So many questions: Are these books copyright by Tom Lesley, ChatGPT, OpenAI, ...? Should he share the money with OpenAI, or with the internet? Do we need to update the definitions of plagiarism? So it goes.
The case/example mentioned is that a single image from image generation software cannot be copyrighted, but a series of related images being used to tell a story can be copyrighted, because it's a human carrying out the creative task.
[1] - https://arstechnica.com/tech-policy/2023/03/us-issues-guidan...
More precisely, only human works are subject to copyright. The law simply doesn’t apply to works by animals, AIs, or anything else, even when they are as creative as a human.
> "the requisite level of creativity is extremely low; even a slight amount will suffice"
Literally throwing paint at a canvas is enough to suffice.
Arranging the prerequisite conditions and coaxing a monkey into taking a selfie is probably enough as well (as long as you don't hire PETA's dumb lawyers who want to argue for assignment of the copyright to the monkey)
There's not really enough information to determine whether this author would qualify. We don't know how much editing he did if any, nor have we seen these specific cases tested in a court yet.
I think it will ultimately hinge on whether courts find the totality of the process to be creative enough to constitute a “modicum of creativity”
My prediction is that courts will eventually rule that something like writing a prompt is enough. I think too much would be upended to rule otherwise.
If an LLM reaches a sufficient capability that this can be achieved while maintaining a roughly comparable level of readability, then you'd have just completely legalized 'piracy by proxy'. And the same would come to every single medium from movies to software. This is going to make it an extremely difficult question to answer.
I expect what it come down to is determinism. LLMs are completely deterministic - same model + same input + same seed = same output. And so any seller will be obligated to retain any model composition/data, and any copyright holder of an input training item will have a copyright claim on any output item. In other words, you can only train on stuff you already have the copyright to, or rights to.
Exponential growth!
So ChatGPT? The book includes unvetted output that was copied verbatim (As an language model blablabla)
ChatGPT can't, legally speaking only humans are capable of claiming copyright. In the monkey selfie case a few years ago it was determined animals and machines can't claim copyright. In cases where there is no human author, the work is public domain.
If I made a drawing with one of those old plastic spirograph tools I'd never wonder if the copyright was mine or the spirograph's.
To me GPT is either very close to or already crosses that line.
> OpenAI hereby assigns to you all its right, title and interest in and to Output. This means you can use Content for any purpose, including commercial purposes such as sale or publication, if you comply with these Terms.
I didn't say that it did, just that OpenAI disclaims ALL rights (including, but not limited to copyright).