Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Penguin Random House underscores copyright protection in AI rebuff

thebookseller.com

Post reply on HN
Print view View original

51–57 of 57 posts

Re: Penguin Random House underscores copyright protection in AI rebuff

#51

Post by polotics »

kranke155 wrote: post #38
polotics wrote: post #35

Earlier quoted context omitted.

You wrote "thinking machines"? seriously? You do know that attention is all you need, right? ( https://en.wikipedia.org/wiki/Attention_Is_All_You_Need ) Having a Generative Pretrained Transformer regurgitate pattern-matched input into mostly logical-appearing output does not make a thinking machine! ( https://en.wikipedia.org/wiki/Generative_pre-trained_transfo... ) case in point, look at this, and think: """ It's tr…

I can spot a pattern that people who think AI should be able to violate copyright often anthropomorphize it as a silly defense. If you mean what I wrote, it's irrelevant what you call them. If governments think there is an AI race that they will lose if they enforce copyright, they won't enforce copyright.

ok, good point!

Re: Penguin Random House underscores copyright protection in AI rebuff

#52

Post by portaouflop »

dartharva wrote: post #46
portaouflop wrote: post #16

Penguin Random House is a predatory corrupt business that the world would be much better off without. If AI means businesses like this shut down I need more „AI“

Scorched earth is never the way.

Why not? Plenty of examples in history where it was the way - and if it was never a viable strategy it wouldn’t have a name.

I’d rather see no one profiting from producing culture than orgs like Random House.

Re: Penguin Random House underscores copyright protection in AI rebuff

#53

Post by lelanthran »

AStonesThrow wrote: post #4
gruez wrote: post #2

IANAL, but this seems kind of pointless? Either AI training is copyright infringement or it's not. A tiny disclaimer isn't going to affect that. The AI companies contend that it's fair use, which would probably override any fine print on the inside cover.

> The AI companies contend that it's fair use Do they? "Fair Use" is an affirmative defense, so the only time we're going to get into that is in a court case, where it'll be tested through legal means. I would say it's even more nuanced: if LLM training involves merely reading a dataset, but it is not strictly necessary to copy , or even store it verbatim to be useful, then does it even fall under copyright protectio…

> It is a strange copyright protection indeed that explicitly or implicitly prohibits certain types of machine readings, while allowing many others.

As far as I know, licenses can discriminate on whatever constraint they want to[1].

A license that is basically "This work is for $FOO only. If you want a $BAR license please contact us." is perfectly legal right now!

There is no restriction in law that $FOO cannot be "human consumption" and $BAR cannot be ""machine consumption".

If the AI companies are not arguing the "fair use" argument, then they are arguing for AI companies having a special exemption for themselves carved out in law whereby a copyright holder is forced to license their works to the AI companies specifically.

IOW, their only recourse is to argue "fair use", because any other argument boils down to "please make a special exemption for us in copyright law, defying hundreds of years of precedent", which is a particularly hard sell.

[1]Excluding discriminating against protected classes.

Re: Penguin Random House underscores copyright protection in AI rebuff

#54

Post by CaptainFever »

NeoTar wrote: post #15
CaptainFever wrote: post #5

If it's fair use in the US, it doesn't matter how many "X is prohibited because of copyright" clauses they add, AI training is allowed. For EU TDM, this opt out probably works to disallow commercial entities from using it for training, but not researchers. For SG TDM, this opt out can be ignored. All legally-acquired material can be used for AI training purposes, period. Disclaimer: IANAL.

SG = Singapore, TDM = Text Data Mining?

You're correct. Apologies for the vagueness.

Re: Penguin Random House underscores copyright protection in AI rebuff

#55

Post by CaptainFever »

rahimnathwani wrote: post #7

Which country has the most favorable 'fair use' laws, and why wouldn't big companies train their models there?

In the context of AI, it's possible that OpenAI is already doing this. From what I know, Singapore has the most pro-AI law currently, where any legally-acquired material can be used as training data, and you can legally ignore all opt-outs (e.g. in the ToS, noAI, etc). OpenAI has recently opened an office there. I'm not sure if it's related, but it's possible.

Maybe this is why LAION was in the EU, too? They have a similar law too, though from what I know they require commercial entities to respect opt-outs.

(Related to sibling comments; Japan seems to have one too.)

https://store.lawnet.com/blog/post/understanding-the-text-an...

Re: Penguin Random House underscores copyright protection in AI rebuff

#56

Post by CaptainFever »

kranke155 wrote: post #34

Always happy to see the copyright experts in HN jumping out of the woodwork whenever a thread like this shows up. Whether AI can use copyrighted material as training data is legally undetermined. There's only an argument that it could be. It will be decided in court.

> Whether AI can use copyrighted material as training data is legally undetermined. There's only an argument that it could be. It will be decided in court.

Note that this is U.S. specific.

https://store.lawnet.com/blog/post/understanding-the-text-an...

Re: Penguin Random House underscores copyright protection in AI rebuff

#57

Post by kranke155 »

CaptainFever wrote: post #56
kranke155 wrote: post #34

Always happy to see the copyright experts in HN jumping out of the woodwork whenever a thread like this shows up. Whether AI can use copyrighted material as training data is legally undetermined. There's only an argument that it could be. It will be decided in court.

> Whether AI can use copyrighted material as training data is legally undetermined. There's only an argument that it could be. It will be decided in court. Note that this is U.S. specific. https://store.lawnet.com/blog/post/understanding-the-text-an...

It seems to me that a lot of countries in the western sphere just follow US law, even if indirectly. So I think the US judgement on this will be decisive.
Post reply on HN