When the 'tech giant' starts crying: NVIDIA's 'boy who cried wolf' farce

The AI race in 2025 is playing out like an absurd drama: Jensen Huang, the head of Nvidia, the global computing power giant, is on one hand proclaiming that 'China will defeat the U.S. in the AI race,' while on the other hand pleading with the U.S. government, 'Stop building walls, the market is collapsing.' Behind this farce is a fierce clash between American technological hegemony and market forces, and above all, a dark humor story of 'tying your own hands while blaming your opponent for being too strong.'

Huang's warning can be described as a 'technological hegemony version of the boy who cried wolf.' He first lamented that the new AI regulations in various U.S. states would create '50 new standards,' fragmenting technical norms; he also complained that China’s 'electricity is almost free,' allowing domestic chips to operate at lower costs. Most strikingly, he even attributed the decline of the U.S. technology system to 'cynicism'—as if the U.S. government were not a policymaker but a victim hijacked by pessimism.

The absurdity of this performance lies in the fact that NVIDIA itself is a beneficiary of technological restrictions. Over the past five years, it has made huge profits by selling specially tailored chips to China, and in 2023, sales of H20 chips in China grew by 42%. Now, with China’s chip localization rate soaring to 42% and the replacement rate of training chips rising from 5% to 15%, Jensen Huang suddenly laments that 'we are losing the market.' This is not a warning; it is plainly the hysteria of a technology monopolist facing market backlash.

The actions of the U.S. government are even more like an absurd play. The Trump administration simultaneously banned NVIDIA from selling its most advanced chips to China while demanding 'allow cooperation but ban cutting-edge technology'—this kind of 'wants the horse to run but not to eat' logic exposes the hypocrisy of technological hegemony. Even more ironically, American companies have long voted with their feet: Oracle warned that the blockade policy would cause U.S. companies to lose 80% of the global market share; NVIDIA’s top two clients contribute 39% of revenue, a risk hanging like a sword; meanwhile, China’s Huawei, in collaboration with ten domestic EDA companies, launched a full-process design toolchain that has already saved 2 billion yuan in chip R&D costs and shortened the development cycle by 30%.

The ending of this farce was already destined: as the U.S. attempts to use political means to cut off the world’s largest AI application market, China is rewriting the rules with a 'replace time with scenarios' strategy. Huawei Ascend 910B chips surpass NVIDIA H100 by 30% in energy efficiency in smart city video analysis scenarios; the DeepSeek-V3 model improves government approval efficiency by 400%; China’s automation penetration in short video creation reaches 70% (only 45% in the U.S.). What the U.S. technology system is losing is not just market share, but the definition of next-generation AI chips.

Huang’s warning exposes a fatal rift in the U.S. AI strategy: when 50% of AI researchers are Chinese, and the scale of China’s developer community is the largest in the world, attempts to maintain technological leadership by splitting the market are essentially defying the laws of innovation. History has long proven that from semiconductors to the internet, the leaders of every technological revolution are those countries that embrace the global market and integrate global talent. If the U.S. wants to continue leading, what it needs is not walls, but a more open platform—to attract global talent, embrace global markets, and iterate global standards.

The lesson this absurd play offers the world is that in AI, which is reshaping the global landscape, the ultimate battlefield of technological competition has never been the laboratory, but the global market; the ultimate weapon for winning competition has never been blockade, but openness. While the U.S. is still struggling with 'whether to allow China to use the most advanced chips,' China has already been spreading AI technology to the Middle East and Southeast Asia through the 'Digital Silk Road.' Perhaps as Jensen Huang helplessly admits: 'China is only a few nanoseconds behind the U.S. in AI'—but this nanosecond gap, due to U.S. closure, has become an accelerator for China’s technological self-reliance.

In this civilization-level marathon, no country can lead by building walls. If the U.S. wants to avoid the fate of being 'defeated,' perhaps it must first learn to set aside the arrogance of technological hegemony and heed the ironic laughter of market laws.