Chinese AI labs kinda seem like a plot twist in LLM evolution. Their models are quite capable now. They’re not at the levels of American labs’ flagship models yet, but the gap has been narrowing quite a lot.
When OpenAI and Anthropic models are only marginally better, but much pricier, then I would think they’ll gradually shed users (followed by investors).
Ironically, I could imagine a possibility of Nvidia “saving” American AI. If they can take the lead with Nemotron (in like a “post-OpenAI/Anthropic” future when open-source models dominate), then maybe they can survive on chip sales… Though they’d probably have to compete with Chinese chips at that point.
Chinese Big Tech isn’t really known for innovation, they take existing tech and push to make it more efficient by just throwing people at the problem. It’s basically because they have a culture where critical thinking is not welcome. Makes it difficult to think outside the box. It’s why they still haven’t gotten an EUV machine out of the prototyping phase
Their only real innovative industry is their battery industry.
For the first time among those watching these issues closely, the technological “choke point” strategy adopted by U.S. authorities across the late 2010s and early 2020s has now been shown to have failed, as Chinese government and R&D officials, as well as key state-backed and private sector firms, have been able to respond to the challenge forcefully and effectively. Leading the response are key policymakers: Vice Premier Ding Xuexiang, and the semiconductor team at the National Development and Reform Commission (NDRC), overseen by Vice Minister Xiangli Bin. A new AI-focused group at the NDRC overseen by Vice Minister Huang Ru is also increasingly important, as semiconductor and AI-related industry policies increasingly dovetail.
…
Leading domestic foundry SMIC, for example, has faced pressure to manufacture Huawei’s most advanced chip designs by stretching existing foreign equipment to its limits. This includes deep ultraviolet (DUV) lithography systems supplied by the Dutch firm ASML, which are being pushed beyond their intended capabilities, often resulting in low and inconsistent yields. The urgency stems from Huawei’s need for system-on-chip (SoC) processors for its consumer devices—especially smartphones, as well as for advanced AI chips in its Ascend 9XX series.
Remind me, again. Who else was experimenting with deep ultraviolet (DUV) lithography at scale prior to 2024? Who else was a front-runner in developing and deploying system-on-a-chip or AI embedding?
That’s before you get into the EV sector, SMRs for bulk shipping, or the Chinese airplane and aerospace development.
India, Korea, and Japan have all been in a scramble to keep up with the Chinese industrial programs. Meanwhile the US/EU don’t even seem to bother trying.
This is moronic and entirely divorced from the facts. Look at key players in the Chinese LLM space like Deepseek: it’s a tiny team of less than 200 people, building models that rival US tech firms with thousands. They make breakthroughs by pushing research first and intensive planning, rather than brute force. These are immensely innovative and creative teams with a great approach to R&D and engineering above all else
This wouldn’t really negate what I’m talking about in terms of their organizational advantages or the argument I’m making about them not just “throwing people” at the problem. But also, I don’t see any evidence that this is true; it seems their hiring strategy is to grab researchers that recently graduated from top Chinese universities as their talent
As far as I understand it is decently true, but not to the extent that they would be incapable of doing what they’re achieving. Either way you’re right, it doesn’t refute your claims in any way - those researchers are still doing work in China for Chinese companies, regardless of where they got their education.
Chinese AI labs kinda seem like a plot twist in LLM evolution. Their models are quite capable now. They’re not at the levels of American labs’ flagship models yet, but the gap has been narrowing quite a lot.
When OpenAI and Anthropic models are only marginally better, but much pricier, then I would think they’ll gradually shed users (followed by investors).
Ironically, I could imagine a possibility of Nvidia “saving” American AI. If they can take the lead with Nemotron (in like a “post-OpenAI/Anthropic” future when open-source models dominate), then maybe they can survive on chip sales… Though they’d probably have to compete with Chinese chips at that point.
Chinese Big Tech isn’t really known for innovation, they take existing tech and push to make it more efficient by just throwing people at the problem. It’s basically because they have a culture where critical thinking is not welcome. Makes it difficult to think outside the box. It’s why they still haven’t gotten an EUV machine out of the prototyping phase
Their only real innovative industry is their battery industry.
This is a racist fairy tale that will get you pwned.
This is the same racist trope we told ourselves about the Japanese and then the Koreans. Obviously only Americans have the Innovation Gene
Innovation under Pressure: China’s Semiconductor Industry at a Crossroads
…
Remind me, again. Who else was experimenting with deep ultraviolet (DUV) lithography at scale prior to 2024? Who else was a front-runner in developing and deploying system-on-a-chip or AI embedding?
That’s before you get into the EV sector, SMRs for bulk shipping, or the Chinese airplane and aerospace development.
India, Korea, and Japan have all been in a scramble to keep up with the Chinese industrial programs. Meanwhile the US/EU don’t even seem to bother trying.
This is moronic and entirely divorced from the facts. Look at key players in the Chinese LLM space like Deepseek: it’s a tiny team of less than 200 people, building models that rival US tech firms with thousands. They make breakthroughs by pushing research first and intensive planning, rather than brute force. These are immensely innovative and creative teams with a great approach to R&D and engineering above all else
Well…those engineers were all training in the USA at MIT, Stanford, etc. but got the boot in 2025.
This wouldn’t really negate what I’m talking about in terms of their organizational advantages or the argument I’m making about them not just “throwing people” at the problem. But also, I don’t see any evidence that this is true; it seems their hiring strategy is to grab researchers that recently graduated from top Chinese universities as their talent
As far as I understand it is decently true, but not to the extent that they would be incapable of doing what they’re achieving. Either way you’re right, it doesn’t refute your claims in any way - those researchers are still doing work in China for Chinese companies, regardless of where they got their education.