DE version is available. Content is displayed in original English for accuracy.
Advertisement
Advertisement
⚡ Community Insights
Discussion Sentiment
63% Positive
Analyzed from 3621 words in the discussion.
Trending Topics
#models#china#open#companies#more#why#same#government#tech#model
Discussion (108 Comments)Read Original on HackerNews
Ant Group: China halted Ant’s IPO and forced a restructuring
Alibaba: China fined and politically disciplined Alibaba
Didi: China punished Didi after its US listing by removing its apps, freezing users, forcing delisting
Tutoring platforms: banned profit from core school-subject tutoring.
Tencent gaming: restricted youth gaming froze approvals
NetEase and gaming companies: licence freeze stopped game companies from shipping games.
Meituan: fined Meituan and forced changes to its labour and platform model.
Huya/DouYu: blocked Tencent’s game-streaming merger, stopping commercial consolidation in a major entertainment market.
Boss Zhipin / Full Truck Alliance: froze new users after listging in the US
Crypto companbies: banned crypto trading and mining, forcing exchanges offshore.
Think it's not happening to the US?
tourism - people afraid to visit
tariffs - wrecking ball to all businesses
defence - why would anyone buy US weapons after Greenland and Canada
internet clouds - Greenland made Europe decide that the US clouds can't be trusted, now sovereign computing matters and MS/AWS/Google are feeling it
finance - no one trusts the US not to turn people into "non members of global society" by banning them from visa and credit card and banking systems
Political priorities and good governance is why we have government.
- Tutoring platforms were a plague on Chineese youth that increased the weight of their already _very_ heavy load (tbh, i think and education reform might have been preferable, this is a stopgap, but at least it is something).
- Ant group was offering predatory consumption loans to rural China, which to me felt a lot like the "revolver credits" that plagued my country in the 80s and 90s and pushed to many to suicide (the surname cam from their english name, "revolving credit", and because my countryside had a lot of hunting rifle available to whomever). Considering how rural china is mistreaded by Chineese state and general government (and imho this is a real weakness in China politics), having this group by a huge fine for their practice and a general debt forgivness was great. Curtailing Ant's power is also good.
- Stopping consolidation is a great way to keep a market free.
- Crypto companies: mining diverted power from villages who couldn't compete on purchasing power to mining wharehouses in some state. The ban is great for the rural population at least. Also, if that can curtail the birth of Chineese cryptobros, great for the mental health of the country.
Huh? US foreign military sales are up at all time highs
"Total exports by the United States, the world’s largest supplier of arms, increased by 27 per cent. This included a 217 per cent increase in US arms exports to Europe, according to new data published today by the Stockholm International Peace Research Institute (SIPRI)"
[https://www.sipri.org/media/press-release/2026/global-arms-f...]
Why China crushed its tech giants https://www.wired.com/story/china-tech-giants-policy/
Why Big Tech May Never Recover in China https://time.com/6973119/china-big-tech-crackdown-backfiring...
Beijing can’t afford another crackdown on its tech companies https://www.cnbc.com/2026/03/11/china-cant-afford-another-cr...
And even if foreign investors are more cautious now, there is plenty of money trapped by capital controls, so that it doesn't look like new tech companies have trouble raising capital anyway.
"wait, not like that"
Seems it blew in their faces and probably the new frontier models will be available only to a select few. Many people predicted this, only a naive person would believe that access to something with these capabilities would be decided by some dude in California.
GLM 5.2 is competitive with Opus 4.6. If the best model I'll ever get is Opus 4.8, then the choice is clear. I'll miss Opus.
Would you levy the same two quote criticism of the reasonable call for regulation?
If they can’t freely sell access to their models and Chinese models catch up to Opus 4.8/GPT 5.5 in 6-8 months - then why pay OAI/Anthropic at all?
If non US citizens shouldn't have the models - wouldn't that cause both Anthropic and OAI to fire non-citizens?
They would do what the thousands of other companies do with their tens of thousands of engineers handling ITAR/EAR-regulated software/hardware every day: compartmentalize their workforces, buildings, and access.
Dont worry though, the rest of the entire world gets access to better chinese models :-), once they get a taste for those the US has lost their little trade game and the future truly belongs to China.
