3D program, addon, sculpting and texture painting developments because this is a high tech field! They often post AI news. I literally can't get away from all this. Whenever I manage to catch a break, news about AI creeps in and the anxiety comes back.
I SUPPOSE the saving grace for me is that I've only been doing this for 5 years, 6 this November. I imagine this is much worse to people who've been doing this for 10 or even 20+ years. Still, this occupation has been so good for me. It's literally changed my life. 5 years is nothing and I barely got started.
I wanted to learn everything pertaining video game art, including 2D. But now I ask myself what's the point? I suppose I should just refine my hard surface skills and hope I still have time to make enough money to soften the inevitable blow because the way I see things, every intellectual job out there is at risk. Architecture, Law, medical research, accounting, you name it. EVEN FUCKING PROGRAMMING! I have a really smart 150 IQ friend that works for Google and he is just as scared as I am. His colleagues too. All that's gonna be left are the lower skill jobs that are not really worth automating right now, like janitor work. I suppose the only job worth doing will be growing food. In short, we're going to be peasants at the mercy of the AI and whoever its masters are.
Oh, I don't think there's reason to be quite so far on the doom-and-gloom side of the fence. And let's face it - pursuing art has always involved filtering out the noise. Trying to figure out if the response you got is from someone trying to pad their post count or an actual artisan who for some reason was kind enough to take the time, no critiques and going from place to place searching for it, getting blindsided by some weird inflationist fetish shit on deviant art...
...this is just... another kind of noise.
It is disgusting to see. I do believe this is ultimately harmful to the world, to humanity, and to any semblance of sustainable happiness for generations to come. But it's not an outright death sentence to anyone with a functioning brain who wishes to use it. To an extent, the same reasons I think this technology is disappointing, and harmful are the same reasons I'm not concerned about it ultimately impacting my livelihood.
When you simplify things, when you devalue things, when everything is "accessible" - people get stupid. A lot of people are stupid. It is really hard to find good candidates for jobs - people who are able to communicate, follow or write a spec/brief, people who...know how to read and write in their native tongue. The world being eager to dumb themselves down to the point of being NPCs just makes the actual story characters all the more valuable.
If anything, @CyberdemoN_1542, I'd say this is more reason to also learn the 2D side of the fence - to learn the other elements of game art. There will be jobs, and if you're the equivalent of 3 separate trained AI programs and aren't a total tool - you're top pick. I am an adequate artist but most of my value as an artist actually comes from problem solving skills and other random shit I was dumb enough to burden my brain with lol
And you know, as depressed as I can get the more I see from AI development - it isn't without its humor. People think that because they can now (or they expect eventually) generate "art", that because they don't have to hire X, Y, Z that they can finally see their brilliant, totally-one-of-a-kind-not-something-that-can-only-exist-because-its-been-done-100-times idea realized without working with actual craftspeople - that they'll be able to make bank! People will see how great their ideas are, now that anyone with "passion" can make a game!
Except that eventuality would mean so could I so why the fuck would I buy from the lazy piece of shit who prompted his idea into existence lol
FYI, that toggle has been there for many years, and was added when Adobe first starting diving into AI, long before the current generative craze. I can't go into much detail because an official response will be published soon, but I am proud to say that Adobe's AI efforts have, are and will be "clean", the data we use to train our models is either proprietary, public domain or licensed (also why it takes longer than simply slapping a web UI on top of LAION).
Well ... it being there for years, and (I believe) turned on by default without any explicit warning to the users is still the opposite of a good look.
As a matter of fact, isn't it pretty much a straight up violation of GDPR ?
The description on that dialog is poorly worded and contains few details, but having an insider view on what was analyzed and for what purpose, there is nothing shady going on. This is meant to allow the apps to process things such as face recognition for auto-tagging your photos or for smart object selection.
Well, to me (and many), the very existence of such an option turned on by default and without warning is the definition of shady - regardless of what is done with it. But indeed, we'll see what the official statement says (I am very curious to see if they'll downplay the issue, or apologize).
