If this really is a war, trump is kneecapping the country with his lawlessness and eroding America’s good will. If the world cannot trust China with their data and they cannot trust the U.S. to provide good reliable service and not turn it into a mafia style negotiation, then winning the AI war is not helping the U.S. countries as much as it potentially can. It’s probably a good thing for more capable areas like Europe which may develop their own tech stack.

In a weird way because the AI stack is so expensive, China helps the world much more than the U.S. with their really capable open source model.

>The US is winning the AI race where it matters most: commercialization

If you ask me, one could name different criteria for winning, and commercialization would not be the first thing to come to my mind:

https://english.www.gov.cn/news/202604/15/content_WS69df29e6...

https://fortune.com/2026/05/03/chinese-court-layoffs-workers...

https://www.reuters.com/world/china/china-moves-regulate-dig...

> It also owns platforms that generate and organize the data of the AI age. YouTube is a video corpus. Google Drive and Microsoft 365 sit inside daily office work. GitHub sits inside software development.

Yeah, okay. China does not have any platforms nor data.

What's the point of leading the race for 90% of it, if they're gonna slip on their own sweat and fall down by the end? In non metaphorical terms, what's the point of spending billions of dollars rushing to get the best AI tech at all costs, when the competition can distil your progress and catch up in 6-12 months while only spending 1% of what you spent.

Even in the aspect the article cares about, commercialization, the US is starting to lose marketshare, I've seen people move from cc/codex plans to use glm/opencode plans due to the recent squeeze the US companies put on plan usage, the US companies are screwed if that sticks, not everyone needs the bleeding edge models, they just want to pay $20/month and have the models be decently capable.

Anthropic, OpenAI and Mistral are just companies that are making money right now (still not profitable), but will lost their tractions and values in the long term.

However, I am more appealing to see how OpenCode Go subscriptions will go in the future: cheaper than big techs, more tokens, and they don't train on our data to (try to) improve...

I understand that America dominates in distribution, integration, enterprise contracts, ecosystems, infra... The article isn't wrong, it's just that that dominance is fragile and requires constant upgrading.

But what is the point of that if you have to infinitely scale because the opposition is right behind you at all times ready to usurp you... You CANNOT scale infinitely, the VC money will run out at some point and then everyone will have to downscale everything to meet the real costs associated with SOTA models, they'll have to be able to use subscriptions, and other monetization to cover those insane costs, we just saw SORA shut down because it was bleeding money far too fast while the Chinese released video models that far surpassed it back to back to back...

EDIT: Hell, one of the most critical aspects is integration of the models into other products, and even on this end open-source is keeping up (and will eventually outpace when the VC money dries out) with these big companies.

In my eyes I would rather use the AI I can run on my own paid infrastructure, so if there's an outage its isolated, or I could potentially have a different region / DC to fallback on.

I'm still surprised that neither Microsoft nor Amazon have made their own models available on their cloud offerings. I guess Microsoft probably does have Phi on there, but it's not front and center, especially with something like Copilot for Devs (seriously Microsoft rebrand that damn thing to be clear what you mean by Copilot!) where they could use the cheaper compute by using something like Phi.

There's a significant amount of innovation happening, but if the market decides this AI thing is not worth funding then I think that'll dry up overnight.

1. https://thenextweb.com/news/anthropic-private-equity-venture...

Article content: “The US are capitalizing on AI the best”

A lot of assumptions there that no one can actually verify as true right now. If commercialization into rent-seeking SaaS landscapes is the endgame, then yeah, the US is winning the AI race. If individualization, local LLMs, and consumer hardware are the endgame, China is winning the AI race. If it’s something entirely different - if LLMs are the wall and research is what grants the next breakthrough, or if compute and memory requirements take a dive, or whatever; then we have no idea who’s winning the race because that stuff is mostly happening behind closed doors.

But the thing is... I could be using any of the llms for my use - I'm using a middleware that lets me change providers only with a configuration change.

So it's going to be tough for USA ai companies to charge 5x to 20x (depending on what you're doing).

It begs the question because both its premise and assertion are already wrong. Has AI improved the industrial capacity of the US in order to improve the lives of its citizens? No it hasn't. Has AI increased the wealth of its citizens by being able to do laundry or any household task in a generalized way? No it hasn't. The only thing it's really done is make very narrow slices of white-collar work more fungible. In what way has AI been able to address existing shortcomings of the US?

AI is not some divine creation. It is built by humans. History has repeatedly shown that China is able to catch up, and often surpass others in the end.

The article never really explains why AI is supposedly so unique that it guarantees the United States will inevitably win.

