Half of the world's AI talents are Chinese. Why does China still face a talent shortage?

"Fifty percent of the world's artificial intelligence researchers are Chinese. You can't stop them from making progress in the field of AI," said Jensen Huang, CEO of NVIDIA, when attending the Taipei International Computer Show last month.

This is not the first time that Jensen Huang has emphasized the importance of the Chinese in the international arena of the AI industry. Behind this statement lies the fierce competition for AI talent globally, especially between China and the United States.

Chinese AI Elites Take Over Silicon Valley

As Jensen Huang said, nowadays, Chinese faces are active in every top AI laboratory in the United States. Among the top AI research institutions in the US, 38% of the talent comes from China, slightly higher than the 37% from the US itself. A report from the think - tank MacroPolo shows that nearly 40% of the top AI researchers in the US graduated from Chinese universities.

Chinese people play an indispensable role in the top large - model teams in the United States.

Last year, when OpenAI launched its first native multimodal model, GPT - 4o, 6 out of the 17 key team members were Chinese, coming from universities such as Tsinghua University, Peking University, Shanghai Jiao Tong University, and the University of Science and Technology of China.

Among the 12 founding members of Elon Musk's AI company, xAI, there are also five Chinese. At the press conference for the release of the Grok 3 reasoning model, Musk gave the center position to Jimmy Ba and Wu Huaiyu, which shows his great importance attached to Chinese scientists.

Live - broadcast picture. Second from the left is Jimmy Ba, third from the left is Wu Huaiyu. Image source: Internet

In the Gemini technical report released by Google, among the 837 authors, 141 are Chinese scholars, including Ed Chi, a Chinese - American and Google's chief scientist.

The influence of the Chinese is not only reflected at the technical level. In the AI era, they are gradually taking control of the industry's voice, replacing the dominant position of Indian - Americans in the upper echelons of Silicon Valley in the past. The four major US chip giants, NVIDIA, AMD, Broadcom, and Intel, are now all led by Chinese people. The Chinese are no longer just basic executors but real industry leaders.

On the other hand, the Chinese AI elites shining in the Silicon Valley AI circle directly constitute the greatest competitive force for the development of China's domestic AI industry.

Last week, Meta announced a $14.8 billion investment in Scale AI, a data - annotation startup founded by the Chinese - American "genius teenager" Alexandr Wang. This is the second - largest deal in Meta's acquisition history, second only to its $19 billion acquisition of WhatsApp. Scale AI once placed a full - page advertisement in The Washington Post addressed to the US President, saying, "Dear President Trump, the United States must win this AI war."

The success of Chinese AI elites in the United States may not necessarily be a source of pride for China, but it definitely contributes to the victory of the United States.

Half of the World's AI Talent Is Chinese, but Does China Still Lack People?

While Chinese AI elites are thriving in Silicon Valley, the shortage of AI talent in China's domestic market presents a fragmented reality: data shows that currently, the shortage of AI talent in China exceeds 5 million, with a supply - to - demand ratio of 1:10. Among them, the talent in the basic layer is the most scarce, resulting in insufficient momentum for AI technological innovation.

As of 2024, more than 500 universities across the country have offered AI majors, with about 40,000 students enrolled. This is far from enough to fill the huge talent gap in the entire industry.

This shortage is not only reflected in quantity but also in quality. There is a mismatch between the talent output by universities and the needs of enterprises.

Wang Hao (a pseudonym) is an ordinary undergraduate student from a second - tier university. When filling out the college entrance examination application, he chose the newly launched and popular major - artificial intelligence. He has an average math foundation and has no experience in programming. "I had to teach myself Python. There was almost no hands - on practice in the courses of machine learning and deep learning. I had no exposure to real - world enterprise projects before graduation." Wang Hao's "superficial" learning experience is not an isolated case but a common problem in current university AI majors.

