Live data from Hacker News

Things are about to get worse for generative AI

garymarcus.substack.com

301–310 of 769 posts

Re: Things are about to get worse for generative AI

#301

Earlier quoted context omitted.

Search engines are ruled fair use because they use the copyrighted material in a limited way, they provide a public good and they benefit the copyright holder. Generative AI is more or less the opposite of that. It ingests the whole work, generates output that substitutes for the used work and profits the user of copyrighted work to the detriment of the copyright holder. Throw in the fact that it is purley a mechanic…

But transformative use is an exception to copyright. And I think it's going to be pretty hard to argue that the matrix of parameters inside an LLM is not sufficiently transformative from the input image.

If I run a thesaurus over a plagiarised text, it would be a long bow to draw to say that’s “transformative”. I feel like this “oh but it’s transformative” argument is becoming rapidly load-bearing in the context of LLM arguments and I don’t really see nearly enough justification for it.

Re: Things are about to get worse for generative AI

#302

Earlier quoted context omitted.

thats irrelevant since an LLM is not an intelligent entity. Whatever you're arguing about is fiction.

LLMs are clearly a type of lower level intelligence. Intelligence does not require consciousness.

That’s quite a lot of metaphysical speculation for a conclusion that is all but clear.

Re: Things are about to get worse for generative AI

#304
post #13

If I prompt "golden droid from classic sci-fi movie", what else am I asking for if not Star Wars?

If you do "Golden robot holding a lazer gun in a sci-fi setting, cinematic" it will give you a golden robot that doesn't look in the style of C3PO or Star Wars.

"Droid" is actually a Star Wars term [1], and saying you want it from a "classic sci-fi movie" is asking it to reference a real thing that is well known. Reid is intentionally pushing it that way to fill his agenda and these terms are not as generic as he's making out.

[1]:https://trademarks.justia.com/756/52/droid-75652542.html

Re: Things are about to get worse for generative AI

#305

Should not be a problem in the EU. Article 3 and 4 of the „ Copyright in the Digital Single Market“ Directive already regulate this. Summary by Wolters Kluwer: […] Everyone else (including commercial ML developers) can only use works that are lawfully accessible and where the rightholders have not explicitly reserved use for text and data mining purposes. AFAIK they are discussing something like a robot.txt to flag s…

The EU cannot agree that the Do Not Track flag on web browsers is legally binding but big content should be able to create legally binding flags on their websites to avoid scraping of data? Seems odd!

Re: Things are about to get worse for generative AI

#306

Earlier quoted context omitted.

If a child is instructed to read a copyrighted work at school, which later becomes a factor in his own derivative works, he won't be in breach of copyright. Why should other intelligent entities be prevented from reading copyrighted works and gaining whatever there is to gain from those works the way any human might?

If llms are intelligent entities legally equivalent to a human child, then they incur an even more serious legal problem, as we are all in violation of the 13th amendment.

Hey, don’t look at me. I always say please and thank you when I play with LLMs.

Re: Things are about to get worse for generative AI

#307

Earlier quoted context omitted.

People that create art, must eat and sleep under a roof.

and we can definitely feed them. Plenty of resources lying around.

We don't feed or house tons of people right now, despite those resources.

Re: Things are about to get worse for generative AI

#308

As I understood it, the legal precedent for generative AI is the same one that allows google to scrape websites in order to index them for search for the common good. Google also can display cached versions of websites which is the original content of those sites. No one is going to say that google is copyright infringement just because it is showing content from other websites verbatim. So I think this is a weak arg…

Wonder. Do Cliff Notes have to pay royalties to the underlying material?

Cliff Notes contain quotes, and citations.

Does the cliff note company, when producing Cliff Notes for "Into The Wild", pay royalties to the publisher?

For that matter, does any paper, article, etc.. that may contain a quote from another, have to pay royalties to the source of the quotes?

Re: Things are about to get worse for generative AI

#309
post #298

Earlier quoted context omitted.

> No one is going to say that google is copyright infringement just because it is showing content from other websites verbatim Journalists [1] and Getty Images [2] did in the past [1]: https://yro.slashdot.org/story/03/07/14/025216/web-caching-g... [2]: https://www.theguardian.com/technology/2016/apr/27/getty-ima...

And lost, if memory serves.

No, Google agreed to a licensing agreement and removed the direct links to the images.

Re: Things are about to get worse for generative AI

#310
post #77

I might be a bit idealistic, but I've always believed that the core purpose of art and publishing should be to influence culture and society, not just to make a heap of money. That's why I feel original work needs its protection, but it should enter the public domain much sooner to fuel creativity and inspiration. We should be thinking in terms of a few years for this transition, not decades.

So what do you suggest artists have for dinner.

Let's advocate for robust protections and support systems for artists, ensuring they can secure a sustainable and comfortable livelihood from their creative work.

Once they hit the tipping point of broad cultural absorbtion (think Banksy) AND/OR raking in absurd amounts of cash, move their IP into the public domain more aggressively (think Disney, NYT, etc.). How exactly this would work should be debated.

They'd still own the IP and have all the rights to use it commercially, but other's would be able to use it as inspiration, remix and maybe even resell it if attributed (or cheaply licensed).

In other words: "IP-Tax" the unproportionally successful.

Post reply on HN