IANAL, and for a real opinion on contract law you should definitely consult a specialist contract lawyer, but this section:It's potentially a valid argument as the Agreement does not have a stipulation on using the images specifically for AI generation. Also the agreement identifies that the contributor still maintains all copyright etc, for the images.
The license to our end users may include the right to modify and create derivative works based upon the Work, including but not limited to the right to sell or distribute for sale the Work or any reproductions thereof if incorporated or together with or onto any item of merchandise or other work of authorship, in any media or format now or hereafter known.
There won't be any backlash... people are just getting a few peanuts per month from their work and if they are lucky they might get one or two commissions for weddings or ads. Imagine spotify but much worse... The worst is that I have to opt out for training images with whatever I make in the adobe ecosystem.The Adobe Stock Contributor Agreement isn't at all clear about the use of contributor photos for this kind of thing. I expect backlash from stock photographers about the use of their images to train Adobe's AI model here, justified or not.
There was no such thing, now there is. The dictionaries will be updated accordingly.
For those of you concerned about AI art dominating the industry, this might give some comfort, if you haven't seen it. Bear in mind that a lot of questions remain and this is far from definitive, but I find it very interesting. If you can't copyright an image from an AI, that would dramatically reduce the appeal in today's IP-driven creative industry. The argument is interesting too, although I'm skeptical that it will hold up under closer scrutiny.
It has nothing to do with philosophy. zunipus said there is no AI art "as per every dictionary ever written". It's semantics.Well, if we go into the philosophy of art, we can consider the Grand Canyon. Is it art? No, but it is beautiful. It was made by complicated processes and even a picture of it would be art, but it itself is not. Now, one might ask if an AI graphic of a canyon then is art or not? There's no expression of creative skill or imagination, for instance. So in a technical sense, we can argue it is not art, but merely looks like art, much as an earthquake knocking over a paint can might create something that looks like art.
Nope. Not my argument.
Try again to support the position that a bot is an artist.
You're using dictionaries as your definitive source on what art is? Seriously? When the question what is and isn't art has been hotly debated for over a hundred years, with multiple schools of thought. Look at this part from section 5 of the Stanford Encyclopedia of Philosophy's article on the definition of Art:As per every dictionary ever written:
There is no such thing as AI "art", and there never will be.
I have seen in the Louvre students with easels sitting in front of an old master trying to imitate the artist style, so clearly it's been a large part of the art education for centuries. I'm pretty sure Leonardo's students painted a few...That's simply not how these image generators work at all. They're not sampling. The AI trains on tens of thousands of images and keeps exactly none of them, not even a portion. Unless they've somehow found a way to compress hundreds of terabytes of images down to a few GB.
For that matter, style is not copyrightable, even if a particular style is attributable to a specific person. A specific work in a style is copyrightable. I can imitate Terry Pratchett's style of humor without owing anything to his estate.
It's literally not theft if the artists have specifically signed over to Adobe the rights to use their work and relicense it as they see fit, for any purpose specifically including making derivative works and specifically including purposes that haven't been thought of yet. Which is what the small print in Adobe's documentation that was posted up-thread actually says.Adobe. You wouldn't steal a car. Right ?
I'm not cheering on theft. Ai art is theft... that is the intention of every single company pursuing the tech. These AIs are not creating anything. People can argue about how humans iterate on existing works all they want. Creating software that apes style and composition is not art. Its here now no undoing it... but I won't be supporting any company trying to profit from it either. "Ethically" sourced photos or not isn't really the point. The tech at its core is a theft tool.
Yes, people have been, not advanced AI systems being used by the masses. Anyone who has ever spoken to even just a semi-experienced artist understands that they never show their direct reproductions, let alone distribute and claim them for their own. There have always been unethical artists and artists who recreate unethically, but we call those unethical artists for a reason.people have been training off other people's art since the invention of art. There isn't anything unethical about it.
"AI might put the mediocre ones out of a job though."Eh, if the art someone can produce can be simply replaced by a robot then I’m not sure they were much of an artist to begin with.