On a personal level, I simply do not trust the US anymore. I won't host any of my personal data in a US company. I don't want the US govt invading my personal privacy, and their corporations are constantly leaking and selling private data. I consider US to be rapidly approaching complete autocracy (on par with China) so US-hosted AI is a non-starter. And let's not forget local inference keeps getting more efficient, with higher context and TPS in the same amount of RAM. Within a year even small consumer machines will run local models good enough for basic coding, and in 3 years RAM prices will lower and everyone will be able to afford a decent rig.

Finally, open weight models are now good enough for daily work. They may never be as good as SOTA (SOTA will just keep increasing indefinitely), but that doesn't matter; my car may not be as fast as a Porsche but it still gets me to the grocery store and back. So I use non-US hosted model providers which provide open weights, which are both significantly cheaper than Anthropic/OpenAI, and actually allow me to use my subscriptions without a moat.

But yes, Anthropic/OpenAI are absolutely the new Oracle. They will win for US govt and Enterprise contracts. But that's far from the only users of AI.

The winner here will be whoever can move atoms with AI not take notes at the daily standup.

i.e. Think boston dynamics vs unitree

They're both doing well but I'd lean towards China is winning on atoms in light of a huge manufacturing base they can AI-ify.

I feel like the author (and perhaps many here on HN) are on a different planet than almost everyone I interact with.

>Frontier cyber models may push states and defense firms toward the opposite logic: security by obscurity, with closed software, closed tooling, closed firmware, and closed chips. If a model cannot train on the code and architecture of a target stack, it will usually have less context and less speed. That does not make systems safe, but it does raise the value of proprietary stacks all the way down to hardware.

Is this really true. Are there any experts who can weigh in on this.

Should we interpret this to mean that in the new world Windows is more resistant to attacks than say Linux.

Thats like Microsoft saying "Don't use Linux because selling an operating system is what matters"

Even if any of the US corporations would eventually end up in a scenario where their revenue is at least as high as their inference cost, what harm would that do to the other contenders? It's not as if there is any kind of network effect here that would exlude them from market participation.

Is it just that the subject line alone is a springboard for casual discussion? If so, maybe that's fine, but then, it feels like we'd be better off cultivating these discussions as "ask HN" posts instead of boosting this kind of web content.

The USA is very good at loosing very, very expensively....

Where are these profits of which you speak?

It's like the USA Librem 5 vs PinePhone. About the same HW for $1600 vs $150.

Sure will not pay 10x for "US" thing just because it's a US thing.

Correct. "Revenue" is the wrong scorecard when they're selling 20$ bills for 15$. I too can make a bajillion dollars in revenue with that strategy.

Show me a company not speed running the uber/doordash playbook and we can talk.

If only people would stop thinking in terms of who is winning, jeez.

Opening up comments to see top comments are 90% "NO U" without any substantial discussion - you disappoint me, HN.

Cultivating an ecosysyem of strong capital protections, wealth creation through extraction, and tax advantages for AI finance is what we should be looking for. Commercialzation may be a step towards that, but isn't the destination. We have to create a system where those with money can multiply it, not simple add to it.

Does any of the US companies earn money on LLMs? No, they bleed money. Github Copilot is switching to token based pricing, which will be costlier than hiring juniors.

Anthropic also is switching enterprises to token based pricing from their subscription one.

From the big three only Codex is still in somekind of subscription pricing, but they'll shift eventually (usage limits are a kind of that, but they have them less stricter than Claude ones)

There is one winner in this race - China. Trump with his agendas and wars makes it even more likely that China will lead this new market.

Sorry, nobody's winning that AI race.

I'm not certain that racing China in AI is the right reason but it might get us... somewhere.

Not only is the investment that keeps US AI companies flying high slowing, I suspect in two or three years, we'll all mostly be using open models and the people making money will be the hardware manufacturers. Even the small models will keep getting more capable. I'd guess a model you can run on a high end, but not outrageously overbuilt, developer desktop or laptop (something like 128GB of unified RAM), will be competitive with the current frontier when it's allowed to search the web and do research and write test code. You can't fit as much knowledge in a small model (80GB of weights can't store the world's knowledge), but I don't have the world's knowledge in my head, either, and yet I can figure out most problems with a little googling and experimentation. The reasoning and tool use abilities of smaller models is where the gap is closing, and that's what will make the huge models obsolete for huge classes of problem.

Already, there are many classes of problem that the easily self-hostable Qwen 3.6 27B can solve that required a frontier model a year ago. When the self-hosted options reach Opus 4.5-ish levels of capability, the argument for paying for tokens for most work begins to look a lot less compelling. And, looking forward, 1.58 bit models are coming. Incredible intelligence density, and still a lot of improvements happening.

