Hacker News Reader

Comment by pavon | Hacker News Reader

It goes further than that. Kris Kashtanova made a comic book named "Zarya of the Dawn" using midjourney for all the art. When registering for copyright she claimed copyright over the art for herself, under the theory that she was prompting the AI, and it was just a tool. When the copyright office learned about this they revoked her copyright registration[1]. The view of the copyright office is that this is no different than the photographs taken by the monkey - the monkey can't claim copyright because they are not human, and the owner of the camera can't claim copyright because they didn't have sufficient creative input, so the work is ineligible for copyright at all.

She is eligible for copyright on the text she wrote, the composition and editing performed on the images, but not the raw output of the AI.

[1]https://arstechnica.com/information-technology/2023/02/us-co...

View on HN

I think there's a second potential future legal element to that case because of the blatant use of an existing human's image as the character.

https://www.reuters.com/article/ai-copyright-legal-idUSKCN2V...

> “Zendaya leaving gates of Central Park,” Kashtanova entered into Midjourney, an AI program similar to ChatGPT that produces dazzling illustrations from written prompts. “Sci-fi scene future empty New York....”

> From these inputs and hundreds more emerged “Zarya of the Dawn,” an 18-page story about a character resembling the actress Zendaya who roams a deserted Manhattan hundreds of years in the future. Kashtanova received a copyright in September, and declared on social media that it meant artists were entitled to legal protection for their AI art projects.

This is so weird. Can artists who splash paint randomly at a canvas not copyright those artworks? Or ones who attach a paint bucket to a rope and swing it over the canvas, giving it just one push?

This is actually a good point. I've seen artwork where the artists just had random passerbys inadvertently make changes to the canvas until they considered it 'done'.

She wasn't just prompting the AI. That comic required significant work on the user's part to edit the images and re-synthesize portions of each frame repeatedly while doing things like inpainting, outpainting, re-prompting, etc. And likely required use of some other tools like Photoshop or Indesign (midjourney absolutely did not generate the bubbles or comic book text).

I’m guessing that the copyright equivalent is whether the editor of someone else’s photograph or the director of a photo shoot should have copyright privileges. The answer is pretty clearly not.

It’s not that she didn’t do (a lot) of work, but that the creation of the underlying work happened by AI. If it were human then it would hold the copyright, but it’s not so it doesn’t. The logical misstep is thinking, “because it can’t have copyright, someone else can.”

But then by this logic photos shouldn’t have copyright either because a camera can’t hold copyright

I think prompting the AI is like taking a photo in that you set the parameters and subject and press the button to get an image

Exactly, generative AI, like the camera, is just a fancy way of transforming ideas from the real-world into a digital representation, but with a lot more parameters.

> That comic required significant work on the user's part to edit the images and re-synthesize portions of each frame repeatedly while doing things like inpainting, outpainting, re-prompting, etc. And likely required use of some other tools like Photoshop or Indesign (midjourney absolutely did not generate the bubbles or comic book text).

The issue is that the AI's output is deterministic, and the prompts it uses are commonly too short to be eligible for copyright, not least because someone else could plausibly use the same prompt independently. With the same seed, the AI would then produce the same image. Or produce a similar image in response to a similar prompt.

If you could copyright that, you then have an enforcement problem, because the usual way to show copying is similarity, but you can then have two images that are similar yet independent. And some light Photoshop work is not going to eliminate the similarity.

If someone builds a device that sprays paint onto a canvas based on some parameters, and then a person use those parameters, writes a small controller that sprays the paint a specific way, manually adjusts the position of the device over multiple sessions, and does their own touchups to create a piece of art, would they not hold the copyright on the resulting piece?

Yes, there is a deterministic component (although, to my knowledge, the AI tools aren't actually deterministic over their parameters). But the resulting work involves a lot of human creativity and intervention.

Suppose you have a printer with a programmable head which can be used to print out arbitrary images. Can you copyright the pictures it can print, even though the printing is done by machine? Obviously. But then it's not the machine choosing what you want on the page.

Perhaps my analogy wasn't clear. I meant, the painting device uses your parameters, but also supplies some modifications. Like a stenograph.

At the end of the day, generative AI, Stenograph, and this painting device are all tools.

If you shake a kaleidoscope and then take a picture through the lens, I'm pretty sure you have copyright on that photograph, even though someone else made the kaleidoscope

It's not the presence of a machine which makes it ineligible for copyright. It's the absence of some independent creative expression. If you add significant creative expression to what the machine does, that's not the same thing as the output of a machine produced from a minimal input.