The good artists will still get paid and if they’re smart they will use AI for inspiration or to do more tedious or labor intensive parts of their job. AI might put the mediocre ones out of a job though.
"AI might put the mediocre ones out of a job though."
But on the other hand, it might put us atrocious artist into a job. I have the artistic ability of a tree stump. I do, however, have ideas, ideas for which I lack the physical skills to express visually. AI image generation might help me--the worst artist in the world--express my ideas.
By extension, I can also see it being a godsend for the physically disabled who want to give artistic visual expression to their ideas.
I'm not INHERENTLY opposed to AI art generation. I'm opposed to theft of course, and while teaching a human to draw using previous samples is one thing, it's another when it's a corporate owned soulless entity. That does change things morally. Further, putting artists out of work is a threat under our current social systems. Provide a basic minimum standard of living and eliminate poverty, THEN we'll talk about how this tool could benefit people. As it stands, it threatens instead to put artists out of work and replace them with... nobody. Not even you. Some corporate executive will just say "make the thing" and an AI will do all of it for free. No art staff, amateur or otherwise, required. It evens the playing field by putting talented artists out of work, not by putting those with ideas but no means of expressing them INTO work."AI might put the mediocre ones out of a job though."
But on the other hand, it might put us atrocious artist into a job. I have the artistic ability of a tree stump. I do, however, have ideas, ideas for which I lack the physical skills to express visually. AI image generation might help me--the worst artist in the world--express my ideas.
By extension, I can also see it being a godsend for the physically disabled who want to give artistic visual expression to their ideas.
The other day in my art class, my teacher was talking about how unfeasible it is to be an artist today, and how even a professional with ten+ years of industry experience might needs a secondary source of income if they're not involved in a very exclusive and small group that can be employed full time by a massive corporation (where they will likely never own most of what they make). Being an art teacher has become the dream as at the very least you get to stay within your field for your work.snip
The point I'd make is that people talk about all the artists being put out of work. Are they though? I'm not an expert in the field, but from what I can tell, it's already a brutal field. Many of them seem to have day jobs and do art on the side because they want to. At best, this might reduce the likelihood of people running a side gig or alternatively, AI assisted art might make provably human generated art that much more valuable. It could boost the incomes of the mid-tier / top-tier artists.
He's correct. The paintbrush is not the artist.I'm not sure this philosophy of art position is going to work when it comes to these systems in the real world, but I appreciate the effort at least.
I hear the pain and respect the point. I tend to view the problem (making a livable wage) as a 1st world problem. Let me explain by example.There's no reason why it should not be a feasible profession that pays a livable wage.
Your contention takes my statement out of context. What I said was that using automated tools or lackeys to produce work, and THEN subjecting it to mass reproduction diminishes the role of a so-called artist. And I stand by that statement, and shake my fist at clouds.You specifically said mass reproduction diminishes the work. Which means producing copies of the same work. Like making prints of a painting or copies of a book. What you seem to really mean is mass production of distinct AI images based on random prompts from users, and that's different.
Your contention takes my statement out of context. What I said was that using automated tools or lackeys to produce work, and THEN subjecting it to mass reproduction diminishes the role of a so-called artist. And I stand by that statement, and shake my fist at clouds.
And that applies to professionals or to hobbyists. Again, the involvement of the tool-user at each step, with active engagement and editing and re-direction, is the bright line I draw. Offering a prompt and taking the end result as presented doesn't cross that line.
Workman-like product can be produced with that easy approach, and I am happy to accept it as craft, but not as art.
I think we can argue whether or not it's art all we want, but critically, the real argument is whether or not it was created by someone. I'd argue no. In the same way a monkey taking a picture shouldn't give the man who owned the camera a copyright on that photo (nor the monkey, unless the monkey asks for it), neither should any "artist" get credit for telling an AI to make something, and then the AI makes it. They get, at BEST, copyright over the specific text they typed into the prompt. Who does it belong to? I'd say it's best if art generated by AI belongs to no one at all, unless the AI asks.Your contention takes my statement out of context. What I said was that using automated tools or lackeys to produce work, and THEN subjecting it to mass reproduction diminishes the role of a so-called artist. And I stand by that statement, and shake my fist at clouds.
