Hacker News

points by iamnothere 1 day ago | 0 comments

It’s possible that they have different interpretations of how generated code affects licensing, but I think in general everyone is now abandoning old positions on licensing. It’s an exciting time for those of us who dislike copyright, something is bound to shift soon.

Whatever its stance towards the GPL, Linux has become the default home of GPL land with the exception of the few dedicated souls still working on Hurd. (I hope they keep making steady progress.)

[–] flossly 1 day ago

Linux (Linus T himself iirc) is very GPLv2.

They are not onboard with the innovation that was GPLv3. That innovation was mostly about software patents iirc. But the result was also that big corps liked GPLv3 even less than GPLv2.

On top of that it was going to be very hard (near impossible) to get all devs to agree to a v2->v3 license bump.

[–] jraph 1 day ago

> They are not onboard with the innovation that was GPLv3. That innovation was mostly about software patents iirc

I think what Linus is against in the GPLv3 is mostly the anti-tivoisation mechanisms. He wants tivoisation as a feature for the Linux kernel.

[–] voakbasda 15 hours ago

Some contributors were already dead by the time the decision would have come to a head. Getting them to sign an agreement would have been challenging.

[–] NoidFonsense 17 hours ago

As a gplv3 advocate, I disagree. Anti-Tivoisation is thr main innovation of gplv3

[–] ender341341 1 day ago

> On top of that it was going to be very hard (near impossible) to get all devs to agree to a v2->v3 license bump.

My understanding was this was never a problem, almost all of the kernel was licensed as "GPL2 or newer" until GPL3 was released and they removed the "or newer" from new releases

[–] tredre3 1 day ago

You're mistaken. Here's a direct quote from Linux Torvalds dated 2006:

    The "version 2 of the License, or (at your option) any later version"
    language in the GPL copying file is not - and has never been - part of the
    actual License itself. It's part of the _explanatory_ text that talks
    about how to apply the license to your program, and it says that _if_ you
    want to accept any later versions of the GPL, you can state so in your
    source code.
    The Linux kernel has never stated that in general. Some authors have
    chosen to use the suggested FSF boilerplate (including the "any later
    version" language), but the kernel in general never has.
    In other words: the _default_ license strategy is always just the
    particular version of the GPL that accompanies a project. If you want to
    license a program under _any_ later version of the GPL, you have to state
    so explicitly. Linux never did.

https://yarchive.net/comp/linux/gpl.html

[–] ender341341 6 hours ago

Thanks for clarifying, I'd been misunderstanding that for 20+ years.

[–] asveikau 1 day ago

I think what you're finding is that people who don't care about low quality of LLM outputs also don't care about licensing. If you're in the faction that supports all of this, you don't see a lot of people around you still caring about licensing or quality.

[–] iamnothere 1 day ago

I am in neither faction, as I already indicated. I’m not an LLM user except for search engine summaries.

I also already mentioned that some formerly GPL die-hards have been jumping ship from Linux to BSDs simply because of the respective stances on AI. That’s enough for me to say this works both ways.

[–] g-b-r 1 day ago

The prospect of being sued because the LLM outputted training code you're not allowed to use, or to relicense, is exciting?

Or to be forced to change license to large portions of your project?

With Linux I meant the Linux kernel

[–] iamnothere 1 day ago

Copyright is based on laws, and laws can be changed or clarified. Given the amount of overinvestment in AI, I expect them to change. The EU is already making moves in this direction: https://noyb.eu/en/ai-eu-member-states-approve-digital-expro...

I expect the overall copyright regime to be weakened over time due to the effects of AI. It’s clear that the intent is to replace copyright gatekeeping with model/compute gatekeeping. This is a battle between two groups I greatly dislike, and I don’t really side with either, but there may be a small benefit for the average person who just wants to “download a car”.

[–] g-b-r 1 day ago

It could be, but:

- For now the laws are in force, and I don't even know of discussions to change them (that EU proposal only wants to allow training on personal data)

- changing intellectual property rights law would typically require changing international treaties

- however much companies are in love with AI now, I'm not sure they'd support abolishing legal protections on which their business relies

[–] ethin 1 day ago

> The prospect of being sued because the LLM outputted training code you're not allowed to use, or to relicense, is exciting?

