By Mike Magee
Last week, after a visit to the White House, Jensen Huang instigated a wholesale reversal of Trump’s policy that was blocking Nvidia sales of his H20 chip to China. What did Jensen say?
We can only guess, of course. But probably shared the results of A patented report Of researchers of AI noticed in Digital science That suggested that an immediate correction of the policy course was critical. Beyond the fact that approximately 50% of all A -researchers are currently in China, their study documented that “in 2000, China -based academics produced only 671 AI documents, but in 2024 their 23,695 publications related to AI (6388), Thedomic, Thedputput or Thedput of TheDPut of Thedpput of Thedput of Thedput of Thedput of Thedput of Thedput of Thedput of Thedput of Theating) (2747) (2747) and the European Union).
David Hook, CEO o Digital science It was declarative in the opening of the report, stating that “the influence of the United States on AI investigation is decreasing, since China is now dominating.”
China now supports 30,000 IA researchers Compared to only 10,000 in the United States. And that number is being reduced thanks to the US rate. And to visa mischief, and administration surveys in our main academic institutions.
Economics teachers David Author (Mit) and Gordon Hanson (Harvard), known for “his research on how globalization, and especially China’s emergence, remodeled the US labor market,” described the elements of “China Shock 1.0”. In 2013. It was “a singular process of the transition from the late 1970s of the 1970s of the Central Maoist Planning to a market economy, which quickly transferred the work and capital of the country of collective rural farms to capitalist urban factories.”
As a result, A quarter of all the manufacturing work of the United States Disappeared between 1999 and 2007. China’s manufacturing workforce It covers 100 million, eclipsing the manufacturing work count of the USA of 13 million. These numbers reached their maximum point a decade ago when the supply of low -cost labor of China reached its maximum point. But these days, China is clearly looking forward while this administration and its advisors stay in the rearview mirror.
Welcome to “China Shock 2.0” wrote Auto and Hanson in a recently New York Times editorial. But this time, their leaders are focusing on “key technologies of the 21st century … (Y) will last while China has resources, patience and discipline to compete fiercely.”
The very respected Australian Strategic Policy InstituteFinanced by its Defense Department, it has been tracking the volume of innovative technological research published in the United States and China for more than a quarter of a century. They see this as a measure of the opinion of experts in which the greatest innovations originate. In 2007, we lead China in the previous four years in 60 or 64 “Border technologies”.
Two decades later, The table has turned, with China very much ahead of the USA in 57 or 64 measures categories.
In the AI algorithm investigation, China had 29% of the documents for our 12%. In high performance computing, they had 36% to our 13%. And in automatic learning, they tilted the 40% scale compared to our 11%.
According to Author and HansonChina is the “Predator Apex” in a battle that combines “economic Darwinism with Chinese industrial policy.” As Musk was dismantling our National Planatus, the Chinese government has been pointing to our most innovative sectors, including “Aviation, AI, Telecommunications, Microprocessors, Robotics, Nuclear and Fusion Power, and Solum.
Warnings may sound strident, but this is a bad reading of historical proportions. And everything is reduced to a word. Duty. And when it comes to Trump’s rates, no country has treated with a bone with greater hostility than China, which attracts this reprimand of intentional isolationism of the United States in January 2025 by the Harvard Business Review: “Also with the privacy and data security of the data: American LLMS will ignore the threat of interruption of Chinese LLM at its own risk.”
In the newspaper, Harvard Professor Business School, Raj Choudhury partnered with the Professor of Business Management of the University of Syracuse, Natarajan Balasubramanian and the economist at the University of Tsinghua, Mingtao Xu, to answer the timely question “Why Deepseek shouldn’t have been a surprise.”
For those who are not familiar with the controversy, the generative superstics of the United States had an excessive leg in recent times, with huge palmadites on the back and philosophy (between the rocket trips and the reflections of Mars), when a little Chinese, a little better known, Deepseek, Mold the Apple cart with its open source release from a smaller, more efficient and much cheaper llm. “
The moment could not have been more uncomfortable for our technocratic elites: Claude de Anthrope, Google Gemini, Open Ai’s Chatgpt, Microsoft Copilot, Musk’s X.AI and Metai.
These owners and self -processed geniuses are completely committed to eliminating each other. But that does not prevent them from adopting the environment of Sam Altman from Open AI, who recently wrote an opinion article entitled “The intelligence age” In what he explained, “technology took us from the stone age to the agricultural era and then to the industrial era. From here, the path to (Ai enabled) Intelligence age It is paved with computation, energy and human will. “
And these American billionaires have put their money where their mouth is, burning huge amounts of capital as they build a completely new and completely founded industry: the creation of generative factories of AI that replaces both the investment of cont in erythlant Nuclear energy industry Now it has passed through the clouds, and Musk has officially registered a new city for them and their followers of AI, Starbase, Tx.
Meanwhile, Rand and Kyle Chan researchers, Inform That “China is applying state support throughout its AI technological battery, from chips and data centers to energy.” The Chinese government strategy focuses on three coordinate construction blocks: 1) Computer capacity, 2) integrated and accessible Chinese data, and 3) qualified engineers.
At the same time, the government has financed private laboratories networks committed to public/private association, and encouraging local government leaders in cities through the country to support hundreds of new AI research companies. A city in southern China, Hangzhouthat already married giant AlibabaCreated a section of the city, called Dream townnow the birthplace and permanent residence of Deepseek.
Kyle Chan de Rand recently marked the fact that China has converted adversity (Trump tariffs) into a motivating tool. By maintaining their discoveries of open source LLM (since US companies are Leandard’s own controls to improve future profitability), authoritarian China, which is supposed to oppose transparency and accessibility, skirt, outline, outline, The AtlanticTo move “exactly in the opposite direction of where the American technology industry is directed.”
Why deep see now? Management experts Let’s say this is only “classical theory of interruption in the game.” The case in question was what happened to the steel industry decades ago when integrated steel plants served to the high -end steel customer were exceeding less efficient electric arc furnaces. In response, older plants, with lower, personalized and specialized technology to address low -end tasks, such as creating steel reference bars to reinforce concrete construction. The holder was forced to “give up market share to the disruptor.”
In this case, Depseek can be lower technology than the next best version of Chatgpt, but it seems to be good enough to offer “Specific industry applications Driven by their LLM that are deeply integrated into China’s digital ecosystems. “The Chinese LLMs are more specific and less general puro.
How low? For software developers around the world, Depseek-R1 is 95% cheaper than using the O1 Open AI product and as an effective. Satya Nadella, The Microsoft CEO says that the child’s child will make the AI “shoot, turning it into a merchandise that we simply cannot get enough or.”
In a nutshell? Under the nose of the United States, Porcelain It has become “democratic” to the opening of its code, cheated global software developers and users, and placing the main United States competitor (at least for the moment) in the driver’s seat.
Mike Magee MD is a regular medical and correspondent historian in THCB. Hey is the author of Blue Code, within the United States Medical Industrial Complex. (Grove/2020).