Its almost like they are serving it up on a silver platter.
ofc they are not, they are just betting all in their models will be better, which is unlikely. (just look at the chinese law and all the names atop of advanced AI papers...)
And given how willy-nilly they are operating I see no reason they won't clamp down on open source. All it takes is someone with connections/political contributions wakes up one day and realizes that open source is a threat to their power or bottom line and it will be declared an imminent threat with no oversight or debate.
Any company providing the models will be deemed a threat to national security.
No need to block the download.
> Any company providing the models will be deemed a threat to national security.
Any company providing specifically-controlled models to foreigners would hypothetically be prosecuted.
Why wouldn't they just tell Hugging Face that they need to abide export restrictions directly - they're an American company?
Doesn't sound dystopian enough without a second compelled entity?
wondering when more people will raise their voice and get engaged
It’s only a matter of time before companies start to acknowledge the huge cost of tokens and look for a cheaper alternative with basic cost-benefit analysis.
My F500 company is getting local infrastructure going to host open models and I’m sure many will just switch to bedrock + the best open models.
It’s foolish for companies to let three companies dictate the price of tokens, I just don’t think they are aware of this now by and large.
IMO AI is different from everything else. It is a weapon as potent as nuclear. It is only natural that it be treated as one.
I'm very pro-west, but at this point okay, I guess the rest of us have to side with China, not because we remotely like it, but because they don't try to be quite so antagonistic to us in everything they do.
Just because the many headed dragon is trying to bite your sailors' heads doesn't mean you should pilot your ship into the whirlpool
- instant, total world war if it's not coming from USA
- let's finish all oil's reserve first otherwise
But it is odd that this administration has learned absolutely nothing about the mid- to long-term effects of export restrictions on other countries' ability to compete with the US.
In all seriousness, I can’t believe the AI firms are abiding by this peacefully. If I truly loved my company, and I felt we were on the bleeding edge of incredible, life changing products, why would I allow my company to be set up for failure by remaining somewhere that clearly wants control over the sovereignty.
The US gov sees these AI companies as bartering power, not as innovation. Wouldn’t you as a parent always want what’s best for your child, not for yourself?
It also feels like they can’t just relocate out of the country, as the administration will surely sanction anyone from business within the country again. These firms are so over inflated with evaluations and opex, they’ve dug themselves into a corner.
This is not to say regulation does not exist in any other country, but it’s clear now after what’s happening at Anthropic + OAI that the US gov has taken these companies hostage.
This is only further playing into the hands of open source and the outside models; the US gov is going to be to blame for when they all lose the race to low cost/free.
Which, like the US, uses export controls when it finds them advantageous: https://nam.org/china-imposes-export-controls-on-u-s-mineral...
> In all seriousness, I can’t believe the AI firms are abiding by this peacefully. If I truly loved my company, and I felt we were on the bleeding edge of incredible, life changing products, why would I allow my company to be set up for failure by remaining somewhere that clearly wants control over my sovereignty.
So, locate in China, where every company of importance is essentially required in practice to maintain ties to the CCP?
I personally think the US has gone too far with its use of export controls and sanctions as a political tool, but it's foolish to believe that it's different anywhere else on the planet.
In China, it has even been reported that top AI talent is restricted from overseas travel.
https://www.thinkchina.sg/technology/china-tightens-control-...
Bottom line: if you're working on cutting-edge technology that is deemed to be of critical national security importance and has military or dual use implications, you're going to be a hostage no matter where you go.
AI has long existed in many countries around the world without this type of behavior from the government. Deepmind in the UK, Mistral in France, DeepL in Germany - the governments don’t seem to be forcing employees to get their deploys approved by a government official.
My argument is that the US gov does not like that these companies have too much influence which they do not feel they can mandate. It’s slowing the entire country down at a very critical sink or swim inflection point in this tech.
-the US is the only place where you can raise the kinds of money you need to run a lab like this.
-a government that won’t let you sell products to customers abroad will probably object even more to you moving abroad.
Even if you made the move abroad, that government might no longer let you access US data centers.
-This basically affects OpenAI and Anthropic, which make the only LLMs most people consider frontier nowadays. Since most open weights models rely on distillation of frontier models, it may genuinely entrench those companies more.
