Key Takeaways

  • Chinese AI companies are achieving commercial dominance in specific areas, like video generation, with Cling raising $2.8 billion at an $18 billion valuation and doing $500 million in Q1 AR, while OpenAI’s Sora shut down due to cost. This shows real market traction, not just academic progress.
  • The 'Great Firewall' and US government restrictions on access to leading American models (like OpenAI and Claude) prevent Chinese users from accessing these tools, forcing local developers to build their own. This creates a protected, competitive environment for homegrown AI.
  • US chip export bans, intended to hobble China's AI progress, are instead accelerating its path to self-sufficiency in hardware. This move pushes Chinese companies to develop their own chips, which has the unintended consequence of creating a commercially competitive, independent AI industry.
  • A surprising six of the top open-source models on OpenRouter today are Chinese, contrasting sharply with US frontier models that dominate the closed-source space. This signals a bifurcated global AI market where open-source innovation is thriving in China.

The Unseen AI Giant

Forget the narratives about the US leading the AI race. A different story is unfolding, one where China isn't just catching up, but pulling ahead in critical commercial sectors. Take Cling, for example. Harry Stebbings pointed out that it “raises 2.8 billion at an $18 billion valuation. Is the biggest AI video business in the world. It's doing 500 million in Q1 AR wise.” This isn't just a big number; it’s a direct contrast to OpenAI’s Sora, which struggled with cost-effectiveness and shut down. While American companies battle for closed-source frontier models, China is quietly building a dominant commercial presence, especially in open-source AI. Stebbings asked a pointed question: “The most commercially successful AI video product on earth is Chinese. the top six models as of today on open router or Chinese. Do you think China's running away with the model there?” This isn't theoretical; it's a real-world market shift.

The Firewall's Unintended Genius

China's AI ascendancy isn't happening in a vacuum. It's directly tied to government policy, both domestic and foreign. The 'Great Firewall' plays a bigger role than many realize. Jason explained the stark reality: “When you're in China, OpenAI and claude will not and anthropic will not serve you. It's not just a question of being blocked. You cannot access it.” This forced isolation acts as a powerful incubator. With no access to leading US models, Chinese developers must build their own. It's a classic case of necessity breeding invention, but on a national scale. This isn't about choice; it's about survival in a technologically isolated market. The result? A rapidly maturing, self-reliant AI ecosystem that operates independently of Western influence.

Chip Wars and Self-Sufficiency

US restrictions on chip exports were meant to slow China down, but the long-term impact appears to be the opposite. Jason referenced Jensen Huang's (NVIDIA CEO) perspective: “Jensen's point was you better let the GPUs go over there, right? Or they're going to just do it themselves. And you know, literally the fact that even in Hong Kong, which is much more open than China, I just couldn't use, you know, chat GP or Claude or the or the APIs, what do you expect?” This forced self-sufficiency has a clear commercial consequence: China will become competitive. Rory articulated the geopolitical dilemma: “If you believe there's a national security concern on these models and these chips then and it's legitimate and real and you've made that decision soberly as a government and responsibly, then you can choose to block access to these technologies. But you're right, you can't expect the other side to say, 'Okay, you caught us. We give up.'” The US gamble on limiting access may accelerate, not prevent, China's rise as an independent AI powerhouse, creating a global market split where local champions dominate.

What to Do With This

For ambitious founders, this isn't just an abstract geopolitical story; it’s a critical market signal. Your next move should be to map your competitive landscape for AI models and infrastructure, globally. Don't assume all innovation flows from Silicon Valley. If you're building an AI product, dedicate a day this week to research the leading Chinese open-source models on platforms like OpenRouter and understand how they compare to the US closed-source frontier models you might be tracking. Consider the implications for market entry if your product relies on, or competes with, AI in a bifurcated world. Ignoring China's self-sufficient and commercially successful AI sector means operating with half the data, potentially missing critical competitive threats or future collaboration opportunities.