Things being turned on by default when installing software is definitely a dark pattern. The fact that one has to read everything like a hawk is just wrong, but here we are.
also 1-button-press = game-ready-retopo when? :D I would submit data for that
"There is no copyright violation at any point. Proof: the AI is now in the wild for three months and more. And there is not a single lawsuit. Now guess why :)"
even if this succeeds in the USA, which sounds like an uphill battle to me - other countries are not obligated to do anything. If it does go through, I wonder if Artstation will adopt Youtube's DMCA filing system where it's used as a weapon against other content creators... lol
Losing a market like the US probably soon to be followed by the E.U. isn't something that will be taken lightly. China is also already regulating, for different reasons, but still. 3 huge markets
How are recruiters in general getting on with this anyway? Last I heard it was along the lines of, "AI makes sorting through viable candidates more difficult."
Well, it's not so much a matter of candidates sneaking in anything, and more of a matter of AI bros flooding companies with applications because they're so "10x" compared to artists ...
well, if you're not going to embrace the power of gpt3 it does yeah
wrt recruitment - I've not seen it yet but I'm not hiring concept artists. I imagine it's going to be a huge pain in the arse when hiring programmers in fashionable fields (javascript and it's spawn) cos the market is already flooded with awful copy&paste developers and this is just going to make it easier for the incompetent to appear competent.
The sad thing is that the only real solution to this from a recruitment perspective is tests with feedback rounds
this really doesnt look all that great, but it is interesting for sure, the 3d mesh only looks like a throw away mesh that is only used to get indications for positions of things, but each frame seems generated on its own independent of the geometry and the connected frames. otherwise it would probably look more stable.
not really because that creates an actual metahuman with adapted premade hair etc, look at the camera turns how stuff is flickering, its not a mesh that gets turned, to me it feels more like ai is filling each frame with what it "imagines" what this hair shape would look like using the mesh as a positioning helper
automatic avatar generation is interesting, but who wants to look like a AI generated potato outside of a corporate hell environment? I think more people would prefer to commission their own avatars or make their own. There's actually a decent market for this already happening in VR and Vtuber avatars
It's both : there is a huge market/desire for people to be able to take a picture of themselves and have their Madden avatar look just like them. As a matter of fact these things have been attempted before with crude photoslapping.
And then there's indeed the fantasy/anime avatars for VRchat and Vtube - but I can guarantee that the market for handcrafting these (as is currently done with Vroid Studio + manual touchups in Blender, all quite tediously) will straight up vanish as soon as a way to generate these avatars with a prompt will exist. People paying 3d artists 20USD for a custom Vroid are not doing it out of an altruistic desire to support artists - they pay because doing it is really quite tedious to do, and requires some technical skills.
The thing I find fascinating about this collage presentation ... is just how boring and uninteresting it is. And by that I don't mean the images themselves, which are fine ; but the whole process involving AI and successive "revisions", which evokes no curiosity or wonder or sense of satisfying hard work whatsoever. Not to mention that the original picture seems to be the most interesting by far anyways ! What a waste.
The tech was pretty much still in the realm of sci-fi just a few months ago ... and now it's already the most tired an unoriginal thing one can possibly think of. Funny how things go.
The layout is much more interesting and has more personality than any of the iterations. It's also weird they didn't change the colors or look much after generation, and this looks like something that might get better results if you could tag surfaces like the Nvidia generation tools.
As you can see, Shutterstock and Nvidia are going to collaborate to create an artificial intelligence that can generate 3D models from text. This is an innovation that could facilitate the creation and customization of visual content for different industries and applications. According to Shutterstock, the AI-generated models will have an appropriate level of detail. The models will be available on Shutterstock's website and will also be offered on Turbosquid.com and Omniverse.
Do you think this is going to be a blessing or a nightmare that will lead us to the unemployment line? In my opinion, this technology has great potential to democratize access to 3D and stimulate creativity, but it also raises ethical and legal challenges about the use of artists' assets to train AIs and the impact on the 3D labor market. I believe it is important to have a debate on these issues and seek responsible and fair solutions for all involved.
If you could currently be replaced by an anonymous outsource artist then yes, the robot will take your job and it'll probably happen in the next couple of years.
In practice I strongly doubt that anyone currently employed will be directly replaced by a robot . Instead what will probably happen is that studios will rely slightly less on mass outsourcing and we'll probably also see a reduction in the number of junior artists hired as the technology improves .
