AI, art, and copyright: The fight over 'Théâtre D'opéra Spatial'
Can AI-generated art be owned? The copyright debate surrounding 'Théâtre D'opéra Spatial' isn't as hairy as it feels.
A bit of a shorter and less technical post today. I tend not to focus on current events, but this one is too core to the idea of AI Artistry. Let me know how it hits you. 👀
I wouldn’t call myself an expert on ethics or legalities in AI, but I am fascinated by both. If you haven’t seen it yet, an interesting stand-off is occurring. Artist John Allen won an award, got denied a copyright by the Copyright Office and is appealing said denial for his work Théâtre D'opéra Spatial.
One curiosity led to another, and I began thinking about the implications, both ethical and logical, around an artist and their piece. Allen might have a case (and his appeal efforts go on), but there’s not much more to say other than:
Allen's argument is: He wants copyright for his AI-assisted artwork on the basis of his creative direction.
The Copyright Office's argument is: The office says AI-generated works can't be copyrighted because they're not made by a human.
At first, my feelings were very… complicated. One question led to another, and it recursed all the way back to the idea of copyrighting. What’s the purpose, the deeper meaning even, of the Copyright Office? Is the “how” really something they should care about? It seemed layered and new and ethereal. And the kernel at the center was inextricably fluid based on your beliefs about the Copyright Office. As a newsletter post, it generated a Tony the Tiger outline.
Until I stumbled across the right logical comparison, and it became blindingly clear what my thoughts on it had to be. Think of the camera.
A photographer pushes one button to capture — or create — an image. That image is yours and you can copyright it and license it. You really don’t even have to be in the right place at the right time — like those pesky Pulitzer winners — to do so. A photo is inherently yours when it’s snapped into existence by the click of a button (if you didn’t change a single word of that sentence, it 100% applies to image generators). Where a camera captures, an AI model creates (or imagines, if you like).
The analogy can go deeper.
Think of two artists who take (photograph) the same image from different angles. It’s forever clear whose photo is whose. And those photos are copyrighted on the basis of this difference? If it helps you to think about it like this, the same prompt just creates different angles of a different imagined concept.
So in a very long-winded way of saying ‘I’m right and you can’t disprove it’: John Allen is right and has the rights by all logical right.
But if there is one question that will dictate our ethical feelings on this fight, maybe it’s how we answer this: Should copyright be denied just because a creation happened by chance (and could have as easily been created by someone else)? Because isn’t that the only intangible difference between a DSLR and Dall-e, Midjourney, you name it?
And at that point, whoever has it first can profit. Then it’s just a nine-tenths of the law thing (like it is nine-tenths of the time). But I’m curious to hear your thoughts on it.
Logical comparisons
I don’t think there’s a better logical analog than the photograph, but I did cycle through a few others to get there:
A short story in a found notebook
Captured sounds (licensable and so easy to copy)
A poem that a medium channels (if you believe in that sort of thing)