Commentary
The New York Times calls it a “
conundrum.” America wants to beat China in artificial intelligence (AI), in part by attracting the tech talent of Chinese nationals to U.S. shores. But some of them come, learn AI, and bring it back to China. They are now so prominent in the field that they are almost leading it not only in China but also in the United States.
While U.S. AI continues to have advantages over China, for example, in the power of U.S. computer chips, we could become perilously dependent on the massive pool of AI talent in China that is increasingly in short supply relative to the needs of cutting-edge military and intelligence applications here at home.
University of California–Berkeley Professor
Pieter Abbeel, who started an AI company, told The New York Times that for leading businesses and academics, working with large numbers of Chinese researchers was “just a natural state of affairs.” Some work in places such as Google, Stanford, and MIT, which, when hiring Chinese citizens, thereby risk transferring cutting-edge AI tech to adversaries. In one case of a Chinese national engineer at Google, the Department of Justice alleges that he
illegally transferred sensitive AI tech over the course of a year to a company in Beijing that paid him secretly.
Over the past several years, Beijing has used a combination of state support for domestic companies and exploitation of U.S. academic and business openness to push China ahead on some AI metrics.
According to new
data released by the MacroPolo think tank on top AI researchers globally, 47 percent have an undergraduate degree from a university in China, while only 18 percent have an undergraduate degree from a U.S. university. Many of those 18 percent are likely from China. China’s percentage has increased rapidly since 2019, when it graduated just 29 percent of top AI researchers. Over that same period, the U.S. fell further behind. In 2019, we graduated 20 percent of the top AI researchers.
The accelerating spread is even greater for the elite (top 2 percent) of AI researchers. While China graduated a few of these in 2019, the country jumped to 12 percent in 2022. On that metric, too, the United States fell from 65 percent to 57 percent over the same period.
China has rapidly advanced in AI partly because of the more than
2,000 AI programs added to undergraduate universities in China since 2018. Many of these programs are geared not toward crowd pleasers such as ChatGPT or the Dall-E art generator but toward industry, manufacturing, and, most likely, the Chinese Communist Party’s (CCP’s) intelligence and military apparatus.
In 2023, China increased its AI industry by
18 percent to 4,400 core enterprises valued at more than $80 billion. In early March, the CCP published a paper revealing
plans to further support its AI, quantum computing, and big data analytics industry.
According to an early March report in the Financial Times, AI companies in China take advantage of the black market to purchase high-end U.S. semiconductors. Relatedly, the CCP has plans to build
10 data center clusters that will utilize brokers to give priority access, including for these banned chips, to government-sponsored AI research.
“Traces of China’s underground chip trade can be found on such brokering sites,” The New York Times reported. “The Suzhou one, which partners with AliCloud and Tencent Cloud, is advertising computing clusters that include advanced Nvidia H100 chips, which the US has always banned from sale to China.”
With the CCP’s unprecedented and illicit access to global technologies of information and control, through, for example, the widespread adoption of AI paired with the Internet of Things, the United States and our allies are increasingly at risk. For policymakers, the concern should be whether these Chinese nationals who research AI in the United States and our allied countries are helping or hurting us more. It is certainly a “conundrum,” in which we find ourselves increasingly dependent on a country that treats us like an enemy. We have not prioritized science, technology, engineering, and math (STEM) students from allied democracies, and now, the chickens are coming home to roost.
In hindsight, it all traces back to policy mistakes from the 1970s, when the CCP sought to
modernize China, including by negotiating with the Carter administration to admit hundreds of Chinese national students. Asked in 1978 whether he worried about Chinese nationals wanting to stay in the United States after their studies, as with Soviet citizens, CCP leader Deng Xiaoping responded that his citizens were more loyal than those of other countries.
Chinese nationals in U.S. universities grew in number, and with time, U.S. universities became dependent upon them for the high tuition they paid. China has rapidly modernized because of those few who do return to China, to the point at which some of the People’s Liberation Army’s capabilities—for example, in hypersonic missiles—rival or exceed those of the United States.
Now, the notion that Chinese nationals in the STEM fields could be
banned from U.S. universities in favor of U.S. and allied nationals is wrongly seen by some as racist. In fact, there is no such concern about Chinese Americans or Taiwan nationals. Rather, the concern is with a nationality, not a race.
Given that the CCP is in control of students if they return home—and is totalitarian, genocidal, and seeks global hegemony—consideration of evolving academic admissions policy vis-à-vis China is simply about democracies trying to determine how best to survive in a rapidly changing and unpredictable technological environment.
Views expressed in this article are opinions of the author and do not necessarily reflect the views of The Epoch Times.