This is kind of inevitable in any industry - eg. there's not a lot of demand for people who can write html any more.
An informed and well-reasoned article on Lawfare about the copyright implications, and an argument for granting copyright only to human-generated works.
If we allow computers to make new copyrights, we should expect an AI-generated version of every melody and chord change possible to be authored and copyrighted, followed immediately by automated lawsuits to defend them. Every new hit will be followed immediately by a spate of lawsuits showing prior art. It could be the end of human creativity as we know it, and the rise of the AI copyright troll.
Indeed, an informative articulately written piece.
There are a lot of trade-offs, and no easy answers. Artists and creators will be harmed economically, and it’s nearly certain that industries will be disrupted or even destroyed, although perhaps it’s too early to say which ones and how soon. It’s possible that the solution is right in front of our faces:Governments should restrict copyright to humans, and deny it to computer-generated works.
Although therein, a likely stone in the shoe!
For instance, when viewed through a fractured geopolitical lens aka: G7 - G20 and/or OECD vista, trying to formulate a 'workable' broad consensus would I fear be akin too herding cats.....in a timely manner.
Replies
I have to disagree with you. I don´t think that truth or righteousness is in numbers. See how many morons would support AI just because they don´t need to know sh*t? I wouldn´t call him a moron because of the numbers. Don´t get me wrong - you are right. But for different reasons.
An important video showing the harassment directed at an artist once AI-babies got their hands on the xeroxing tech :
Hey let's keep it civil please. No need for name calling. Otherwise I'll be forced to close the thread. But I'd much rather see polite debate, and keep it open.
https://polycount.com/discussion/63361/information-about-polycount-new-member-introductions/p1#respect
I do feel that posting here on polycount and also making a mention on LinkedIn likely works better than artstation if you are looking for employment.
I understand that recruiters do browse artststion but every recruiter I've spoken to found my artstation profile from LinkedIn.
Wow, have some empathy for people and respect for CC-0 man.
Regarding building and hosting your own website(s), the "noai" system isn´t to be implemented in the robot.txt file but
in the html header of your own website(s).
https://github.com/rom1504/img2dataset#opt-out-directives
Websites can pass the http headers X-Robots-Tag: noai, X-Robots-Tag: noindex , X-Robots-Tag: noimageai and X-Robots-Tag: noimageindex By default img2dataset will ignore images with such headers.
*(This one is an official option for those AI systems;
To disable this behavior and download all images, you may pass --disallowed_header_directives '[]'
tl:dr
Whether going the route of building your own website,or hosting it at big(ger) platforms, it doesn´t matter, you can´t protect
your data against AI with any noai tag since it is at best optional.*
Been had this one on ignore, and by the looks of it rightly so, seems to have stirred up the rest. ANYONE for it is against your well being if this is the only source of income and for many of us it seems it is, the poor argument of you shouldn't put all your eggs in one basket is fighting for the same people that put us in our current situation... The scheme being played out everywhere is, *terminate all sources of income across all networks*. When you realize it then you'll get this post. continued on my sketchbook...
AI "prompter" overinflating typing time to make the result sound harder to generate than it is :
https://www.reddit.com/r/StableDiffusion/comments/ztddoi/this_started_as_a_complete_accident_took_8_hours/
Wouldn't it be far-fetched to imagine Japan banning this tech - or rather, culturally rejecting it ?
While I personally disagreed with using premade assets in 3D for a beginner I totally understand why it's useful for more experienced 3D artists simply because it saves time filling a scene. I would be lying if I said it wasn't useful to use premade assets and assemble them in a scene, it's working effectively. Whether it's pulled from previous projects or paid for by another artist to be used. Something I don't see is artist's, who have had their work run through a training data set, haven't been compensated for their work in that regard... something far removed from what you're referencing in terms of premade assets. I don't feel the sting of guilt when I load in premade assets I had purchased for a scene because the effort someone else put in has been compensated for it. Can Ai say the same?
The thing about Daz3D that I remember is how I slowly saw it fade out when I was just starting to get into 3D. Probably because you could easily tell a Daz work over something more personalized, but that's just my opinion and even then Daz gave off that low effort vibe to me. However, if people want to use Daz I'm not going to stop them, I just don't care because it doesn't hurt anyone.
