Why This Award-Winning Piece of AI Art Can’t Be Copyrighted
Why This Award-Winning Piece of AI Art Can’t Be Copyrighted::Matthew Allen’s AI art won first prize at the Colorado State Fair. But the US government has ruled it can’t be copyrighted because it’s too much “machine” and not enough “human.”
Did you read the article? In this case he put in quite a bit of work to generate and alter the image:
He sent a written explanation to the Copyright Office detailing how much he’d done to manipulate what Midjourney conjured, as well as how much he fiddled with the raw image, using Adobe Photoshop to fix flaws and Gigapixel AI to increase the size and resolution. He specified that creating the painting had required at least 624 text prompts and input revisions.
And he is essentially claiming that the work should be transformative enough to be copyrightable. Even if the original image is not.
That all makes this case more interesting then a lot of others in the past as it is about AI generation with some human input. Not just someone generating vast amounts of work to find something they like (which likely will never be copyrightable). When this goes to the courts will will help to define the line of how much and what type of alterations are required to claim copyright over the works.
Not all AI work is the same, but I am glad that the copyright office is pushing back on these claims. Putting the burden of proof onto the author that they did have enough input into the work. The big open question ATM is how much input is needed and what that input can look like.
Who is they? Do you have links?
I’m not trying to be critical or skeptical, just want to be as informed as possible.
Given the presence of stolen artwork in the training data I don’t see why it should be copyright able.
Also award winning? It honestly looks like the kind of liminal mindfuckery most models could output. There’s nothing particularly impressive with the piece.
Eh. I’ve seen abstract art that people are in awe with throughout my life. And like the uneducated swine I am, I’ve never thought they were impressive either.
Art appraisers are weird.
But according to the article, it wasn’t generated in minutes. The artist went through over 600 iterations of tweaking the prompt to get what he wanted. Sounds like days or even weeks of work probably. And then made additional tweaks via Photoshop.
Not too say that makes it any more impressive, but it wasn’t something that was without effort.
Point taken. In that case, I guess one can recognize the effort. Still, the impressive part of the piece is the style, which, if one were to assume was made with actual oil paint, it would be impressive.
With AI, I would explore styles that are inherently difficult to produce digitally.
Which I can also do. “Imagine a cave full of cats. The first cat is pink with yellow dots. The second cat glows in the dark. The third cat…”
I guess art involves expressing what you want to express it, sure. But also how you express it is part of it as well. If you make the strokes yourself, it’s more impressive. A machine? You gotta do better than making it look like someone painted it.
It’s like 3D printing. A 3D-printed statue? Neat. But not terribly impressive. A 3D figure that defies all optical illusion explanations? Now we’re talking.
Funny situation indeed. Thoughts:
So your main argument against it is “anyone could do it”?
What? Lol
To file an infringement suit they’d need to have paid registration for each image which, even for the exorbitantly rich, wouldn’t be remotely feasible for all logical arrangements of words/images. There’s probably not even enough space in the Universe or time until its heat death to generate and store all such images.
Even if they did, copyright doesn’t protect against against independently created works that happen to be similar or even identical - so they wouldn’t be exhausting some limited set of possible works by doing so.
For my understanding artistic works get copyright from the moment of their creation. This would allow one to pick battles based on how lucrative they may potentially be.
You dont really need art museum of babel for this but you just tons of different works that may contain unique characters, structures or objects similar to what someone might be able to imagine or has already imagined.
You may draw fan art with disney characters but its actually illegal to sell said art work without Disneys aproval until copyright expires. Now if any one can start churning out for example A.I generated web comics left and right the chances for almost identical designs increases by a lot.
For my understanding artistic works get copyright from the moment of their creation. This would allow one to pick battles based on how lucrative they may potentially be.
In the US, you need your copyright to be registered in order to file an infringement suit or be granted statutory damages. This must be done prior to the infringement, so they wouldn’t be able to pick and choose which to register after the fact. The fact that (unregistered) copyright arises from the moment of creation is true, but not particularly useful here.
You dont really need art museum of babel for this but you just tons of different works that may contain unique characters, structures or objects similar to what someone might be able to imagine or has already imagined.
Copyright is not the same as patents or trademarks; someone coincidentally creating something very similar or even an exact replica of your work is not infringement.
If whether you copied from their work or independently made similar choices is under question - then close similarity of the works could skew the balance of probabilities. However, the courts will be able to see that coincidental similarity is far more likely if a colossal number of images have been registered.
You may draw fan art with disney characters but its actually illegal to sell said art work without Disneys aproval until copyright expires.
It’s still copyright infringement even if you publish it non-commercially, but a Fair Use defense would likely hold up.
The museum of babel already exists.
However with generative A.I you dont need artwork for every combination of pixels. Machine learning seems to be really good at finding patterns in everything we may do or think.
With generative A.I we can use this information to create increasingly more human like output. In terms of art mimic and blend art styles, create new designs based on existing ones etc.
Much more elegant and way cheaper than using brute force algorithms.
Depends on how radical we’re talking.
On the less radical side - UBI, donations, contracts for creation of art that people want even if it’s not copyrighted afterward, Convenience Factor (often used by FOSS projects), and also the fact that a lot of art is created and intended for free redistribution anyhow ^.^
On the more radical side: re-examining our entire concept of work and labour and rent, decentralused gift economies, automation, the destruction of capitalist structures as a whole, etc.
The artwork, Théâtre D’opéra Spatial, was created by Matthew Allen and came first in last year’s Colorado State Fair.
No. No he didn’t create it. He put words into a black box.
“He didn’t create it. He moved a mouse.”
“He didn’t create it. He put commands into a keyboard.”
“He didn’t create it. He pressed the camera trigger.”
“He didn’t create it. He threw store-bought paint at a canvas.”
“He didn’t create it. He cleaned some dirt off the wall.”
“He didn’t create it. He was inspired by gods.”
Where you see a categorical difference, I see a qualitative one. AI-generated art can be nothing more than putting words into a blackbox, but it can also be a day-long process of tweaking dozens of parameters to get what you want from the words you put into the box. A child can slather paint onto a canvas without much thought - but that doesn’t mean great artists drawing complex, intricate paintings isn’t art, does it?
Generative AI is a tool. It can do more than most tools, but still, it is something wielded by an artist.
Even if I were to grant you that generative AI is just “describing an illustration”: other people say there is a world of difference between painting something with your hands and using a mouse, yet I think digital illustration is as real as physical illustration. Yet other people say there is a world of difference between creating something from the ground up and using store-bought materials and tools, yet I don’t discount artists who do just that.
But I don’t grant you that, because if I simply describe an illustration, the generative AI will not give me anything close to what I want. I have to learn the prompting language of the model (what words and phrases result in what?), I have to learn the influence the many different parameters have on the output, and I have to learn how to use things like prompt weighting, negative prompts and the like to get what I want. It’s something completely different from describing an illustration.
And that’s ignoring things like variant generation, inpainting, outpainting and the many different things that are completely removed from just “describing an illustration”.
Learn how to make a digital illustration, learn to make an oil painting and learn to make an AI image. Then we can talk.
I can do all three (worked on comission basis as a digital illustrator and did make Sci-Fi illustrations with acrylics and ink in the past). Generating AI images is not even in the same universe as the ballpark were digital illustration and traditional painting are playing.
It’s like saying watching someone’s Let’s Play of playing GTA is “kinda similar” to driving a Formel1 sports car yourself.
L4sBot
ram
Johanno
ParsnipWitch
ThǝLobotoʍi$T
Dojan
maegul (he/they)
Vipsu
Corkyskog
Robaque
sapient [they/them]
Fisch
raoulraoul