I'm struggling to parse the double negative in that statement, haha.

Are you saying that you believe that untested but technically; models trained on GPL sources need to distribute the resulting LLMs under GPL?

▲

advael 3 days ago | parent | next [-]

Yes. Double negative intended for emphasis here, but apologies if it's confusing

▲

eru 3 days ago | parent [-]

Well, most companies never distribute their models. So GPL doesn't kick in.

▲

vova_hn2 2 days ago | parent [-]

I think that the claim that they make is that once a model is "contaminated" with GPL code, every output it ever produces should be considered derived from GPL code, therefore GPL-licensed as well.

▲

Tadpole9181 2 days ago | parent [-]

So GitHub and Windows and IDEs need to be open source because they can output FOSS code? That's obviously rediculous.

If an AI outputs copyrighted code, that is a copyright violation. And if it does and a human uses it, then you are welcome to sue the human or LLM provider for that. But you don't get to sue people for perceived "latent" thought crimes.

	▲	vova_hn2 2 days ago \| parent \| next [-]
		First of all, I'm not advocating for this claim, I'm merely trying to clarify what other people say. That being said, I don't think that your analogy is valid in this case. > GitHub and Windows and IDEs need to be open source because they can output FOSS code They can output FOSS code, but they themselves are not derived from FOSS code. It can be argued that the weights of a model is derived from training data, because they contain something from the training data (hard to say what exactly: knowledge, ideas, patterns?) It can also be argued that output is derived from weights. If we accept both of those claims, then GPL training data -> GPL weighs -> every output is GPL > If an AI outputs copyrighted code Again, the issue is not what exactly does AI output, but where it comes from.
	▲	eru 2 days ago \| parent \| prev [-]
		It would be relatively easy to scan the output of the LLM for copyrighted material, before handing it to the user. (I say 'relatively easy'. Not that it would be trivial.)

▲

gottheUIblues 2 days ago | parent | prev [-]

If that theory holds - have to ensure that the models have not been trained on any code that is licensed incompatibly with the GPL, in which case the models could not be distributed at all