While I see where you're coming from, I have much more respect for kitbasher's over prompters... anybody can load in premade assets into a scene and call it a day, doesn't mean it's going to look good. Kitbashing still involves good lighting, presentation, positioning of objects in a scene, what's the story? what's is the purpose of what I'm seeing? I feel like you get that sort of control. What I don't get is how is prompting an image tells me about your creative decisions and why you did the things you did to get there, because if a pretty image was your goal then you didn't have much to say to begin with.
Just my two cents. Sure wish AI could do retopo and UV mapping... that'd be pretty cash money though.
Here's the Italian/European equivalent to the American CAA campaign :
https://www.gofundme.com/f/help-protect-our-art-and-data-from-ai-companies
I am not against AI per se, but I am deeply concerned about the way it is being used. The way it is being used also sparks much bigger concerns about the future. The people creating these systems seem to have utopian views about how this will turn out. They promise us a world in which we will never have to work again and we can have everything we want, whenever we want. Ignoring the fact that not everyone thinks this is desirable, and that arguably, a measure of struggle arguably makes someone better, and that cultivating a skill over many years ennobles them, they promise us a theoretical future good, at the expense of current suffering. They argue that everything we enjoy right now was made by sacrificing the livelihoods of people in similar situation such as medieval book copiers, artisans but just because I enjoy technology and some advancements were made by a kind of symbolic human sacrifice, does not mean this this necessarily has to happen every time a technological revolution happens and that we can't do things in a better way! The present is real but the future is theoretical and they are causing real pain right NOW in favor or some murky, theoretical future.
This COULD have been done in a way that does not spit on the work of people who devoted their lives to cultivating a skill! If the utopia is inevitable, then why not do it in such a way that it does not cause tremendous short term suffering to people? Short term can mean anything from 5 to 15 years, time in which huge swaths of people can be in extreme vulnerability. IF you are a good person who wants to help people and humanity, why not use public domain art out of PRINCIPLE and not just because there is a loophole in the law that lets you do it just because? Why not do good all the way instead of waiting for this sacrifice, that no one asked for, to pay off in the future. Your god might accept your sacrifice or be offended by your hubris and presumptuousness! Even if you don't believe in such "silly fairy tales", because you are men of science after all, you can't deny that these silly stories show a pattern of life and that pattern has a habit of punishing hubris! Besides, this so-called, imminent "utopia" seems to have a creepy religious undertone to it, like it's a messiah whose coming is inevitable.
Limiting the model to public domain might slow technological progress in the short term but In terms of history that can mean 5, 15 or 50 years and what does such paltry time mean if the utopia is inevitable? If an artist loses his job and is forced to flip burgers for the next 15 years, how is he supposed to "adapt or die"? The AI can just shit out 5 results in 5 seconds. How can you compete or adapt to that? If you were a monk in the 15th century who lost his job copying bibles, you could still have an illustrious career writing books about theology, and philosophy! OOPS, YOU CAN'T DO THAT ANYMORE! Your papers will just be fuel for the AI! Even IF you invent a new art style (which is a monumental task in itself) your art will just become fuel for the AI. You will never be compensated and no one will EVER be grateful for your work!
All this is done in the name of so-called "progress". You keep hearing the same mantra online: "you can't stop progress". But this word is very murky, ill-defined and problematic. Progress according to whom? What IS progress? And what are we progressing towards? So far, no one has given me a coherent answer of WHAT "progress" is and WHERE we are "progressing" to. All I get are answers that point to some kind of techno-utopia that they just presuppose is good because "progress".
It's sort of like social media. On paper, it sounds like a really good idea. It would bring people together, make friends across the whole globe, and it enables us to have this conversation but if you really look at it, social media has only made people more depressed and more lonely, exactly the opposite of what it was intended to do! I am not making an equivalence here, I am just saying that based on this precedent (and there are many many more), just because something is new, does not mean it is better. Technological progress does not automatically equate to social, spiritual or even human progress!
