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

Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

writings.hongminhee.org

Post reply on HN
Print view View original

491–500 of 623 posts

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#491

Post by jph00 »

palmotea wrote: post #455
esrauch wrote: post #365

Earlier quoted context omitted.

It's not at all obvious whether copyright net protects or destroys the little guy. It definitely does some of both, and we have no obvious measure or counterfactual to know otherwise. You also have to take into account not just if optimal reform or optimal dismantle is better, but the realistic likelihood of each, and the risk of the bad outcomes from each. Protect even more conceptual product ideas seems pretty stro…

> It's not at all obvious whether copyright net protects or destroys the little guy. It's super obvious, unless your perspective basically stems from someone who was mad they couldn't BitTorrent a ton of movies. I mean, FFS, copyright is the literal foundation for open source licenses like the GPL. My sense is a lot of the radically anti-IP fervor ultimately stems from people who were outraged they could be sued for…

GPL was created as a workaround for copyright - it wouldn’t have been needed if there wasn’t copyright. There are complex arguments both for and against copyright and there’s no reason to simply assume it must always be just as now even as circumstances change.

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#492

Post by re-thc »

satvikpendem wrote: post #124
re-thc wrote: post #121

Earlier quoted context omitted.

> Blanchard is very clear that he didn't write a single line of code He used Claude to write it. Difference? The fact that I write on the notepad vs printed it out = I didn't do it? > Signing GitHub commit doesn't change that. That's the equivalent of me saying I didn't kill anyone. The fingerprints on the knife doesn't change that.

I'll take a commit authored by someone else and then git amend the author to myself, did I write that commit then? By your logic I did apparently.

> I'll take a commit authored by someone else and then git amend the author to myself, did I write that commit then

I did say co-author didn't I? Even if you added 0.000000001% to something you did so technically, yes.

> By your logic I did apparently

If you take someone's email and forward it did you write that email? Instead of debating that imagine you took a trojan email and forwarded it to someone and they opened it - do you think you'd be held up in any way?

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#493

Post by tosapple »

dogcomplex wrote: post #383
jrochkind1 wrote: post #279

> If source code can now be generated from a specification, the specification is where the essential intellectual content of a GPL project resides. Our foreparents fought for the right to implement works-a-like to corporate software packages, even if the so-called owners did not like it. We're ready to throw it all away, and let intellectual property owners get so much more control. The implications will not end up b…

We should be removing IP law entirely, not strengthening it to cover entire classes of problem even when implemented entirely differently. Same for anyone trying to claim "colorful monster creatures" as innately Pokemon IP. Just because someone climbed a mountain first doesn't mean they own it forever. Nobody should be honouring any of these claims. Nor should we be treating AI models themselves as respected IP. They…

[dead]

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#494

Post by prohobo »

zmmmmm wrote: post #162

The really interesting question to me is if this transcends copyright and unravels the whole concept of intellectual property. Because all of it is premised on an assumption that creativity is "hard". But LLMs are not just writing software, they are rapidly being engineered to operate completely generally as knowledge creation engines: solving math proofs, designing drugs, etc. So: once it's not "hard" any more, does…

From what I understand, LLMs can't really generate anything meaningful that doesn't implicitly rely on the operator's choices. It's hard to make the right novel choices as soon as you leave well-defined problem spaces.

In terms of math and biochemistry the cost of generating candidates has collapsed, but the cost of validating them hasn't.

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#495

Post by motbus3 »

To me, there is a confusion of what "copying" and "using" means.

You can copy the idea and not use the source code. This has been ruled ok many times already and would be quite dangerous if that was not the case.

But this is not what this is. To generate the new program, another program, the AI, must have an input which then becomes part of the program itself. It does not really matter much if the generation does not contain the source code itself or a similar reimplementation. One could rewrite a full version of the Lord of the Rings changing all the words but having the same elements, it would still be plagiarism. No reason to think this is not the case here. It is evident that the source code was the base, hence, this is a derived work.