The problem with this is that if someone sues you for this they have to somehow prove that you actually infringed. They can't just point to the code and say "See, mah, infringement!" The legal system does recognize that it is very possible for two people, completely independently, to arrive at a nearly identical solution.

[–] g-b-r 1 day ago

Of course we're talking of code for which that's implausible.

I think in most cases twenty lines of identical code are more than enough for that to be the case (at least if it contains a few non-single-letter identifiers).

[–] ethin 1 day ago

I mean, if it uses the exact same identifiers and maybe even comments, then sure, that's pretty good evidence. However, 99.99999999999 percent of the time this will not be the case. As in, the super-super-super-majority of LLM-generated code is customized for the specific use case it's being employed in. As such, claiming any kind of infringement, let alone proving it, will be insanely difficult.

[–] g-b-r 1 day ago

We're talking of the case when it happens (and it's detected).

I don't know if that's in the 0.000000000001 percent of the time, but it can happen.

[–] graemep 1 day ago

How many cases of an LLM outputting training code have you heard of? let alone training code that you are not allowed to use or relicense?

Why are proprietary software developers entirely unbothered by this?

[–] mschuster91 1 day ago

> Why are proprietary software developers entirely unbothered by this?

Because the chance of someone auditing the actual source code is near zero, but for open source software it's near 100% that someone or, worse something aka an AI bot looking for a revenue stream, will have a look at your code. And EVEN IF you get caught stealing by whatever random incident (say, your code repository being 0wned) - it's a matter of a settlement payment. If you are not a company, but an individual person... well, just look at what happened to Aaron Swartz RIP.

And for the WINE/ReactOS/console homebrew folks, it's vital to be able to claim no code contributor ever got exposed, even indirectly, to proprietary code.

[–] graemep 1 day ago

On the other hand it could be very expensive if caught. It could be very difficult to remove the code too.

Your point about being caught is valid, but it depends on AIs producing identifiable chunks of code, and that is incompatible with your license. AI training is going to lean towards open source, because its available to train on.

I also think the widespread use of AI will mean accidental copying will become more acceptable. Even as it is, its going to be regarded as an accident.

[–] g-b-r 1 day ago

I heard many cases of LLMs outputting their training data verbatim; they sure don't check that such data can be legally used by you before doing it.

As others said, for closed source software it's a lot harder to detect plagiarism.

And it's in general very hard to detect plagiarized code; there's some product that does it, but it of course has only a partial inventory of the existing open-source code.

I think that some big company does use those products, though.

[–] graemep 22 hours ago

> I heard many cases of LLMs outputting their training data verbatim;

I have heard of this happening in response to prompts designed to do so, not inadvertently.

How many case of proprietary code inadvertently being included in open source projects, or even of copyleft code being included in permissively licensed projects do you know of?

I can just imagine it happening to a purely vibe coded project that lets an LLM do huge chunks of stuff with minimal instructions doing something like copying a chunk of a library, but its otherwise hard to just copy code that will work in a different context.

We read of huge LLM assisted projects on HN all the time. How many of those have been sued?

[–] ltph0 1 day ago

They did in 2023 and the AI thieves built logic to rephrase the plagiarism so it isn't immediately detectable.

Proprietary developers don't care because their code wasn't trained on. Is that a trick question?

They do care now that reverse engineering is possible.

[–] graemep 1 day ago

Can you explain further? What was the code affected?

[–] Vaushite 1 day ago

You're being pretty, I'm not sure the right word, scaredycat?, about licensing. Licensing doesn't matter unless somebody sues you, which they won't in most cases. The AI era has exposed this truth that you're extremely unlikely to get sued about copyright violation - so unless you're the kind of person who avoids car trips because you could crash, you can ignore it.

[–] g-b-r 1 day ago

Uhm, you're being pretty, careless?

I'm not sure that that's how legal departments think.

[–] jeremyjh 1 day ago

Legal departments definitely think about their practical exposure. That is because they are employed by businesses.