If you really want to make such extremely radical changes to society, then you need to find a universal definition to what progress is and not ONE of the tech bros seems to be interested in that! These scientists are all like a bunch of sailors on a ship and we are the passengers. They are all very good sailors but they have no idea where the fuck they are going! At best they have a general idea where they are going. But be careful, there might be sirens on the way or you might land on an island of cannibals! I would be extremely cautious of people who promote ideas of so-called utopia, they are capable of doing anything and stepping on anyone to reach their goals, with a smile on their faces too! Just because I like technology and technological progress does not mean I have to accept everything that's coming my way!
I repeat, I am not against AI, and I was actually very excited for this technology two years ago! I saw those upscaled 4k 60FPS videos from the 1900s, I saw AI upscaled textures and sprites to my favorite 90s video games and I thought to myself how these remasters would be better than human made ones because the remaster wouldn't be "tainted" by subjective impression of the artist. The original vision of the game would just be objectively better. I kept thinking how some tedious tasks like UVs, retopo, making non-tileable textures tileable would be automated. However, it seems like things are going in the opposite direction. Instead of automating the tedious boring, jobs, they are taking away the creative, meaningful jobs, leaving us with either scraps or forcing us to flip burgers because for now, constructing a robot that flips burgers instead of a human is more expensive than hiring a human. You can't eat a digital burger!
I am a 3D artist and I have a very good relationship with my employer and my client. I started 3D 5 years ago so the blow won't/wouldn't be so hard for me but this job is what turned my life around and what gave me a measure of confidence. How am I supposed to grow as a human and what am I supposed to aspire to if the AI does everything better than me, on the backs of people who never consented their works to be used in such a manner? I wanted to learn 2D art as well and since I am extremely picky about who I learn from, Steven Zapata's course is exactly what I need but it's going to take at least 5 years to get decent at it and right now AI is more than decent in just 1 year. Who knows how good this will get in 5 years? Why even bother? Even if I do it for free, it's going to be extremely disheartening to see an AI shit out two results in 5 seconds. No one will appreciate the hard work that I put into this. If I want to switch careers (again), don't be fooled that is coming just for art. If you can replace an artist, you can probably replace any other kind of highly skilled, intellectual job like an architect, lawyer or whatever. You'll just be flipping burgers until the so-called utopia comes.
A year ago I felt like I have the whole world ahead of me but now it seems like I need to plan an exit strategy. I can't even get away from it. I need to hone my craft and I am subscribed to many pages and at least half of them share AI stuff on my feed and the anxiety just starts again. But don't forget, it's all in the name of "PROGRESS"!!!
Sorry to be a downer again, but here is where things are at the moment :
https://www.traffickinggame.com/ai-assisted-graphics/
First off, it shows that believing that "it will never get better than humans" is beyond pointless really. It is (obviously) lacking the personal touch of an artist in terms of their own references and personal visual library, personal back story and taste ... since the machine is using an immensely bigger set of references and visual libraries (equivalent to millions of combined artists), and is relying on a sense of taste perfectly weighted by popularity. But in terms of execution it is absolutely equivalent to high-end production art.
Secondly and most importantly it IMHO 100% shows that these ML models would have not been able to create these outputs had they not been using copyrighted work without consent - since public domain datasets would have never included modern graphic design and layout styles, digital rendering techniques only widespread since the 2000s and available as highres images, and themes/prompts like "character turnaround sheet" and "cyberpunk".
Now all that said I have no doubt that such extra layers of training could have very well been added on top of an "ethically trained" model for sure, by injecting more recent imagery (with properly acquired rights) on top - but that's kindof besides the point for now.
(On a side note, the fact that this person-typing-words-to-get-AI-art finds the first batch of images to be "not great" says a lot about how the tech is already devaluing art ...)
- - - - -
All this leads me to believe that now that such tools are out there, and that there is a high chance that they'll keep being used (regardless of how regulation around scraping evolves), perhaps the only ways for illustrators and concept artists to keep being relevant is to :
- Develop a strong counter-culture always on step ahead of AI tools (if that is even possible).
- Seek longer-term contracts with studios that value human craft and can afford the cost of human labor (as, after all, concept art and promotional art still remains a very small fraction of production costs even if made by a slow meat-person).
- Perhaps shifting to a "post internet" way of looking at things, trying to find means of living away from the online crazyness of AI-vomit.
