> If scrapping content is legal, model distillation should be legal too.
No, because legality should be determined by what's in the best interests of Athropic and OpenAI's business models.
Hopefully they're working on RLHF their models to insert clauses making that reality clear into any legislation their models generate or review. That way it's only a matter of time until the confusion is cleared up.
I suppose model distillation is technically legal, in terms of copyright, because LLM output is automatically public domain.
It's only "illegal" from a standpoint of breach of contract given its against the terms of use/service, which is to say its not illegal at all, there's no criminality there.
Yeah I considered whether I should use the term "illegal" in my original post, but in this case, I believe these models are actually banned for use in China, right? Like there are probably export controls (at least with the NVidia chips)
I honestly don't know ... yeah if it's just technically a terms of use violation (which isn't illegal, just a violation of one company's rules, for which Anthropic has every right to stop), or do we now have export controls applied from the various government actions, etc making them truly illegal now.
There are so many China born Chinese employees at Anthropic and OpenAI and I think quite a lot of them have already been recruited as spy . So it is almost impossible to keep secrets from Chinese government.
At what point though doesnt somebody stand back and say "wow, thats really dumb!" I think its probably more an indication of a dev having too much time on their hands rather than being in a hurry.
If scrapping content is legal, model distillation should be legal too.
> If scrapping content is legal, model distillation should be legal too.
No, because legality should be determined by what's in the best interests of Athropic and OpenAI's business models.
Hopefully they're working on RLHF their models to insert clauses making that reality clear into any legislation their models generate or review. That way it's only a matter of time until the confusion is cleared up.
I suppose model distillation is technically legal, in terms of copyright, because LLM output is automatically public domain.
It's only "illegal" from a standpoint of breach of contract given its against the terms of use/service, which is to say its not illegal at all, there's no criminality there.
Yeah I considered whether I should use the term "illegal" in my original post, but in this case, I believe these models are actually banned for use in China, right? Like there are probably export controls (at least with the NVidia chips)
I honestly don't know ... yeah if it's just technically a terms of use violation (which isn't illegal, just a violation of one company's rules, for which Anthropic has every right to stop), or do we now have export controls applied from the various government actions, etc making them truly illegal now.
There are so many China born Chinese employees at Anthropic and OpenAI and I think quite a lot of them have already been recruited as spy . So it is almost impossible to keep secrets from Chinese government.
> steal the models or illegally distill them
The irony.
At what point though doesnt somebody stand back and say "wow, thats really dumb!" I think its probably more an indication of a dev having too much time on their hands rather than being in a hurry.
Not totally new territory; there was a highly compressed period of panic about encryption 35 [0] years ago:
https://en.wikipedia.org/wiki/Pretty_Good_Privacy#Criminal_i...
[0] f**k I'm old
> steal the models or illegally distill them
Oh no, they're trying to steal the models that were trained on stolen data? That's horrible, I feel so bad for Anthropic.