Why isn’t there a competitive US or European open weights model?
The Chinese open weights models are copies of the American closed ones. Nobody in the US is trying to do that, I guess.
Startup founders urge U.S. government not to shut off Chinese open weight AI
521–530 of 983 posts
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#522Why isn’t there a competitive US or European open weights model?
Europe is actually catching up right now. Not at the frontier like kimi and qwen 3.6 (yet) but it's just a matter of time until the European A3B models catch up, too.
Apertus and Soofi, both open source (not only open weight) models:
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#523I‘m not even sure what the argument for banning Chinese models/open weights even is supposed to be? 1. if it’s to stop hackers doing hacking things with „uncontrollable models“ then, well… they’re already doing something illegal to begin with, why would they care about breaking another law running these models? 2. if it’s to stop foreign actors, then that ban would not apply to them anyway 3. it’s not stopping distil…
How would it even be stopped if you could ban it? It's not like you're banning just one company. Anyone with the resources can still obtain and run the models. Sure, big models like Kimi K3 will be less accessible until a foreign version of Openrouter with no incentive to listen to US law begins offering it. Or they offer it and obfuscate what model it actually is behind codenames. It wouldn't surprise me if that kin…
The interest to target large corps who pay for LLMs. I don't think Antrhopic/OAI care what you or me are running in basement.
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#524Earlier quoted context omitted.
Suppose a foreign actor deliberately builds an unaligned model, and dumps it at a very low cost through subsidies. Since the cost is lower, everyone integrates it into their business. But now the foreign actor has control; perhaps the model checks for instructions to execute, perhaps it has backdoors, perhaps... who knows? I don't have a view on whether this is a sufficient argument to ban models from potential adver…
So some kind of „sleeper agent“ model? Seems a bit far fetched, given that I’ve never seen any PoC/research on this topic. Got any pointers? To be fair the fix to this seems to be that American labs would need to adjust to the new „fair market price“ then, right? Subsidy-backed competition is still legitimate competition in capitalism (e.g. Uber). And you could still ban specific, known malicious models/labs instead…
It is much more difficult to embed a "sleeper agent" in an LLM than it is to do the same in any closed-source program. Any proprietary program, e.g. from Microsoft, is much more likely to contain hard-to-detect "sleeper agents", than any LLM.
An LLM is a collection of inert data, it is not an automaton.
You obtain an automaton only using an inference program, like llama.cpp or anyone of the many others, and then using some harness around the inference program.
You can put in the harness all kinds of guards, to detect any possible harmful action and prevent it.
In something like a CPU from Intel or AMD it is easy to put some undetectable backdoor that would allow the remote control of the computer, unless you disconnect any antennas and you filter all wired network traffic through an external firewall that would allow only whitelisted connections.
On the other hand an LLM is a much less plausible attack vector, as long as you host the inference yourself. Only an LLM that runs externally, e.g. at OpenAI or Anthropic, can be really dangerous.
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#525Earlier quoted context omitted.
Sanctions would cancel the whole business rationale for the Chinese government having/supporting open weight models. You can't commoditize your competitors with open-source if your product is blocked.
The open weights models would be even more effective by incentivizing people and companies to switch jurisdictions. It might even break SFs monopoly on AI, or at least weaken it, everybody isn’t there exclusively to funnel money to openAI and Anthropic. It would also boost research in non US jurisdiction. Who is going to be wooed by “come to our lab where you’re only allowed to work with closed models!”
What researchers want to join a company that is in a race to the bottom to serve inference tokens? If any open weight companies start actually producing state of the art models then maybe, but if it's just copying what others have done for cheaper, you aren't going to find many researchers interested in that
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#526As bad as it will be no one will be surprised when this happens. This administration does not care about the law. The president wants to keep using his position to make a generational bag. This is the same leader of the free world peddling shitcoin. Everyone knows the game someone is going to cut him a check and suddenly the Chinese models are going to be a national security threat. The only saving grace may be that…
Everything he does benefits his adversaries because they're all playing him like a fiddle because he's basically the dumbest person alive.
I don’t think he’s dumb, I think he’s very cynical and only does things he benefits from. He’s not dumb for selling out the country because he doesn’t care about the country, he’s accomplished his goal of enriching himself.
I can’t find it now, but I could have sworn Trump made a comment one time where he said something very roughly like, an earthquake can just happen and a million people in India die. And he was making a point about the senseless chaos of life, and to me it seemed like a glimpse into one of his core beliefs.