A large number of "non - double - first - class" graduates majoring in AI face the cruel reality of "unemployment upon graduation" during the job - hunting season. Many core AI positions require a master's degree or above. Even if their majors match, they are screened out at the resume - screening stage.

Not only are students confused, but university teachers are also groping in the dark. The development of AI is changing rapidly, with new tools emerging constantly and the curriculum system being continuously adjusted. The teachers of AI majors are basically part - time teachers from computer science majors, teaching while learning themselves. The lagging teaching system makes it difficult for the students it cultivates to meet the latest needs of enterprises.

The result is what we can see: The low - threshold positions are saturated, while there is a shortage of high - end talent. The high - end talent that China truly needs is still flowing overseas. As the world's largest exporter of AI talent, only 12% of the most elite AI talent in China choose to work in China first. For example, Yao Shunyu, a legendary special - award winner from the Department of Physics at Tsinghua University, joined the Claude team at Anthropic. It's hard not to worry if China's AI talent cultivation is once again on the old path of making wedding clothes for other countries?

How Long Can the United States' Talent Siphon Effect Last?

For many years, the United States has been attracting high - tech talent from around the world with its superior scientific research and employment environment. It is still the top choice for the world's most elite AI talent to work.

However, in recent years, the US government has increased restrictions and scrutiny on international students, especially Chinese students. This month, the White House announced visa restrictions on international students at Harvard University, which may affect a large number of new students' enrollment. In addition, the government has significantly reduced research funding and stopped funding more than 400 research projects, leaving a large number of doctoral students and post - docs facing career crises.

These risks not only deter many students who originally planned to study in the United States but also make Chinese scholars already in the US consider returning to their home country or going to other countries, staying away from this "troubled place". According to a survey by the world - leading academic journal Nature, up to 75% of US scientists are considering leaving the United States.

The cold wave in the US scientific research community coincides with the generous offers from Chinese enterprises, which is beneficial for China to attract international AI talent and for Chinese scientists in the US to return. Recently, Alex Lamb, a former scientist at Google DeepMind and a student of a Turing Award winner, announced that he would join Tsinghua University. Luo Jianlan, a 90s - born expert in the field of robotics who has worked at Google and DeepMind, joined a Shanghai startup, ZHIYUAN ROBOTICS, as the chief scientist. Luo graduated from Wuhan University of Technology as an undergraduate and obtained his doctorate from the University of California, Berkeley.

The departure of top AI talent from the United States may indicate a reshuffle of the global AI talent landscape.

The Shortage of AI Talent Is a Global Problem

If we compare the architecture of artificial intelligence to a kitchen, then data is the ingredients, algorithms are the recipes, and computing power is the gas. These are the three cornerstones for the efficient operation of the kitchen. However, it is the "chef" who really determines the quality of the cooking results. They are responsible for screening data, designing model architectures, and scheduling computing resources. Therefore, talent is the core asset that countries compete for in the global AI race.

As the United States loses its talent dividend, it's not only China that can "pick up the bargain", but also European countries, Japan, and Singapore.

French President Emmanuel Macron called on social media platform X, "Scientists from around the world, choose France, choose Europe!" Japan provides high - value subsidies for Indian students in the field of AI, with an average of up to 145,000 RMB per person. Singapore has launched the NAIS 2.0 strategy, hoping to attract the world's top AI talent to work in Singapore and increase the number of AI practitioners to 15,000.

During World War II, the influx of top scientists such as Albert Einstein and Wernher von Braun brought a technological leap to the United States. Today, the flow of AI talent may reshape the new pattern of global technological competition in the era of the "Fourth Industrial Revolution".

In this race, China has a large talent base, the ecological advantage of rapid implementation of AI applications, and strong support at the national level. In the battle for AI talent, China not only needs to "recruit talent" but also needs to retain and attract talent by improving the talent cultivation system and the entrepreneurial environment.

In the AI era, whoever has the talent has the future.

This article is from the WeChat official account "Hundun University" (ID: hundun - university). The author is Hundun Academy. It is published by 36Kr with authorization.