Weird how the article doesn’t mention a key factor: the Chinese labs have really low headcount compared to their Western competitors, and organizationally they’re much more focused on model building than sidequests. Deepseek had only a couple hundred employees until recently, and accordong to leaks only had one or two people maintaining their consumer facing app.
The meat of the article is behind the paywall.
My theory is that
-
China has better integration between academia and industry - I’m mostly basing this off how in the US the academia is chronically underinvested so that the private sector can get it’s plunder of smart graduates, but when it comes to AI development you actually need academic knowledge of intelligence otherwise you’re just approaching it like an engineering problem
-
US companies are incentivized to spend more, the more broke they are the smarter they seem, so there isn’t really any incentive to improve in smart ways because you can just rack up a bigger bill with your investors and keep the circular economy going.
Would love to know what the recipe says though.
My theory is the article is not accurate. China needs articles saying how cool China is.
Yeah, the US uses no propaganda whatsoever. :)
No one said so tbf.
-
Being the pioneer means doing the heavy lifting and making mistakes. So makes sense, China can learn for others mistakes
That’d only make sense if this was a total investment thing, but it’s not. America is investing at a higher rate yet China is getting more returns. Who started first is irrelevant here.
Im not sure how that cancels out what they said. One of their first models said “I’m ChatGPT” when asked who are you and it was rumored they used OpenAI model outputs for training.
I think of it like drafting behind another vehicle on the speedway. The car behind gets higher fuel efficiency
But the problem there is that China is clearly ahead of America now when it comes to AI, so if that was it they wouldn’t be getting higher fuel efficiency anymore.
China is clearly ahead of America
Well that’s debatable. The article implies that the US is still ahead. The root comment of this thread implies that the US is still ahead. Your earlier comment made no mention of this idea. Clearly if that was part of your argument you should have brought it up earlier
The article implies that the US is still ahead.
The whole point of the article is that it’s not (or at least not enough for the disparity we’re seeing).
The root comment of this thread implies that the US is still ahead.
No, it implies it was ahead at one point (which is obviously true)
Its investment lags far behind America’s. Its models do not
Their models appear only fractionally less powerful owing to this frugality. K3, an advanced model launched last month by Moonshot AI, a Beijing-based startup, is 95% as clever (on widely used benchmarks) as Fable 5, a frontier model from Anthropic, another top American lab. It is also 70% cheaper to use. On August 3rd Alibaba, a Chinese tech giant, released a model which reportedly scored among the world’s best by some measures.
See also: Chinese models showing up at or near the top in leaderboards. Admittedly it seems I was wrong about China being clearly ahead, but what little gap exists is far from enough to explain the disparity in investment efficacy.
little gap exists is far from enough to explain the disparity in investment efficacy.
This is difficult to say. From what I’ve seen AI gains are slowing down, so catching up might not be that hard anymore. We’ll see from here if China is able to pull a clear lead. If so, then that’s clear evidence that they aren’t just copying
China loves to steal tech.
Oh come on all of AI is built on stolen tech, you can’t get mad at China for playing by the same rules as everyone else.
I’m mad at China and at OpenAI, Anthropic, Facebook, Google, …
Less mad, more statement of fact
I’m not mad?
Which is great. Intellectual property is innovation-stifling cancer. I’d certainly rather China have the tech to make all those EVs than leave the West to hoard it.
I don’t necessarily disagree, but you’re a trashy person stop talking to me mrdown
I’ll never understand why people keep saying I’m Mrdown.
Edit: Oh it’s not “people” it’s just you. Still, why Mrdown in particular? Like what’s the thought process here?
You know why
No I don’t?
The problem with not doing the heavy lifting yourself and using derived technology and ideas is that in evolutionary terms you are not selecting for innovators. Down the road this approach becomes ingrained both philosophically and in reality as you let your own innovators wither on the vine.
Yeah China’s VERY good at reverse engineering stuff and playing catch up. Haven’t really seen them take a clear lead in developing new technology in any field though.
That’s more due to the lack of new technology these days. AI aside most modern engineering is incremental improvements to existing technology—that’s the innovation everyone is doing.
GenAI? Self-driving cars? Robots? Materials science?
All science and innovation is incremental, but some have bigger impacts than others.
In order:
GenAI?
I said “AI aside” for a reason.
Self-driving cars?
Fair enough, though I don’t hear much about those these days.
Robots? Materials science?
No idea what’s going on with materials science, but robots aren’t exactly new, so this is incremental improvement to existing technology.
i distilled this article
Summary of the article “How China gets better bang for its buck than America in AI” (Aug 3 2026)
-
U.S. AI spending is massive – Bloomberg Intelligence estimates U.S. data‑centre capital outlays could exceed $740 billion in 2026, with Nvidia alone negotiating a $250 billion financing deal for a $500 billion data‑centre run by OpenAI. Alphabet announced a $205 billion AI budget.
-
China spends far less – Chinese tech firms are projected to invest less than one‑tenth of the U.S. amount in data centres. Yet their models perform only slightly behind U.S. equivalents. For example:
- K3 (Moonshot AI) scores ≈ 95 % of Anthropic’s Fable 5 on common benchmarks while being 70 % cheaper to run.
- Alibaba’s newly released model ranks among the world’s best on certain metrics.
-
Why Chinese spending is efficient
- Lower input costs – Land, construction, equipment and labour are cheaper in China.
- Model distillation – Chinese labs often train models using outputs from expensive U.S. models, reducing the compute needed.
- Hidden spending – Some expenditures on high‑end chips are masked as “cost‑saving” techniques that make inferior hardware achieve higher performance (e.g., DeepSeek’s efficiency tricks).
-
Export restrictions limit Chinese capital use – U.S. bans on advanced AI chips (Nvidia designs, TSMC manufacturing) prevent China from buying the most powerful hardware.
- Chinese firms are pushed toward domestic alternatives (Huawei, SMIC).
- Sanctions also block access to cutting‑edge chip‑making equipment, forcing costly work‑arounds and capping production capacity.
-
Domestic demand constraints – Chinese enterprises spend < 10 % of what U.S. firms spend on IT, despite China’s GDP being two‑thirds of the U.S. (or a third larger in PPP terms). This throttles revenue prospects for AI providers, curbing their willingness to invest heavily.
-
Strategic focus differs – The Chinese Communist Party emphasizes diffusing AI across the economy, not pursuing a race toward artificial general intelligence (AGI). Fewer than ten Chinese firms target AGI, compared with dozens of U.S. players.
-
Investor attitudes – Chinese investors have historically punished over‑spending on AI, whereas U.S. investors once rewarded aggressive budgeting. This cultural difference keeps Chinese AI budgets modest.
-
Potential bottlenecks for China – Despite restraint, China may face compute shortages:
- ByteDance experiences ten‑hour processing times for some videos.
- Alibaba Cloud, Zhipu AI, and Moonshot’s K3 have long waiting lists or quickly sell out capacity.
- Over‑restriction could stifle growth if AI services cannot meet user demand.
Overall takeaway: China achieves comparable AI performance to the U.S. while spending a fraction of the capital by leveraging cheaper resources, model‑distillation techniques, and a strategic focus on wide‑scale diffusion rather than raw computational power. However, export bans, limited domestic chip capacity, modest corporate demand, and cautious investors together create both an efficiency advantage and a risk of under‑provisioned infrastructure.
-