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#496

Post by vbarrielle »

derangedHorse wrote: post #408
vbarrielle wrote: post #205

Earlier quoted context omitted.

Everyone writes as if he just fed the spec and tests to Claude Code. Ignoring for now that the tests are under LGPL as well, the commit history shows that this has been done with two weeks of steering Claude Code towards the desired output. At every one of these interactions, the maintainer used his deep knowledge of the chardet codebase to steer Claude.

Is this perspective implying that the maintainer might be legally culpable because he, the *human*, was trained on the codebase?

Well I'm implying that someone who's been reading a codebase for 10+ years is the worst person to claim an "independent reimplementation".

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#497

Post by vbarrielle »

marcus_holmes wrote: post #409
dogcomplex wrote: post #383

Earlier quoted context omitted.

We should be removing IP law entirely, not strengthening it to cover entire classes of problem even when implemented entirely differently. Same for anyone trying to claim "colorful monster creatures" as innately Pokemon IP. Just because someone climbed a mountain first doesn't mean they own it forever. Nobody should be honouring any of these claims. Nor should we be treating AI models themselves as respected IP. They…

Good news! LLM output cannot be copyrighted. Everything that an LLM produces is automatically, irrevocably, in the public domain.

Not quite in my opinion. The output of an LLM from a simple prompt falls into the public domain, but if you also give a copyrighted work as input, the mechanistic transformation performed will not alter the original license (same as encoding a video does not change its license).

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#498

Post by scheeseman486 »

throwaway290 wrote: post #469
matheusmoreira wrote: post #401

Earlier quoted context omitted.

> Where does your confidence come from? From the fact that copyright infringement is trivial and done at massive scales by pretty much everyone on a daily basis without people even realizing it. You infringe copyright every time you download a picture off of a website. You infringe copyright every time you share it with a friend. Everybody does stuff like this every single day. Nobody cares. It is natural . > GPL its…

> You infringe copyright every time you download a picture off of a website. You infringe copyright every time you share it with a friend. respectfully yoy have no idea what you are talking about here.

You might be thinking of fair use, but that's an affirmative defence. Every time someone has copied someone elses artwork and modified it into a meme, that's copyright infringement and remains so even if is eventually ruled as fair use. If you make a fair use claim, you don't deny infringement, you make the claim that you were allowed to infringe.

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#499

Post by alex1sa »

wvenable wrote: post #329
peacebeard wrote: post #204

Earlier quoted context omitted.

The big question is: if copyrighted material was used in the training material, is the LLM's output copyright infringement when it resembles the training material? In your example, you are taking the copyrighted material and giving it to the LLM as input and instructing the LLM to process it. Regardless of where the legal cards fall, this is a much less ambiguous scenario.

There's a couple of different issues here that all get mangled together. If you're producing effectively the same expression that's infringement. You draw Captain America from memory, it's still Captain America, and therefore infringement. If you draw Captain Canada by tracing around Captain America that's also infringement but of a different type. When it comes to software, again it's the expression that matters --…

[flagged]

Re: Is legal the same as legitimate: AI reimplementation and the erosion of copyleft

#500

Post by NewsaHackO »

gspr wrote: post #460
NewsaHackO wrote: post #257

Earlier quoted context omitted.

>Has the model really performed an extreme transformation if it is able to produce the training data near-verbatim? Sure, it can also produce extremely transformed versions, but is that really relevant if it holds within it enough information for a (near-)verbatim reproduction? I feel as though, from an information-theoretic standpoint, it can't be possible that an LLM (which is almost certainly <1 TB big) can contai…

> I feel as though, from an information-theoretic standpoint, it can't be possible that an LLM (which is almost certainly It doesn't need to for my argument to make sense. It's a problem if it reproduces a single copyrighted work (near)-verbatim. Which we have plenty of examples of.

Do we? Even when people attempt to jail break most models with 1000s of prompts they are only able to get a paragraph or two of well known copyrighted works and some blocks of paraphrased text, and that's with giving it a substantially leading question.
Post reply on HN