What we're talking about here is definitely art where a human is providing independent creative expression, in combination with generative tools

"The issue is that the AI's output is deterministic"

Point a digital camera at a still scene under controlled lighting conditions and it's pretty deterministic also.

Anyway, it's easy enough to add randomness to the AI image generation process (and Midjourney does do that by default, as a matter of face), so the process need not be deterministic at all.

It's not about the physics of light, it's about what the artist chooses to have in the scene.

And it's not randomness that fixes it. Do you even want to get into the information theory problems if you could copyright the output of a random number generator?

What you need is for someone to be making creative choices with enough entropy to be considered original and not be reproducible by feeding a short sentence to an existing algorithm.

I really like how you put it. Good thought! But I don't quite agree that entropy is a criteria of originality. Because originality is not ranodmness, it's... distinctiveness.

If I commission an artist and give them exacting directions, they are still the artist and copyright owner of the work. They can then assign it to me, but I am not the artist simply because I commissioned the work and gave input.

Typically commissioned works are "work-for-hire" where the person paying for the work gets the copyright, no?

Same as how when you freelance, the customer (usually) owns the code you write for them.

It's an interesting argument, treating the AI more like a commissioned artist than like a technical tool.

But in the end, this seems like the opinion of one mostly irrelevant government office? Copyright protection is automatic in the US since 45 years now, you don't need the involvement of the copyright office at all. For establishing that you created the work, a letter from the copyright office refusing you is as good as a letter accepting your work, and in most cases it's redundant anyways. The only relevant opinion is that of the court where you sue someone else for copyright infringement.

I understand that the individual AI generated elements can’t be copyrighted, but in this case the final product is a huge amount more than that. It’s a collage of such elements, plus cropping, filtering, colour grading, composition in the page and overlay with speech bubbles and such. It seems to me that the end result should be copyrightable.

Then the fact the original elements aren’t covered isn’t a problem. You just delete them.

What is weird about this is that even if AI works aren't protected by copyright, under Feist a collection of uncopyrightable information can be protected if they are selected coordinated and arranged. That is almost certainly met when creating a comic book or graphic novel.

The monkey taking a photo isn't a good example because they never tried to register it in the US, they never actually ruled on it. There is plenty of other case law showing "the modicum of creativity" is a low bar, copyright does not require "sweat of the brow".

The problem I have with this is who is the arbiter of what is "AI"? Couldn't I just claim I used "autonomous state machines" and not fall under this categorization?

It's also a literal pandora's box because somehow an AI itself can be copyright, but what it produces can't be? Surely it isn't very hard to have an AI that produces other AI? So who owns those?

This seems dumb. If she hadn’t told them how she created it, it would have been copyright? Is the copyright office going to start asking artists how they create their work now? This is just a bad decision on their part.

There are laws and rules for copyright, and when you claim copyright, you make a legal claim to the government that those rules apply to yourself and the piece in question. TDIL: When I make legal statements to the government, such as that copyright applies to a work I am submitting, those statements must be true.

> If she hadn’t told them how she created it, it would have been copyright?

No. OTOH, the copyright office might have letter her register the copyright.

If it is copyrightable, it is copyrighted at creation, and if it is not, registration doesn’t change that.

Moral of the story is never reveal how you produced a particular art work.

I think in the relatively near future, video of the artist creating a piece of art will be part of what makes a given piece of art valuable. The provenance will be just as important as the art itself.

AI tools will soon be able to generate videos showing a specific human producing a specific piece of art. Should we require a video of the person taking the first video also?

I don't think it's going to be all that soon. Human images are much more difficult to fake than generic "art", much less human video + a piece that is physically created. Plus the price to produce that complex of a fake will be higher than the cost of an artist to setup a tripod and start recording.

Recorded public performance of art creation may become the norm after that and then actual robots may be making physical art with impressive skill and intrinsic value...

...Zima Blue. Interesting times.

This would be hilarious then, because it'd be a video of each artist creating an absolute breathless masterpiece in one single sitting. If forgeries have taught me anything it's that creating visually duplicate works of art is easy.

Well, if it was a forgery then it would be a video of not-the-artist creating a visually duplicate work. That wouldn't work very well.

I wonder if someone screen printing images created by an AI tool would satisfy the requirements for ‘human authorship’.

Edit: screen printing is a manual printing process (in line, I think, with pop artists recreating commercially printed labels, etc), so this seems like a reasonable question.

Reply on news.ycombinator.com