Is the spotlight shifting? Will the Chinese dominate the AI era?
In the summer of 2025 in Silicon Valley, the sun was still scorching hot, but it couldn't stop a "talent war" around Chinese AI talents.
On the other hand, China's domestic AI is also rising at an accelerated pace.
Beyond the "talent war", it is an industrial paradigm revolution led by Chinese scientists, engineers, and entrepreneurs.
The global AI chess game is entering a new era where the Chinese hold the pieces.
01 The Talent War
In 2023, after Elon Musk parted ways with OpenAI, he founded the artificial intelligence company xAI. On July 9th, he publicly launched the fourth - generation upgrade of xAI.
However, at the press conference, it wasn't Musk who occupied the center stage, but two Chinese scientists - Wu Yuhuai and Jimmy Ba.
Wu Yuhuai was born in Hangzhou. As a post - 1995 academic overachiever, he graduated from the University of New Brunswick in Canada with a perfect GPA and then obtained a doctorate in machine learning from the University of Toronto, studying under the "Father of Deep Learning", Geoffrey Hinton.
Jimmy Ba also graduated from the University of Toronto. After completing his undergraduate, master's, and doctoral studies, he stayed at the university as a faculty member and is currently an assistant professor in the Department of Computer Science at the university. He is also a favorite student of Hinton.
After the press conference, a group photo of the xAI team quickly circulated on social platforms. In the photo, up to 80% of the faces were Chinese, which was quite remarkable.
In fact, among the 12 - member founding team of xAI, Chinese scientists accounted for five seats: besides Wu Yuhuai and Jimmy Ba, there were also Yang Ge, a top student from the Department of Mathematics at Harvard University, Dai Zihang from the Department of Computer Science at Tsinghua University, and Zhang Guodong, an alumnus from the Department of Electronics at Zhejiang University.
Yang Ge is from Hunan. He went to the United States to study since junior high school. He studied in the Department of Mathematics at Harvard University for both his undergraduate and master's degrees, studying under Professor Shing - tung Yau. When he graduated, he planned to join Google, but was persuaded by Yau's words "Don't go to a lousy company like Google" and instead joined the Microsoft Research Institute.
Dai Zihang graduated from the Department of Computer Science at Tsinghua University. He obtained a doctorate in computer science from Carnegie Mellon University in 2020 and then joined the Google Research Institute. He has in - depth knowledge in the field of natural language processing (NLP) and is especially good at optimizing and innovating the Transformer architecture.
Zhang Guodong studied information engineering at Zhejiang University for his undergraduate degree and then went to the University of Toronto to pursue his master's and doctoral degrees. After joining xAI, his main research focuses on improving AI training efficiency and model stability.
Now, the proportion of Chinese in Musk's team is still increasing, and Chinese scientists are the "brains" of xAI.
Even more eager for Chinese scientists than Musk is another Silicon Valley tech giant - Mark Zuckerberg, the chairman and CEO of Meta.
Currently, he is initiating the most intense "talent war" in Silicon Valley in the past decade, with Chinese engineers as his primary target.
The first person he poached was Pang Ruoming, the head of Apple's basic model team. According to Bloomberg, after Pang Ruoming switched to Meta, he signed a sky - high contract worth a total of $200 million. Following him, his deputy, Tom Gunter, and several core members also jumped to Meta one after another.
According to insiders, Zuckerberg is personally overseeing a project codenamed "Super - intelligent Laboratory", targeting the world's top 50 AI experts. A industry analyst privately described it as: "This is a contemporary 'Manhattan Project', but this time, the target is not physicists, but Chinese artificial intelligence experts."
Among the 14 - person poaching list announced by Meta so far, as many as 8 are Chinese, mostly from OpenAI and Google DeepMind.
The first batch of confirmed new members of Meta includes Zhao Shengjia from Tsinghua University, Yu Jiahui from the Juvenile Class of the University of Science and Technology of China, Bi Shuchao from Zhejiang University, and Ren Hongyu from Peking University. They are all the technical pillars behind OpenAI's core breakthroughs such as GPT - 4.1.
Wired magazine reported that Meta offers top AI researchers a total compensation of $300 million over four years, with more than $100 million payable in the first year, including high salaries, stocks, and signing bonuses. Sam Altman, the CEO of OpenAI, said bluntly:
"The treatment Meta offers to AI experts is comparable to that of NBA stars or CEOs of Fortune 500 companies."
In addition, Google and Microsoft are also taking action, and they are also focusing on Chinese scientists. For example, Google poached He Kaiming, a top vision expert who graduated from the Department of Physics' Basic Science Class at Tsinghua University and has a tenure at MIT; Jensen Huang also recruited two Chinese students and promoted one of them to vice - president and chief scientist.