Larry just fired 30% of his people at Oracle because, apparently, he is in an immediate need for cash. Because Oracle's early AI bets aren't paying off.

puke

Yeah, go ahead and run your country into the ground because of hypercapitalism and hypercommercialization, you're almost at the end game now! While the rest of us try to figure out how to actually build societies worthwhile to live in and experience, with healthcare and not waging war on our neighbors.

I don't know how people can seriously publish stuff like this and not feel like they're actively trying to make the world worse. Is money really the single thing y'all can focus on? Is there nothing better in life you can chase, even if it's also a number? So sad to see stuff like this.

Mass unemployment and an eventually collasping economy is winning?

The FSF was not an attack on commercialization, it was about giving users more freedom with their own copy.

AI commercialization is why we will always be a few steps ahead in AI.

The Chinese and Russians are free to join us. It's a pickup game.

That doesn't count as winning at all.

Chinese culture is quick to embrace the benefits.

It's like people forget the entire point, perhaps even definition of technology is "doing more with less."

The "brute force" of power and cycles is almost certainly the least important thing, perhaps even a hinderance.

We've invented a new term here too: revenue backlog. OpenAI and Anthropic in particular need to recover probably at least $2 trillion to recoup their capex investments. Now Claude code has had an impact on software engineering but for a lot of AI uses you're just not going to recover $2T on $20/month subscriptions. It reminds me of Twitter trying to dig itself out of a $44B hole and losing half their ad revenue with $8/month blue ticks.

The only commercial product AI sells is labor displacement and the resulting wage suppression. You lay off 10-20% of your staff and nobody is asking for raises. The people left are happyt o still have jobs (and thus a house). They'll work even harder doing unpaid labor of the displaced workers to keep those jobs. That's what OpenAI and Anthropic are selling.

The problem is that if these companies get their way, 10-20% of the population is going to be out-of-work and society is going to fall apart. Data centers are going to be the targets of increased societal desperation and anger as this gets worse.

There was a report this week that roughly 50% of singles in the US aren't dating because they can't afford to [1]. This goes well beyond the well-understood problems of not being able to afford a house let alone start a family. This is a birth rate death spiral in the making.

So, back to OpenAI and Anthropic, the only way they justify their valuations and can make up the "revenue backlog" is if they have a moat. And I don't think that's going to happen. Hardware will get cheaper. Nobody is talking about how the generation of AI hardware will write of trillions in investments for some reason. I don't know why.

But the dark horse here is China. DeepSeek when it was first released (early last year?) was a shot across the bow. We have it and toher models (eg Qwen) that will close the gap with whatever OpenAI and Anthropic produce such that no company will "own" AI in the way that OpenAI and Anthropic need to. In the coming years, China's chipmaking is rapidly closing the EUV gap and Western companies have zero penetration into this market. China doesn't want to be dependent on foreign tech that can be withheld at any moment.

Don't believe me? Just listen to the NVidia CEO say the exact same thing [2][3]. Huang realizes this is such a problem that he's gone on Air Force One to this week's Trump summit in China to try and convince the Chinese to buy NVidia chips.

[1]: https://parade.com/living/nearly-50-of-single-americans-not-...

[2]: https://www.youtube.com/watch?v=aJmHfmrRMUE

[3]: https://www.youtube.com/watch?v=Hrbq66XqtCo

It’s all about adoption and the bigger picture. The US is an untrustworthy, isolated island in the AI future if you vote another idiot into office in a few years. If you’ll still be able to vote at all, that is.

The largest part of the world is not the US. The cutting-edge US models are way too expensive for most parts of the world, and that also shows in adoption.

China is building an ecosystem of open-source models that are both cheap and good enough for most use cases. While most of the US AI sphere will collapse under the pressure of making profits, which means having their models and infrastructure adopted by as many enterprises and individuals across the world, China’s models will have become global standards and hard to displace.

If Beijing’s AI pitch centers on universal access and cost-effectiveness, then Chinese AI firms do not need the latest chips to win the global AI race. They also don’t need the expensive US-run infrastructure. If you watch Chinese AI adoption closely, they already want as many Chinese people as possible to be able to build and try with AI, whereas for most Americans, US models for productive use are already too expensive.

Kimi K2.6 sits within touching distance of Opus 4.7 and GPT-5.5 while costing about $4 per mil output tokens. That is six to eight times cheaper than cutting-edge US models. If you run hundreds of agents, that’s a significant opportunity to get the same work done for a lot less.