- - - -
@CyberdemoN_1542
As far as I am concerned the hardest thing to do is indeed to manage/ignore this online craziness. It's like being in an abusive relationship : both the tech and the output are downright amazing, but the toxicity of it all really puts a huge weight on one's mind and ability to focus, and does hinder ones ability to deal with the art block knowing full-well that the tech is a threat and it only does what it does because it's been fed art without consent. It's infuriating.
Perhaps artists will need to find/build their own safe spaces (yuck) away from anything AI-related to be able to focus on their work. Maybe the key is to focus on team projects between like-minded individuals. After all, this guy developping his own Cyberpuk game by itself will not experience the pleasure of bouncing ideas with someone else, building a shared universe, and so on.
As a slow-meat-fellow traditional portraitist via another life, I'm utterly gobsmacked in just a brief period of time how far this algorithm contagion has spread.
Personally, I don't see those scraped 'remastered' pretty online pic's as visual creativity in the truest sense but rather nothing other than two bit knockoffs carpet bagged by ai-bros + cronies alike - so kinda needed a shower after seeing this
Which makes one both terribly angry and sad, all at once.
@pior
I CAN'T! I have to keep up with all the 3D program, addon, sculpting and texture painting developments because this is a high tech field! They often post AI news. I literally can't get away from all this. Whenever I manage to catch a break, news about AI creeps in and the anxiety comes back.
I SUPPOSE the saving grace for me is that I've only been doing this for 5 years, 6 this November. I imagine this is much worse to people who've been doing this for 10 or even 20+ years. Still, this occupation has been so good for me. It's literally changed my life. 5 years is nothing and I barely got started.
I wanted to learn everything pertaining video game art, including 2D. But now I ask myself what's the point? I suppose I should just refine my hard surface skills and hope I still have time to make enough money to soften the inevitable blow because the way I see things, every intellectual job out there is at risk. Architecture, Law, medical research, accounting, you name it. EVEN FUCKING PROGRAMMING! I have a really smart 150 IQ friend that works for Google and he is just as scared as I am. His colleagues too. All that's gonna be left are the lower skill jobs that are not really worth automating right now, like janitor work. I suppose the only job worth doing will be growing food. In short, we're going to be peasants at the mercy of the AI and whoever its masters are.
Oh, I don't think there's reason to be quite so far on the doom-and-gloom side of the fence. And let's face it - pursuing art has always involved filtering out the noise. Trying to figure out if the response you got is from someone trying to pad their post count or an actual artisan who for some reason was kind enough to take the time, no critiques and going from place to place searching for it, getting blindsided by some weird inflationist fetish shit on deviant art...
...this is just... another kind of noise.
It is disgusting to see. I do believe this is ultimately harmful to the world, to humanity, and to any semblance of sustainable happiness for generations to come. But it's not an outright death sentence to anyone with a functioning brain who wishes to use it. To an extent, the same reasons I think this technology is disappointing, and harmful are the same reasons I'm not concerned about it ultimately impacting my livelihood.
When you simplify things, when you devalue things, when everything is "accessible" - people get stupid. A lot of people are stupid. It is really hard to find good candidates for jobs - people who are able to communicate, follow or write a spec/brief, people who...know how to read and write in their native tongue. The world being eager to dumb themselves down to the point of being NPCs just makes the actual story characters all the more valuable.
If anything, @CyberdemoN_1542, I'd say this is more reason to also learn the 2D side of the fence - to learn the other elements of game art. There will be jobs, and if you're the equivalent of 3 separate trained AI programs and aren't a total tool - you're top pick. I am an adequate artist but most of my value as an artist actually comes from problem solving skills and other random shit I was dumb enough to burden my brain with lol
And you know, as depressed as I can get the more I see from AI development - it isn't without its humor. People think that because they can now (or they expect eventually) generate "art", that because they don't have to hire X, Y, Z that they can finally see their brilliant, totally-one-of-a-kind-not-something-that-can-only-exist-because-its-been-done-100-times idea realized without working with actual craftspeople - that they'll be able to make bank! People will see how great their ideas are, now that anyone with "passion" can make a game!