And that applies to professionals or to hobbyists. Again, the involvement of the tool-user at each step, with active engagement and editing and re-direction, is the bright line I draw. Offering a prompt and taking the end result as presented doesn't cross that line.
Workman-like product can be produced with that easy approach, and I am happy to accept it as craft, but not as art.
It seems we are more or less in alignment on this contentious issue.As far as the output of AI image generation is concerned my personal feeling is that it shouldn't be subject to copyright protection unless it's had substantial human modification. Apparently the Copyright Office feels the same way.
To me art requires intent and the generator doesn't have intent. It has a model for interpreting written prompts. It's a more complicated, automated paintbrush.
I'll accept that.I think we can argue whether or not it's art all we want, but critically, the real argument is whether or not it was created by someone. I'd argue no. In the same way a monkey taking a picture shouldn't give the man who owned the camera a copyright on that photo (nor the monkey, unless the monkey asks for it), neither should any "artist" get credit for telling an AI to make something, and then the AI makes it. They get, at BEST, copyright over the specific text they typed into the prompt. Who does it belong to? I'd say it's best if art generated by AI belongs to no one at all, unless the AI asks.
Apparently Adobe doesn't feel the same way. And now that this thing is so "ethical", the copyright issue is resolved, and image generator is being integrated into the most widely used software for artists the Copyright Office will have little choice in the matter.As far as the output of AI image generation is concerned my personal feeling is that it shouldn't be subject to copyright protection unless it's had substantial human modification. Apparently the Copyright Office feels the same way.
To me art requires intent and the generator doesn't have intent. It has a model for interpreting written prompts. It's a more complicated, automated paintbrush.
People are still doing it. AI isn't sentient. People are controlling the AI and telling it what to do.Yes, people have been, not advanced AI systems being used by the masses.
omg you sound exactly like the people who say 3D art isn't real art, that only 2D art is real. Or that digital art doesn't count as art, only traditional art does. People having to do less work to make something is a good thing! Its not bad!A lot of the ordinary people who aren't artists getting their hands on this technology don't understand what it means to labor for years, to study other artists and their processes, and painstakingly work to develop a distinct style and voice with a bedrock of good influences.
Of course there's such a thing as unethical art. Art created by abusing people is unethical. We even have numerous laws meant to decompile industries built around this sort of unethical art. Do you want me to spell it out further?There is no such thing as 'unethical art' and suggesting that there is ... that's terrifying.
There is also no such thing as an ethical Adobe product and suggesting that there is ... that's terrifying.
Why is Adobe unethical?"Ethical" and "Adobe" are not words one often sees used together. I may need a moment to process this.
Alexander Dumas gave some one a job, even if a very shitty one.Hate to bring Elon Musk into only a tangentially related conversation, but Adobe's example shows why buying Twitter might pay dividends, after all. Twitter gives him a ton of data with rich context, on which he can train language models without running into ethical or legal issues. OpenAI might capitalize on this.
100%. Really it's a pretty basic problem to solve, so much layout is based on very established grid principles, and unless you're reading old issues of Raygun or something not particularly innovative overall. Certainly at the very least there's a lot of boring drudge work before you get to the more creative choices.
This train is leaving the station, one way or another.
I'm a Creative Cloud subscriber, and I don't have massive love for Adobe or their subscription system. And yet, I like using the tools for the most part, and as a professional the (tax write off) value I get out of them is more than worth it. Replacing Photoshop would be a huge lift for me, I'm very fast in it and it's extremely powerful.
I signed up for the Firefly beta, figured I might as well see what's coming.
I don't agree with your assessment of the similarity of human cognition and machine learning. However, this comment was so outrageous I made an account just to respond.correct. Much like how I can commission an artist to draw an image of the pokemon Poochyena. They would use official art as reference, but not actually use the image for making the art.
