Comment by khalic
1 day ago
Given the US government’s latest stunt with Fable, this is looking more and more like the future.
Can’t rely on strategic products if they’re gated by capricious actors.
Open weight models are basically immune to that
1 day ago
Given the US government’s latest stunt with Fable, this is looking more and more like the future.
Can’t rely on strategic products if they’re gated by capricious actors.
Open weight models are basically immune to that
> Open weight models are basically immune to that
Somewhat. The US Gov can make it illegal to transact with, download, use, etc. foreign open weight models.
Of course, enforcement will be difficult for individuals (businesses will comply by default, and they would all be pulled off Github and other US based hosting locations if they went the sanctions route). But, we are also quickly going down the road of frightening levels of mass surveillance, which could aid enforcement.
The Fable situation sets a very dangerous precedent, and I'm not looking forward the future here. We are losing the fight for information and computing freedom.
I think that this is what OpenAI/Anthropic want but they wont say it publicly. The will be OK with the US banning regulating and banning open source models as it let's Anthropic and OpenAI charge huge premiums to American business clients for their models.
Also the marketing of them getting to say "our models are so dangerous" only a few companies or select users are allowed to use (benchmark) them would help keep their valuations high.
> I think that this is what OpenAI/Anthropic want but they wont say it publicly.
Won't say it publicly? Anthropic is openly and explicitly saying it publicly. Here: https://darioamodei.com/post/policy-on-the-ai-exponential
> AI companies that develop advanced AI models must have strong security standards that protect their model weights
If the model is open-weight then there's nothing to protect, so the only way to fulfill this requirement is to make open-weights illegal.
Since I am not familiar with the law, can you expand on the mechanism by which the US government could making downloading openly licensed files illegal? How would the government avoid denying people their first amendment rights by doing this?
There's a few different levers they can pull, most of them economic & commerce. IEEPA and OFAC sanctions primarily.
They don't have to criminalize the act of downloading open weight models to effectively block access (to foreign open weight models, they have less levers to pull for US based models).
With sanctions and commerce rules though, they can unilaterally prevent all US based businesses from hosting & using them. They will need to be pulled off huggingface, github, gitlab, etc. ISPs could be put on the hook for folks torrenting them as well because technically that could be considered providing serivces to a sanctioned entity. There doesn't need to be monetary exchange.
Likewise, they can use export controls & sanctions to prohibit US companies and individuals from contributing to foreign open source projects as well.
If it went to court, the DOJ would argue that model weights are not speech because it is machine-readable parameters, and not used as a medium of human communication like source code.
Lastly, first amendment rights are unfortunately not absolute since the PATRIOT act. US Gov just has to declare a national security threat and all your rights go out the window.
I mean my state has been making it illegal to download 3d models of pieces that could be used to make guns in a 3d printer
It’s a very broad law and likely not legal, but it’s going to take a long time to be fought through the courts, and in the meanwhile people will probably be arrested for creating or sharing a file for something that may be able to become a gun part.
You’re correct that it shouldn’t be a thing but unfortunately American society is not in a good place right now
Just like we can’t allow Chinese EVs in the USA, because we can’t and don’t want to compete.
VPN usage would go up, to get the banned models.
In the United States it’s illegal to sell Chinese EVs. It’s also illegal to download copyrighted music and movies. Which one do you suppose illegal open-weight models would more closely resemble?
2 replies →
I would so download a car.
Imagine that, people using VPNs to access data inside of China instead of the other way around.
Maybe, but the world and the internet isn’t just the US.
Businesses outside of the US, like the EU, might have significant competitive advantages.
I doubt it, you can easily distill it into "made in USA" model. They're MIT after all.
A lot more expensive thought, but the added benefit is that you can train on your companies data improving performance of the model.
Not if the US is banning capable models. It’s open source so you wouldn’t need to distill anything.
> Somewhat. The US Gov can make it illegal to transact with, download, use, etc. foreign open weight models.
Presumably you mean in the USA (otherwise foreign means nothing)
Yes, obviously. The US has no jurisdiction outside of the US (except for economic sanctions, which the US could in theory put sanctions on other countries that use models from sanctioned countries).
Oh, you think? The US have a habit of imposing sanctions on companies that don't respect their prohibitions- so for example they decided that companies cannot offer services to a certain EU citizen in the EU otherwise they'll be in a sea of troubles. In theory, imposing these so called "secondary sanctions" is against international law; in practice, the EU is so spineless that doesn't even dare to protest.
2 replies →
You seem to forget that there is a rather large world outside of the US - and we very much would be better off with non-gated, open weights models.
One more entry in https://en.wikipedia.org/wiki/Illegal_number
Honestly, banning SOTA LLM services is the best thing the US could do for AI.
It’d force people to run inference locally, and that’d expose the actual $/perf of the models instead of keeping it secret then propping it up with circular revenue and blatant securities fraud.
If we don’t do something like that, we won’t have much of an AI industry post-bubble.
Anyone else remember solyndra?
It’s very likely the Chinese go dark too the second they have parity / lead
It's a very valid point, though being cut off from future development is nothing compared to losing current operative capabilities
I wouldn't bet on it. Chinese live the free market ideals instead of just preaching them but rent-seeking and seeking regulatory capture at the first opportunity. In China business doesn't control politics. Dynamics is completely different and so might be the outcomes.
Well I do hope you're right - that's a brighter future for all
2 replies →
You criticize the government, perhaps rightfully, but give Anthropic a pass. They are the ones fueling this bullshit. Downgrading your results without telling you. Refusing your requests in the name of “safety”. Even if the government didn’t make them pull the model for foreigners, we’d still be in a really shitty situation because Anthropic is really shitty.
I don't criticize based on vibes. The US government is overreaching, seemingly as a retaliation for Anthropic's refusal to let the US use a jailbroken version of their software in autonomous lethal systems. Hegseth is like a drunk vindictive ex