Behind the crazy talent poaching is the increasing anxiety in Silicon Valley.
The AI wave is surging, and top - notch talents are the most important force in determining who will hold the scepter of the future.
When Meta poached Yu Jiahui from OpenAI with an annual salary of $100 million, Chen Xinhan, the research director of OpenAI, angrily wrote in an internal letter: "It's like someone breaking into our home and stealing the most precious thing."
To protect themselves, top Silicon Valley companies have stepped up their efforts: Google DeepMind launched a "shadow compensation system"; and OpenAI set aside $4.4 billion for equity incentives, which is even more than 19% of its total revenue in 2024.
The reason why Chinese engineers are so popular is that they possess the "rarest and most valuable resources" in the AI era. The joke on social platforms, "Chinese scientists are taking over Silicon Valley", is becoming a reality.
02 Unique Advantages
In fact, since ChatGPT emerged in late 2022, Chinese figures have become more prominent on the global AI stage. Now, the global industry is very clear and certain that researchers with a Chinese background have become the core intellectual backbone of leading AI institutions in the United States.
What's particularly noticeable is that while "Chinese scientists are taking over Silicon Valley", China's AI is catching up comprehensively, and even gradually overtaking and taking the lead.
According to a report jointly released by the China Investment Promotion Office under the United Nations Industrial Development Organization (UNIDO) and Shenzhen Dongbi Data, among the world's top 100 AI experts, 65 are from China. Among them, 50 work in Chinese domestic institutions, and 15 work in US research centers.
Correspondingly, Chinese scholars also dominate in the output of core AI achievements.
The 2025 AI Index Report from Stanford University shows that between 2010 and 2023, the number of AI patents soared from 3,833 to 122,511, with a 29.6% increase in the past year alone. As of 2023, China led in the total number of AI patents, accounting for 69.7% of all granted patents.
Chinese - background scientists can shine in the AI era because the technological paradigm shift in the era of large models has magnified the comparative advantages of Chinese engineers. A Chinese engineer who has switched between Meta and OpenAI shared: "It's not that they prefer Chinese people. They've just recognized a reality: Chinese engineers are a triple guarantee of efficiency, stability, and innovation."
From chips to models, from hardware - driven to algorithm - driven, the core of AI competition has shifted to algorithm design, architecture innovation, and data understanding. And these are exactly the strengths of Chinese engineers.
Behind this are both the systematic promotion of education and policies and the deep - seated support of culture and social psychology.
First, driven by national strategy, China issued the "Innovation Action Plan for Artificial Intelligence in Higher Education Institutions" in 2018, comprehensively launching the systematic cultivation of AI talents. So far, 535 universities across the country have established AI majors.
China's long - term accumulation in basic mathematics education has continuously provided "underlying computing power" for the AI field. In the International Mathematical Olympiad (IMO), the Chinese team has won the world championship 15 times, leading the world and laying a solid foundation in mathematical ability early on.
The close cooperation between China's top universities and enterprises such as Alibaba, Baidu, Tencent, and Huawei, the joint establishment of laboratories and research platforms, and the booming development of China's emerging technology industries have also provided support for talent cultivation.
Statistics show that there are currently more than 4,500 AI companies in China, and many entrepreneurs graduated from top - tier universities such as Tsinghua, Peking, and Zhejiang Universities. Behind emerging institutions such as Baichuan, MiniMax, Moonshot, and Zhipu, there are often a group of scholars who still hold teaching positions, running both academic and business operations in parallel, forming a composite research team.
In early 2025, the domestic large model DeepSeek was launched. In just one month after its launch, the cumulative number of downloads exceeded 110 million. It's worth noting that DeepSeek achieved rapid popularity among domestic and international users with almost no marketing investment.
In May, Huawei officially delivered its new - generation AI chip, Ascend 910C, to domestic customers. Its performance far exceeds that of NVIDIA's H20 chip. In specific scenarios, its performance is close to 80% of that of the H100, while the cost is only half of the latter. This breakthrough not only significantly improves the cost - performance ratio of domestic computing power but also breaks the long - term monopoly of US chips in key technologies.
Meanwhile, unicorn technology companies have also achieved revolutionary breakthroughs in their respective fields. Dark Side of the Moon initiated the "long - text revolution" first, and its Kimi intelligent assistant supports context processing of up to 2 million words.
In addition, cultural perception and social psychology are also reasons why AI is more highly regarded and promoted in China.
A report recently released by Stanford University pointed out that only 39% of Americans believe that the benefits of AI outweigh the drawbacks, while in China, this proportion is as high as 83%. This strong "technological optimism" has continuously provided talent and enthusiasm for the AI industry.