I have worked in commissioned art as a freelancer.I was talking a while back to a marketing head at a uni. He was observing that the art & design students liked interning for him because he paid them. Most shops had a pool of interns that had to compete for projects. They designed some page layout / picture / etc and then the company paid for one of the pool.
I listened to an interview with the art director at Fantasy Flight Games during the LCG heyday. She observed that to generate hundreds of art pieces for a monthly or bi-monthly release required many artists. She had a website requesting specific art pieces with styles described and artists submitted pieces. She'd buy some of them and use them. The rest were left unused.
Back to the marketing head at the uni, I asked him what the average graduate made going out into the world with a credible deign degree. He observed that the average pay was around $35k per year in our region (lower cost of living), but compare that to the estimated baseline of $41k at the time for a family (all necessities and a few extras).
The point I'd make is that people talk about all the artists being put out of work. Are they though? I'm not an expert in the field, but from what I can tell, it's already a brutal field. Many of them seem to have day jobs and do art on the side because they want to. At best, this might reduce the likelihood of people running a side gig or alternatively, AI assisted art might make provably human generated art that much more valuable. It could boost the incomes of the mid-tier / top-tier artists.
I never said that any of those mediums aren't real art, and it's either highly dishonest or just plain foolish to act like AI art is the same as digital art. I can fully accept that there are uses of AI in art that serve as extentions of the artistic process, but that's the key here: the process. In art school you are taught not just to improve the product, but how you get there. That is preserved in digital art, but is completely eliminated in most AI generation. A good number of the people using AI aren't using it to generate parts of a work, but are instead generating graphics that are fully completed pieces from the get go.People are still doing it. AI isn't sentient. People are controlling the AI and telling it what to do.
omg you sound exactly like the people who say 3D art isn't real art, that only 2D art is real. Or that digital art doesn't count as art, only traditional art does. People having to do less work to make something is a good thing! Its not bad!
It's also worth pointing out that some of these examples of luddite fear of the future, as in talking about the fear of cars, miss that the dominance of cities designed for cars rather than people really HAS become a social disaster and the fears over cars were in fact quite justified. It's hard to get that across however, because of a generational failure of imagination to see how cities could be designed any other way. The fear here is that if AI art becomes the norm, future "artists" won't be able to even conceptualize how are could be done in different ways. This HAS been an issue in the past in fact. The era of classical realism was so engrained that art that went very intentionally unrealistic was seen as disruptive and, frankly, "bad", reacted to even worse than Wind Waker in the "realism obsessed" gamer culture of the early 2000's.I never said that any of those mediums aren't real art, and it's either highly dishonest or just plain foolish to act like AI art is the same as digital art. I can fully accept that there are uses of AI in art that serve as extentions of the artistic process, but that's the key here: the process. In art school you are taught not just to improve the product, but how you get there. That is preserved in digital art, but is completely eliminated in most AI generation. A good number of the people using AI aren't using it to generate parts of a work, but are instead generating graphics that are fully completed pieces from the get go.
And yeah, people are telling the AI what to do for now, but it seems like no one is considering the long game here. No one is thinking about how it will change or dictate tastes. There is a real possibility that art becomes fully stagnant as it fully becomes a means of entertainment rather than expression, endlessly remixing existing artwork into ways to elicit the greatest amount of pleasure possible. There is a real risk that we are creating a psychological trap that we are not prepared to get out of.
It's so rude that people here are acting like any reservation is irrational, any hesitation or criticism is symptomatic of some kind of personal failure in understanding rather than a failure in the current state/expression of the technology. It's important to be worried, because this is not utopia and giving in completely to the idea that any technological development is necessarily progress is highly dangerous.
Oh sure, if you don't mind the complete loss of CSI's accuracy. You can never use AI to reconstruct the original license plate data in an image if not enough data to show said number wasn't already in the image in the first place. The AI, like us, will "invent" based on it's biases.At the end of the video, I see CSI's "enhance!" is finally here.