It may be playing into the hands of open source OAI/Anthropic dependencies start to look more dangerous, but it also makes building better OSS models harder.
The advantages the AI labs rely on might be less durable than a proprietary process in industrial manufacturing, but it’s still meaningful.
I think the bigger reckoning will come from a different angle: tokens will eventually need to cover cost.
That will likely mean multiplying prices compared to today. And companies already complain now!
Model orchestration and smaller models that can run locally or cheaply will become more important in my opinion.
Right now, you can still default to GPT/Claude and it’s kind of fine, but that will have to change.
Many Americans want AI to fail. The US gov wants to control AI. The AI companies are running out of things to do, and are shipping product after product after product to keep the perceived productivity narrative alive.
At this rate I would not be surprised to see an OAI/Anthropic merger just to throw everything AI the US has to offer to the global markets.
We’re still in the “5$ airport Ubers thanks to VC money” era of AI
They literally asked for this.
What are they going to do about it? Might makes right.
They've already done what little they can: pull access to their models wholesale rather than adopt an export compliance regime.
DeepSeek 4 is a good model for many tasks, but I think it currently lacks the post-training required to become a genuinely great coding model.
I understand if its military hardware and software, that's the property of the US government however this is the property of a private company.. Now seemingly being commandeered and issued at the will of the government, sounds very Russian/Chinese to me.
Is there a precedent for this before in a democratic country ?
The missiles Raytheon makes are export-controlled too, and they're not somehow "property of the US government" - this isn't China.
Is this just upsetting because it's a product you want to enjoy?
> Is there a precedent for this before in a democratic country ?
Try every weapons system, encrypted radio system, FPGAs with high-bandwidth transceivers, lithography equipment, etc. etc. etc. There's plenty of precedent.
No, infact I'm a proponent of open models and being able to run them locally, it just feels strange that a consumer product would be under the same restrictions as military grade equipment and tech which is specifically designed for warfare.
> Try every weapons system, encrypted radio system, FPGAs with high-bandwidth transceivers, lithography equipment, etc. etc. etc. There's plenty of precedent.
If it's the same equivalent then my issue is just that, it feels like trying to restrict the useage of RSA because it could be used by bad actors.
RSA was practically impossible to control (an implementation is what, 100 lines in any language?) and the global benefits outweighed the cost and futility associated with restrictions.
AI laboratories with hundreds of billions of dollars in funding aren't cropping up in every country in the world, and their products and services are easily controlled and not easily replicated.
> Is there a precedent for this before in a democratic country ?
I'd argue US is not very democratic country given how many of what govt does goes against people's wishes. Same as UK
That could be argued but the core principle is freedom of commerce and private companies get a lot of runway. This seems completely counter to tha.
Being told "no" is never fun, but the regulations are not hard to comply with (despite what Anthropic might have you believe.)
> I can't imagine US partners will abide this for long.
What are they going to do? Start their own Anthropic? Go for it. Why is every other country in the world entitled to American technology by default?
Hopefully open-weight models will catch up, hopefully we, as the people, engineers will find the way to maintain those open-weight models on pair with the closed ones.
I try to be optimistic, as we won some battles, against all odds, Linux is flourishing, open source solutions are mainstream.
https://digital-strategy.ec.europa.eu/en/news/commission-sel...
A consortium will train a 400B-class model and get 2.5% on time of the EuroHPC infrastructure (~2000 PFLOPS datacenters). So, even if the Chinese take away the open source there will be some models. Probably not Mythos quality yet though.
Don't get me wrong: I'm all for open models, but I think it will get more and more difficult to distil-train them without (legitimate) access to frontier models.
If anything I bet these people will just use their knowledge to make even more money reselling tokens.
People have no idea and everybody pretends to be an expert and ignore how good China is on AI research
If you want an Opus class model, it makes sense that you would train on what Opus outputs. But, if you want something better than Opus, training on the same data that Opus was trained on with the same architecture will only result in an Opus class model. Then, if your dataset also contains Opus outputs, many of which are wrong, then it makes sense that the model would have reduced performance.
All this to say that I don't think there's such a thing as a "Model Collapse," but there likely is a "Model Stagnation."