Huawei's Genius Youth Program Is Creating Chinas Physical AI Army
Seven years after Huawei began offering up to 47 million yen to recruit teenage prodigies, former participants are now founding the companies driving Chinas physical AI boom. The genius youth program was never just recruitment — it was an incubation strategy born from sanctions pressure.
The 47 Million Yen Bet That Became a Movement
In 2019, Huawei announced it would pay up to 182万 yuan — roughly 47 million yen — to recruit brilliant young researchers, with no required university pedigree. The company called them “genius youth.” At the time, it read like a publicity stunt or a panic response to U.S. sanctions. Seven years later, the answer is clearer: it was something far more calculated.
The recipients of those salaries were not teenagers in the literal sense. Most held PhDs. Many had already published papers or won international competitions. What made them rare was their depth — not breadth — in specific technical domains like AI chips, autonomous driving, and computer vision.
Between 2020 and 2021 alone, over 300 of the roughly 26,000 new graduates Huawei hired fell into the genius youth bracket. That is not a small sample. It is a cohort.
Where They Ended Up
The most striking evidence of the program’s impact is emerging now, not from Huawei itself but from the companies its former recruits have built.
A 2026 analysis by Chinese tech media Zhixidongxi identified at least nine former genius youth who have either founded or taken technical leadership roles at physical AI companies. The publicly reported funding for those ventures totals approximately 77亿 yuan — roughly 180 billion yen.
Peng Zhihui, known online as “稚暉君” for his early YouTube-style videos on electronics and robotics, joined Huawei in 2020 at the top salary tier. While there, he worked on Ascend, Huawei’s AI semiconductor line. In 2021, he built an autonomous bicycle on his days off and posted it online — a self-balancing bike that navigated obstacles without falling. He left Huawei at the end of 2022 and co-founded Agibot the following year. By June 2026, the company had shipped its 15,000th robot. It now stands alongside Unitree as one of Chinas leading physical AI firms.
Ding Wenchao, who joined Huawei’s automotive division in 2020 after earning a PhD from the University of Hong Kong, became a core engineer on the company’s first-generation autonomous driving system. He departed in 2023, moved to Fudan University to study humanoid robots, and in 2025 co-founded TARS with former Huawei colleagues. TARS develops wheeled industrial robots and bipedal general-purpose robots. In April 2026, it raised 455 million dollars in a single pre-A round — a record for Chinas physical AI sector at the time — at a post-money valuation of roughly 130亿 yuan.
Huang Qingqiu, also hired in 2020 for the automotive group, worked on LiDAR recognition algorithms before becoming Huawei’s autonomous driving AI lead. He left in 2025 to co-found Morphi, which builds general-purpose robots. Within a year and a half, Morphi closed an angel round exceeding 10亿 yuan from investors including Alibaba and Tencent, valuing the company at over 70亿 yuan.
These are not outliers. They are a pattern.
The Ecosystem Effect
What makes this remarkable is that several of these founders did not even join Huawei. Chen Yuanpei, born in the 2000s, passed the genius youth selection in 2024 and received an offer — then declined it. He instead co-founded PsiBot, a physical AI startup focused on robotic dexterous manipulation through reinforcement learning. The company, built largely by someone who studied civil engineering as an undergraduate before pivoting to robotics through a DJI-hosted competition, reported raising 100 million dollars at a 14.8 billion dollar valuation in July 2026 — making it a unicorn.
This is the program working exactly as intended, and perhaps even beyond its designers’ expectations. Huawei did not need these people to stay forever. It needed them to absorb its research culture, work on hard problems at scale, and then carry that capability into the broader ecosystem.
Former genius youth from Huawei’s AI research division, the Noah’s Ark Lab, have also begun entering the fray. Li Yinchuan, who led generative AI and large model projects at the lab, left in 2025 to co-found Knowin, which develops decision-making systems for home robots. Zhou Kaiwen, another lab alumnus, joined him. Knowin raised 5亿 yuan in August 2026.
Chinese tech media have started describing Huawei not as a company that lost talent to competitors, but as an incubator that is feeding an entire industry. The genius youth program, in other words, turned a talent drain risk into a talent multiplication strategy.
Why Sanctions Made This Possible
The genius youth program launched in 2019, the same year Huawei’s automotive division was created and the same year U.S. sanctions began tightening. The timing was not accidental.
When the United States restricted Huawei’s access to advanced semiconductors and software, the company faced an existential question: how do you maintain technological leadership when half your supply chain is cut off? One answer was to invest aggressively in the people who could solve problems the old solutions could not.
Paying 47 million yen to a 24-year-old PhD with no famous university credential is expensive. But it is cheaper than trying to build the same capabilities from scratch later, or watching them diffuse to competitors through ordinary hiring channels.
The genius youth program gave its recruits three things Huawei could offer that startups could not at the time: world-class research problems, senior mentors who had built those problems into reality, and access to infrastructure — compute, data, manufacturing partnerships — that no early-stage company could match.
Many of those recruits eventually left. But by then, they had spent years solving exactly the kinds of problems that would make Chinas physical AI boom possible. They had learned to think at Huawei’s scale.
What This Means for the Next Decade
The physical AI sector — encompassing humanoid robots, autonomous vehicles, and embodied intelligence systems — is where the next wave of Chinas tech dominance is likely to be measured. American companies like Tesla and Figure AI are competing in the same space, but Chinas advantage here is different: it is a matter of velocity, cost, and manufacturing integration.
Huawei’s genius youth program appears designed to compound all three. The former recruits are now building companies that can prototype faster, iterate cheaper, and ship earlier precisely because they learned at Huawei how to operate at the intersection of software, hardware, and systems engineering.
A recent parallel story from Japan offers a useful comparison. Japanese real estate firm Gagoken Capital announced a top talent hiring program in 2026 offering newly graduated students a starting salary of 100万 yen per month — roughly 14 million yen annually — for candidates with exceptional records in entrepreneurship, AI, research, or the arts. Over 1,300 applied; about 10 received offers so far. The strategy mirrors Huawei’s: pay above market for rare talent upfront, then wait to see what they produce.
The difference is scale and track record. Huawei has been running its program for seven years and can point to nine companies, billions in combined funding, and at least one unicorn among its former recruits. Gagoken is still waiting to learn whether its bet pays off.
The Unfinished Question
The genius youth program is not finished. Huawei continues to recruit under it. New cohorts are entering the system now, working on problems that did not exist when the program launched — problems in embodied AI, robot learning, and human-machine interaction that will define the next decade.
The question is not whether the program worked for the first generation. The answer is already visible in the funding rounds and product launches of former recruits. The question is whether Huawei can replicate that success with people who are entering a landscape far more crowded than it was in 2019.
Chinas physical AI sector is now attracting capital and talent from every direction. The genius youth who left Huawei in 2023 and 2024 are no longer the only people with Huawei-scale experience building robots and autonomous systems. New competitors are training their own cohorts.
But for now, the first wave is still ahead of the curve. And the 47 million yen they each commanded in 2019 looks less like a recruitment expense and more like the cheapest investment Huawei ever made.