Just a few days ago the Nvidia CEO Jensen Huanghas proclaimed in London that China is about to win the global race for artificial intelligence. According to him, the Chinese are “nanoseconds” away from the United States. The phrase made headlines, but it is worth asking why the most powerful manager in the sector now insists on painting a future where China appears as the inevitable champion.
Huang justified his prediction with the supposed Chinese energy muscle and its abundance of researchers, But he also offered a surprising argument: that Washington’s restrictions on the export of advanced chips could end up benefiting Beijing. An interpretation, at the very least, interested. Nothing scares a US regulator more than the idea that its sanctions strengthen a strategic rival. And nothing would be better for Nvidia than to sell without limits again in a market of 1.4 billion people.
It is true that China dominates research areas such as computer vision. At the ICCV held in Hawaii, half of the entries came from Chinese institutions. Agreed: a clear statistical triumph. But confusing volume with real impact is naive. Chinese academia produces a lot, yes, but a good part of its ecosystem continues to depend on models, software and conceptual frameworks developed outside its borders. Quantitative takeoff does not guarantee qualitative leadership.
China also boasts scale: more users, more data, more applications. But that advantage has existed for a decade. The difference now is another: a centralized state plan that combines public money, political control and an industrial discipline unthinkable in the West. In 2017, Beijing decided that it wanted to lead AI in 2030, and since then it has been pulling all the resources of the State to achieve it. The new National Venture Capital Guidance Fund — $138 billion for AI, semiconductors and quantum — is just another reminder that the boundary between business and government in China is virtually non-existent.
Meanwhile, The United States continues to dominate what really matters: cutting-edge models, private investment and, above all, chips. Nvidia reigns precisely because there is no comparable alternative. China runs, but runs behind. And despite its rhetoric of self-sufficiency, it continues to be trapped by the technological restrictions imposed from Washington.
Its computing centers are growing, its technology giants are developing domestic models, and its researchers are learning to exploit less advanced hardware. All this is true. But it is also that China has computing power that is not even a fraction of that of the United States. And that, without access to the most advanced chips, each technological leap will have an enormous cost and will require more “defensive” engineering than pure ambition.
Chinese talent is increasing – they produce more PhDs in science than anyone else – but innovation under surveillance and with strict limits on academic freedom has a ceiling that is difficult to overcome. The accumulation of data and brains does not always translate into disruptive creativity.
Those who appeal to a “shared race for the good of humanity” are guilty of candor. Competition is fierce, driven by national interests, military rivalry and gigantic corporate profits.
As of November 2025, The United States not only has the ten most valuable AI companies in the world, but 37 of the top 50. Nvidia just became the first company in history to be valued at $5 trillion. China only places four companies on that list. And even if unlisted giants like DeepSeek are excluded, the proportion is still devastating.
The reality is that The United States continues to have the largest computing capacity on the planet, almost half of the global totalwhile China—despite its industrial propaganda—appears lagging even behind India. The Asian country controls many data centers, yes, but with second-rate hardware.
So when Huang dramatizes the situation, it is worth reading between the lines. His implicit message is obvious: he wants Washington to eliminate export controls. That would reopen an immense and thirsty market for computing power for Nvidia. What he describes as a “danger” to the United States is, in reality, a hindrance to his bottom line.
But the era of technological free trade is already over. The United States and China openly compete for control of the critical infrastructure of the 21st century. Pretending that the solution is to let the most advanced chips flow without restrictions is, at the very least, naive. Or, more likely, a pressure maneuver wrapped in an apocalyptic speech.
Huang dirige the most valuable company in history. Maybe I should take it with a little more serenity. The scaremongering, in this case, seems less like lucid analysis than a carefully disguised business strategy.