Except that eventuality would mean so could I so why the fuck would I buy from the lazy piece of shit who prompted his idea into existence lol
Yeah, the so-called "ideas people" will finally get revenge on people with talent and discipline. They'll show us!
Aside from 3D I also dabble in Level Design...I guess I can become a level designer. That is until the AI gobbles up that as well.
I've noticed that the most obnoxious people in this situation are the ones that equate AI art with making millions (usually in cryptocurrency) in Web3
If they are this obnoxious online imagine what it would be like to work with them in person unless you are exactly like them.
Hehehe -
(This screengrab could do with a little bit of AI-upscaling though ...)
FYI, that toggle has been there for many years, and was added when Adobe first starting diving into AI, long before the current generative craze. I can't go into much detail because an official response will be published soon, but I am proud to say that Adobe's AI efforts have, are and will be "clean", the data we use to train our models is either proprietary, public domain or licensed (also why it takes longer than simply slapping a web UI on top of LAION).
Well ... it being there for years, and (I believe) turned on by default without any explicit warning to the users is still the opposite of a good look.
As a matter of fact, isn't it pretty much a straight up violation of GDPR ?
The description on that dialog is poorly worded and contains few details, but having an insider view on what was analyzed and for what purpose, there is nothing shady going on. This is meant to allow the apps to process things such as face recognition for auto-tagging your photos or for smart object selection.
Well, to me (and many), the very existence of such an option turned on by default and without warning is the definition of shady - regardless of what is done with it. But indeed, we'll see what the official statement says (I am very curious to see if they'll downplay the issue, or apologize).
Anyways, more fun stuff :
https://www.instagram.com/p/CnH3-CQoQvB/?hl=de
Things being turned on by default when installing software is definitely a dark pattern. The fact that one has to read everything like a hawk is just wrong, but here we are.
also 1-button-press = game-ready-retopo when? :D I would submit data for that
this is going to be interesting.
"There is no copyright violation at any point. Proof: the AI is now in the wild for three months and more. And there is not a single lawsuit. Now guess why :)"
:) :) :)
I guess it takes more than a week to setup a major lawsuit?
even if this succeeds in the USA, which sounds like an uphill battle to me - other countries are not obligated to do anything. If it does go through, I wonder if Artstation will adopt Youtube's DMCA filing system where it's used as a weapon against other content creators... lol
Oh look, another one! Maybe this isn't as clear cut as some say it is...
Losing a market like the US probably soon to be followed by the E.U. isn't something that will be taken lightly. China is also already regulating, for different reasons, but still. 3 huge markets
How are recruiters in general getting on with this anyway? Last I heard it was along the lines of, "AI makes sorting through viable candidates more difficult."
Why is it making it difficult to select viable candidates?
Are candidates sneaking AI images into portfolios?
Interesting to learn about this variant of the problem, see when someone can post more information on the subject
Well, it's not so much a matter of candidates sneaking in anything, and more of a matter of AI bros flooding companies with applications because they're so "10x" compared to artists ...
well, if you're not going to embrace the power of gpt3 it does yeah
wrt recruitment - I've not seen it yet but I'm not hiring concept artists. I imagine it's going to be a huge pain in the arse when hiring programmers in fashionable fields (javascript and it's spawn) cos the market is already flooded with awful copy&paste developers and this is just going to make it easier for the incompetent to appear competent.
The sad thing is that the only real solution to this from a recruitment perspective is tests with feedback rounds
Microsoft taking a step forward, what do you think of this?
https://80.lv/articles/microsoft-presented-3d-avatar-generation-model-rodin/
this really doesnt look all that great, but it is interesting for sure, the 3d mesh only looks like a throw away mesh that is only used to get indications for positions of things, but each frame seems generated on its own independent of the geometry and the connected frames. otherwise it would probably look more stable.
