Why would ChatGPT be sued exactly? They didn’t publish the picture, they did what the user asked. The user is responsible because they directed the creation and publishing. The user is the entity who should be sued. Or dmca’d. Or whatever.
Think if the user commissioned the art from an outsourced creative shop nobody has heard of. Then they published it. They wouldn’t go after the creative shop, they would go after the publisher.
(I am just addressing publishing here, training on the artist’s works is a different, well discussed issue)
As I said, I am talking about the publishing, not the training. If it produces his signature, which basically claims he created the art, then that is the line he believes was crossed. I am saying that it is up to the user to understand when those lines are crossed. The user could have done this in photoshop as well without ai, would we blame photoshop? Training is a separate argument, it is not what the article tackles.
It is certainly not trademark infringement, I can say that for sure -- no reasonable person would prompt ChatGPT to create a "new yorker style cartoon" and then think the result is produced by the actual cartoonist in mere seconds for their viewing pleasure. So there's no confusion in the marketplace.
The user could cause confusion in the marketplace of course, but that would be her doing, not the app's. Surely we can all agree that suing adobe illustrator for facilitating trademark infringement of logomarks and such would be silly?
It could be copyright infringement, which should drive home how absurd copyright is as a concept. Everyone's all up in arms about Anthropic reporting a user to the police today -- imagine if the thing they were reporting was that she had written a sacred symbol in her personal notebook...
The article is evidence there is confusion in the marketplace - people are contacting the cartoonists about the fraudulent works.
Now the argument can be where does the infringement land - does it land with the person who generates the cartoon, fails to remove the signature and uploads it to instagram? Maybe. But I don't think we should be giving these AI companies yet another get out of jail "oops you made a machine that does a bad thing" card.
I also firmly believe there are contexts in which merely producing the trademark and attaching it to something is enough to be infringement but I'm not going to do the legal legwork to get there to your satisfaction, sorry.
> no reasonable person would prompt ChatGPT to create [...] and then think the result is produced by the actual cartoonist
There's also a narrower situation to consider, where the user describes something they "want"--implicitly to find--but the system generates a fraudulent one instead.
In that case ChatGPT would be committing a trademark violation, at least within a nation of laws rather than lobbying.
The original story is, the Disney movie’s version isn’t.
Edit: This sort of thing is common in Hollywood. For example James Bond (first book) hits public domain in ten years, but not all of the elements we associate with the movies are from there. Q and his gadgets are inventions of the movies and don’t enter public domain. There’s a reason patent/tm/copyright firms make money.
I don't see that as a reasonable argument unless you're claiming the user was lying:
> In the comments section of her post, she wrote she had simply asked ChatGPT to make “a New Yorker-style cartoon.”
If you commissioned me to record music onto a CD for you, and then I put in the credits that Jimi Hendrix recorded the guitar parts without you asking, it seems pretty reasonable that I should get in trouble for that rather than you.
I don’t think she was lying. She created a cartoon and published it irresponsibly. In my opinion, that makes it her fault. And I think that AI is not going to be found culpable for issues like this in the future. The user needs to understand they can’t just publish whatever it creates. There are open source AI models that would have near zero culpability and in that case I believe very, very clearly it will be the users responsibility. And I think that will extend to paid models as well.
It’s their hardware and their web response. Especially more heinous if it’s being served as part of a subscription. I don’t think it’s functionally the same as me opening Microsoft Paint and recreating pixel-for-pixel a New Yorker artist’s signature.
If it's (nearly) all streamed from disk can you use an intel/amd machine and a big fat gpu and use the SSDs (nvme for faster throughput) as your virtual vram? Or something like that?
In principle yes, and the upstream engine already has a CUDA path with expert streaming and residency (that's theirs, not ours — we only measured on this Mac).
Two things carry over: the experts are read from disk per token either way, and the barrier model — a layer waits for the slowest of its 16 reads — is platform-independent.
Two things don't: the 50 GB resident trunk lives in unified memory here, so on a discrete GPU it would need to fit in VRAM or be streamed too; and a desktop's PCIe lanes let you put NVMe drives on the bus directly rather than behind a ~7 GB/s Thunderbolt enclosure, which is our per-drive wall.
Whether that ends up faster is exactly the kind of thing that wants measuring rather than guessing.
Meh. Chess exists to entertain the players. Coding exists to solve problems. I see way too often the programmers think it’s all about the coder and the code. Solve the problem. Don’t write code at all to do that if you can (AI generated or otherwise).
Yeah, it's about incentives. Software development is about solving business problems, but sometimes the solution is "use this thing that already exists instead of paying me to build it."
Similarly, the incentive is to design it so you will need to spend 10 years working on it, instead of 10 hours.
Think if the user commissioned the art from an outsourced creative shop nobody has heard of. Then they published it. They wouldn’t go after the creative shop, they would go after the publisher.
(I am just addressing publishing here, training on the artist’s works is a different, well discussed issue)