wonderingwanderer comments on Linux lays down the law on AI-generated code, says yes to Copilot, no to AI slop, and humans take the fall for mistakes — after months of fierce debate, Torvalds and maintainers come to an agreement

[–] wonderingwanderer@sopuli.xyz 1 point 6 months ago (3 children)

Disagreeing because applying the argument consistently results in an undesirable outcome isn't objectionable.

I'm not objecting to disagreement, I'm objecting to the attempt to apply my argument to a different situation that it wasn't meant for, and then going on as if that's even remotely what I was saying.

That's not "applying the argument consistently", it's removing context, overgeneralizing the argument, and applying a strawman based on a twisted version of it.

Open-source developers using AI trained on closed-source code and closed-source developers using AI trained on open-source code are two different issues. My point was only intended to apply to the former, because that's what we were talking about. Trying to apply what I said to the former is a distortion of my argument, and not the argument I was making.

And to try to conflate the two is to be allergic to nuance, which is honestly just typical and unsurprising, but if that's the case then I'm done wasting my time on this conversation.

permalink

fedilink

source

parent

hideshow 3 child comments

[–] ricecake@sh.itjust.works -1 points 6 months ago (2 children)

I'm really not interested in the topic. I'm talking because I explained what someone else meant and you started responding as though that was an opinion or argument I was making.

That's not "applying the argument consistently", it's removing context, overgeneralizing the argument, and applying a strawman based on a twisted version of it.

It's really not.
It's not unreasonable for someone to think "developers who use copy written code from AI aren't liable for infringement" applies to closed source devs as well as open, and to disagree because they don't like one of those.
It's perfectly valid for you to also disagree and say the statement shouldn't apply both ways, but that doesn't make the other statement somehow a non-sequitor.

permalink

fedilink

source

parent

hideshow 2 child comments

[–] wonderingwanderer@sopuli.xyz 0 points 6 months ago (1 child)

If you're not interested, then why are you still here saying the same thing over and over again?

It's perfectly fine if someone wants to make a claim that "we should apply the same argument across both situations," and then I would give my reasoning as to why different arguments apply. But that's not what happened.

What happened was, I gave an argument applied to the situation being discussed. Someone else tried to apply my argument to a different situation, in order to argue against a point that I didn't make. And ever since that point, this whole conversation has been going in circles in which you and that other commenter keep arguing as if I'm saying something that I never said, and I keep stating repeatedly that it's not what I said.

And if you read back through this chain, I never said it. I even said I can understand the other point of view, and would probably even agree with it, if that's the conversation we were having, and I said we could even have that conversation, but that the sudden change of topic as an attempt to "score points" against me is not a good faith argumentation style.

Is it a problem if commercial LLMs are trained on GPL code, and then used by closed-source developers to generate proprietary code which potentially contains open-source snippets? Yes, I've never denied that. But that's not what this conversation has been about.

From the start, it's been about open-source developers using LLMs to write open-source code, when those LLMs are potentially trained on closed-source code and may generate snippets closely resembling closed-source code.

Those are fundamentally different situations, and if you can't see that then I can break it down for you in minute detail. But the point I made about the one thing was never meant to apply to the other; and arguing against the point I made as if it was meant to apply to a different situation is a bad faith argument.

permalink

fedilink

source

parent

hideshow 1 child comment

[–] ricecake@sh.itjust.works 0 points 6 months ago

Whoah, I never said I wasn't interested in the exchange, only that I wasn't interested in the topic.
As someone who's extremely insistent that it's grossly improper to make any form of inferences beyond what is literally stated, I'm shocked you would make such a leap!

I think you're persistently confusing me with someone else. I perfectly understand your point, and have never had any doubt about what you intended to say. I never even disagreed with you on the topic.
I clarified someone else's point to you, and you started explaining to me how they made unreasonable assumptions, which is what I disappeared with.

Intellectual property laws apply to open and closed source software and developers equally. When you make a statement about legal culpability for an action by one group, it makes sense to assume that statement applies to the other because in the eyes of the law and most people people in context there's no distinction between them.

No one is unclear that you were only referring to one group anymore. That's abundantly clear.

My point is that you're being overly defensive about someone else making a normal assumption about the logic behind your argument. And you're directing that defensiveness at someone who never even made that assumption.

permalink

fedilink

source

parent