Behind Chinese scientists taking over Silicon Valley is the result of a large number of Chinese students going to the United States to study relevant majors. Even in the face of tightened visas and enrollment restrictions, the number of Chinese students in science and engineering doctoral programs is more than twice that of Indian students.
03 Remodeling the Rules
In addition to being the focus of large - company talent wars and occupying high - level positions as scientists, in the new AI era, domestic Chinese talents and overseas Chinese students are also continuously achieving new heights in entrepreneurship.
In July 2025, when Google bought the two founders and the core technology team of the AI programming startup Windsur for $2.4 billion, another AI programming startup, Cognition, led by Chinese entrepreneur Scott Wu, quickly took action and completed the acquisition of the remaining team and intellectual property of Windsur within 72 hours.
This surprise acquisition was called the "Thunderbolt Operation" in the Silicon Valley industry.
Scott Wu is a post - 1990s Chinese. He won three gold medals in the International Olympiad in Informatics (IOI) when he was only 15 to 17 years old and was hailed as a "genius boy".
In March 2024, Cognition, founded by Scott Wu, launched the "world's first AI programmer, Devin" and became famous overnight. One year later, the company completed a $175 million Series A financing, with a valuation of up to $2 billion.
More exciting games are taking place in the field of big - data annotation.
In mid - June 2025, Meta announced that it would acquire a 49% stake in Scale AI for $14.3 billion. The founder of Scale AI is Wang Tao, a 28 - year - old Chinese scientist. After Meta completed the acquisition of Scale AI, Wang Tao, whose net worth had reached $2 billion, would step down as the CEO of Scale and become Meta's chief AI officer, leading the construction of a system for "AI super - intelligence".
Among Silicon Valley's AI startups, compared with Scale, its rival, Surge AI, is now attracting more attention and curiosity.
Surge AI has been established for five years. It has never raised funds and rarely communicates publicly. It is said that it has only one - tenth of the employees of Scale, but it achieved a revenue of $1 billion in 2024. Its founder is also a Chinese - Edwin Chen.
Edwin Chen graduated from the Department of Mathematics and Linguistics at the Massachusetts Institute of Technology. He has worked at Google, Facebook, and Twitter, focusing on machine learning, human - computer interaction, and content review. After founding Surge AI, he targeted the big - model annotation segment with a "high - quality human data platform" as the core.
While these Chinese entrepreneurs are showing their wisdom in Silicon Valley, domestic Chinese entrepreneurs are also breaking the traditional order and becoming rule - makers rather than rule - takers in the new AI era.
When the Chinese open - source model DeepSeek achieved performance comparable to ChatGPT at one - thirtieth of the cost, Silicon Valley giants understand the meaning behind it.
On July 16th, Jensen Huang, the CEO of NVIDIA, appeared in Beijing. Rarely, he took off his iconic leather jacket and wore a Tang suit, and gave a speech in Chinese for the first time. In his speech, he described China's AI market as "vibrant" and mentioned that more than 50% of the world's AI researchers are in China, and it is they who have jointly created the "miracle" of China's AI.
He said, "Chinese developers and entrepreneurs are driving rapid AI innovation. Now, one million developers are involved in this field. Companies like DeepSeek, Alibaba, MiniMax, and Baidu have developed world - class products that are promoting the development of global artificial intelligence."
He also announced that the United States approved NVIDIA to sell H20 chips to China on July 15th. However, the Chinese industry is not blindly welcoming it. Instead, more people believe that China should continue to promote the AI process with domestic chips.
Even the US Treasury Secretary admitted that NVIDIA was allowed to sell to China because domestic Chinese chips have comparable or even better performance than the H20.
This is a change in the rules.
A deeper - level change is that in the AI era, more Chinese scientists are returning to China to start businesses.
Zhu Songchun, a well - known AI scholar from the University of California, Los Angeles, returned to China to establish an AI research institute; Zeng Wenjun, the former chief scientist of Microsoft Research Asia, joined Ningbo Dongfang Institute of Technology; Qi Guojun, an award - winning AI expert, returned to China after working in the United States for more than a decade and joined Westlake University, establishing the "MAPLE Laboratory".
In addition, in emerging AI companies such as Baichuan, MiniMax, and Dark Side of the Moon, there are also a large number of Chinese scientific research backbones with overseas backgrounds. Together, they form an emerging "returning force", driving the rapid upgrade of China's AI in algorithm open - sourcing, model innovation, and large - scale training.
In this global competition of technological forces, Chinese scientists, engineers, and entrepreneurs are no longer just the technical backbones "poached" by Silicon Valley with high salaries. They are the new - generation "steersmen" and rule - makers. AI is leading the global technology industry into:
The Chinese