Even early adopters like Singapore, ditching US models, the government kicked Zuckerberg in the nuts and went for Qwen instead to build its sovereign AI models.

To understand why the US is at a severe disadvantage in this race, you need to understand China’s Belt and Road Initiative (BRI). BRI entails Chinese firms delivering fully financed infrastructure projects in a bid to lock third countries into China’s economic orbit. They use the same approach for their open source ai models, but this time the infrastructure is both invisible and free.

No need to build power plants or buy /build ports. AI dependency is invisible to both policymakers and the population, limiting pushback. No pesky activists in Germany nagging about China buying parts of ports. No African nutbags questioning why the humble Xi is building hospitals in areas Chinese mining companies take things out of the ground for pennies on the dollar.

China is going for a marathon here while the US tries to push their ai tech by sheer force into the throats of the world. As soon as Chinese ai models have become global standards, it’s game over for us ai companies. And China is way better at this game than the US. They have proven this over and over again in the past 50 years.

I recommend reading the China Standards 2035 strategy to get a better understanding of their approach and how smart this is.

https://www.china-briefing.com/news/china-standards-2035-str...

AI is not as revolutionary as you think in terms of our experiences with previous technological advances in terms of trade and economics.

Western economies are locked into U.S. models, while China runs on Chinese ones. It’s the age-old game. But the real war of the AI race will be fought in the global south.

I will give you three examples.

Can you really imagine, if you look at what AI needs to cost to make a profit, that even at the current prices, US models and infrastructure, which are heavily subsidised already, being used in cost-sensitive countries? I am not talking about coders, think really big here for a second.

Secondly, US ai models are trained on Western data. How do you expect them to grasp local contexts in the Southern Hemisphere? Chinese open-source models, on the other hand, can be downloaded and finetuned with country-specific data.

Want an example? Check out AfriqueQwen-14B, which is adapted to the top twenty African languages.

So I think this author is wrong. The ai race to be won is not hardware or cloud infrastructure, my money is on it will be a contest to decide which models and standards become the default infrastructure in countries that are up for grabs.

China neither needs the best models nor does it need the best cloud infrastructure, it just, like so often, only needs to be affordable and good enough to become the default choice in emerging markets.

The right choice would be for everyone to step off the gas pedal and think about whether we are willing to become China in order to beat China. Our ancestors worked really hard to get us here, our rights, our ways of life, culture, all the blood, sweat, and tears.

AI better be worth it in the long run for all of humanity if we go back to survival of the fittest. Because that is what it will take to beat China at their game.

Not even gonna bother clicking through this one, the title is that egregious. And by the way, you can be damn sure that if Anthropic or whichever other American frontier model model is the best of its day was on the cusp of going under, the US gov would either pump it full of government contracts or (less likely) nationalise it.

And that's not to mention the warping of US economic life by the concentration of capital around this bizarre endeavor, with the circular multi-hundred-billion-dollar deals and such.

Unfortunately, the detrimental effects of global warming arrive gradually, and are spread out over the entire globe, so the "AI barons"/tech magnates will probably suffer the least, while island countries will be completely wiped out, whole regions will become too hot to sustainabily live in, tens of not hundreds of millions will have to migrate, biological diversity will suffer, etc. They will look back on these times in a 100 years and will think of us, or at least of US, as the people boarding the Titanic. Hopefully not as the people who board the Hindenburg.

As with another recent example, sometimes in war there is no winning: just loss. This is obviously for us programmers an incredible and wild age, filled with nothing short of miracles. It's incredible. But the prices we are paying, the extreme tensions we are creating, the stress and strain of this all has been incredibly unpleasant, and very very very few people feel like they are seeing upsides to this worrisome menacing age, that promises very few people on the planet anything better coming, and which. Has already made life considerably worse, which no nation has yet directed towards helping its people.

Which one of them all?

If you mean "building models that are very good at coding and as substitutes for search engines", then yeah, sure.

But if you mean: "applying AI to industrial applications and robotics", then China is far ahead: https://time.com/7382151/china-dominates-the-physical-ai-rac...

It remains what benefit, if any, Americans will see from all this...

Just because you are first to do x, doesn't mean you are going to be the winner.

The cost of winning this race has been telling our citizen s we will replace them with robots and there is no hope for their children’s future employment.

The cost has been destroying trust as we tell citizens water and power should go to server farms and not them.

The cost has been naked power telling democracy it’s wrong and dying

I think when we discover the limits of LLM tech and tally its benefits over its cost — we may regret this win.

But don’t let me contradict a bunch of fake techno oligarchs wrapping themselves in war like patriotism to get the investments they need to keep this going.