This seems very similar to the "Mesh to Metahuman" plugin from a few months ago.
not really because that creates an actual metahuman with adapted premade hair etc, look at the camera turns how stuff is flickering, its not a mesh that gets turned, to me it feels more like ai is filling each frame with what it "imagines" what this hair shape would look like using the mesh as a positioning helper
https://3d-avatar-diffusion.microsoft.com/
Well, who knows right ! There's likely lots a fakery involved in these results anyways.
automatic avatar generation is interesting, but who wants to look like a AI generated potato outside of a corporate hell environment? I think more people would prefer to commission their own avatars or make their own. There's actually a decent market for this already happening in VR and Vtuber avatars
It's both : there is a huge market/desire for people to be able to take a picture of themselves and have their Madden avatar look just like them. As a matter of fact these things have been attempted before with crude photoslapping.
And then there's indeed the fantasy/anime avatars for VRchat and Vtube - but I can guarantee that the market for handcrafting these (as is currently done with Vroid Studio + manual touchups in Blender, all quite tediously) will straight up vanish as soon as a way to generate these avatars with a prompt will exist. People paying 3d artists 20USD for a custom Vroid are not doing it out of an altruistic desire to support artists - they pay because doing it is really quite tedious to do, and requires some technical skills.
The first serious commercial production where artificial intelligence is used?
https://80.lv/articles/netflix-angers-the-internet-by-using-ai-in-its-anime/
I kinda expected way more info from 80.lv about the process or images...
Here is more information and it looks like it is img2img.
https://www.cartoonbrew.com/tech/netflix-japan-ai-dog-and-boy-225631.html
The thing I find fascinating about this collage presentation ... is just how boring and uninteresting it is. And by that I don't mean the images themselves, which are fine ; but the whole process involving AI and successive "revisions", which evokes no curiosity or wonder or sense of satisfying hard work whatsoever. Not to mention that the original picture seems to be the most interesting by far anyways ! What a waste.
The tech was pretty much still in the realm of sci-fi just a few months ago ... and now it's already the most tired an unoriginal thing one can possibly think of. Funny how things go.
The layout is much more interesting and has more personality than any of the iterations. It's also weird they didn't change the colors or look much after generation, and this looks like something that might get better results if you could tag surfaces like the Nvidia generation tools.
As you can see, Shutterstock and Nvidia are going to collaborate to create an artificial intelligence that can generate 3D models from text. This is an innovation that could facilitate the creation and customization of visual content for different industries and applications. According to Shutterstock, the AI-generated models will have an appropriate level of detail. The models will be available on Shutterstock's website and will also be offered on Turbosquid.com and Omniverse.
Do you think this is going to be a blessing or a nightmare that will lead us to the unemployment line? In my opinion, this technology has great potential to democratize access to 3D and stimulate creativity, but it also raises ethical and legal challenges about the use of artists' assets to train AIs and the impact on the 3D labor market. I believe it is important to have a debate on these issues and seek responsible and fair solutions for all involved.
https://www.shutterstock.com/press/20477
If you could currently be replaced by an anonymous outsource artist then yes, the robot will take your job and it'll probably happen in the next couple of years.
In practice I strongly doubt that anyone currently employed will be directly replaced by a robot . Instead what will probably happen is that studios will rely slightly less on mass outsourcing and we'll probably also see a reduction in the number of junior artists hired as the technology improves .
This is kind of inevitable in any industry - eg. there's not a lot of demand for people who can write html any more.
A YouTube channel following the advent of Neo-Programmer Art.
https://www.youtube.com/@mreflow/videos
An informed and well-reasoned article on Lawfare about the copyright implications, and an argument for granting copyright only to human-generated works.
If we allow computers to make new copyrights, we should expect an AI-generated version of every melody and chord change possible to be authored and copyrighted, followed immediately by automated lawsuits to defend them. Every new hit will be followed immediately by a spate of lawsuits showing prior art. It could be the end of human creativity as we know it, and the rise of the AI copyright troll.
A great read.
Indeed, an informative articulately written piece.
There are a lot of trade-offs, and no easy answers. Artists and creators will be harmed economically, and it’s nearly certain that industries will be disrupted or even destroyed, although perhaps it’s too early to say which ones and how soon. It’s possible that the solution is right in front of our faces: Governments should restrict copyright to humans, and deny it to computer-generated works.
Although therein, a likely stone in the shoe!
For instance, when viewed through a fractured geopolitical lens aka: G7 - G20 and/or OECD vista, trying to formulate a 'workable' broad consensus would I fear be akin too herding cats.....in a timely manner.