He doesn’t care if his actions lead to the deaths and suffering of thousands of people, he doesn’t care about democracy or the rule of law, he doesn’t care about productivity or building towards a utopia. I think he’s a deeply cynical man who sees no point to anything besides for enriching himself during the brief time he’s here. To that end, his actions aren’t dumb because he’s been pretty successful.
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#527Blocking Chinese models would just be regulatory capture by another name. If we want OpenAI and Anthropic to improve they need to compete not just with eachother but with foreign labs too. I don’t know why anyone would think giving Mango Mussolini the power to decide what AI models they can use is a good idea.
Could foreign labs compete with openAI and anthropic if they weren't literally just copying their outputs?
Can the foreign labs actually advance the state of the art if they put openAI/anthropic out of business?
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#528"Nooooo!!! We already laid off half our workforce and got priced out of models from American companies. We need the cheaper models, even risking giving all our data to a foreign adversary, so our big number goes up long enough for us to get acquired!!!"
It will come for you too, fool. You obviously have no fucking clue what’s the deal with open-weight models. Can I run Fable on my computer? No. Can I run Depseek 4 or glm5.2 on my computer? Yes.
> It will come for you too, fool.
Thankfully, I'm not in a position in which any of this will negatively impact me. But, please, keep seething.
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#529Earlier quoted context omitted.
> funding the conquest of Ukraine It is well known that China is still engaged in trade with Russia- oil and gold going one way, drones and weapon components going the other. But you said funding, and I assume you meant things like grants or loans. Do you have any sources for that? I could only find one that was attributed to indirect open source intelligence, nothing specific.
Trade is funding.
Russia has been trading with a number of countries[0]. Mostly China, but also significantly with India, Turkey and the EU, and in much lesser amounts with 30+ other countries, including the US, the UK, UAE, Brazil, Japan, South Korea, etc.
I don't know if the above data helps to make your original point or not. Would you say that all those countries are funding the Russian war against Ukraine? Perhaps they do so for a combination of various reasons - energy and food needs, business profits, other self-interests, don't care, etc.
Wouldn't a country pushing for global conflict - really global - sow disorder by doing things like: break up alliances, threaten annexations, back illiberal poitical movements, disrupt shipping, disrupt free trade, disrupt energy supplies, start and foment regional wars, etc? Given that all countries tend to do things like this to one extent or another, how many of these things has China done, and how does that compare to other similarly capable countries?
I would say that China is asserting itself as a regional power, like with a navy that threatens Taiwan but can't project power much further than that, wouldn't turn into a global conflict unless other global powers feel so threatened they make it one. Maybe just a play on fear, for power and profit. War-mongering. Wouldn't be the first time.
EDIT: And look at how China has been "funded" China (via trade) over the last three decades... ironic!
[0] https://www.bruegel.org/dataset/russian-foreign-trade-tracke... (summary imports/exports by country in stacked graphs at bottom of page)
Re: Startup founders urge U.S. government not to shut off Chinese open weight AI
#530I‘m not even sure what the argument for banning Chinese models/open weights even is supposed to be? 1. if it’s to stop hackers doing hacking things with „uncontrollable models“ then, well… they’re already doing something illegal to begin with, why would they care about breaking another law running these models? 2. if it’s to stop foreign actors, then that ban would not apply to them anyway 3. it’s not stopping distil…
Suppose a foreign actor deliberately builds an unaligned model, and dumps it at a very low cost through subsidies. Since the cost is lower, everyone integrates it into their business. But now the foreign actor has control; perhaps the model checks for instructions to execute, perhaps it has backdoors, perhaps... who knows? I don't have a view on whether this is a sufficient argument to ban models from potential adver…
> an unaligned model
Beyond certain obvious 'third-rails' even experts have reasonable disagreements on what 'aligned' means.
> everyone integrates it into their business
Few large corporations would sole-source integrate a closed model developed in a country designated as a foreign adversary (which could cut off access at any time) or which their own country might restrict.
> now the foreign actor has control
Your concern is one reason why the Chinese are making most of their models open weight.
> it's not as trivial as "regulatory capture".
You're right, it's not just regulatory capture but the vested interests trying to achieve regulatory capture are certainly leveraging concerns like yours to deceptively achieve that capture. In the case of an open weight, domestically-hosted model, the threat vector you're concerned about isn't a threat in the same way as a Chinese-made, closed source data center router or 5G phone switch.