AI Labs
Liang Wenfeng
Founder and CEO, DeepSeek
The hedge-fund quant whose low-cost models rattled Silicon Valley and Wall Street in a single week.
Score 88/100
Why they’re on the list
Liang's DeepSeek proved frontier-grade AI models could be trained far more cheaply than Silicon Valley assumed, triggering a global market shock in 2025 and repositioning China as a genuine peer in the AI race.
Liang Wenfeng is the founder of DeepSeek, the Hangzhou-based artificial intelligence lab that in early 2025 triggered one of the sharpest shocks the AI industry had yet experienced, when its low-cost, high-performing models raised fundamental questions about how much capital was actually needed to compete at the frontier. Born in 1985 in Wuchuan, Guangdong province, Liang studied electronic information engineering and later information and communication engineering at Zhejiang University, completing his master's degree in 2010 with a thesis on object-tracking algorithms.
Liang's path to AI ran through quantitative finance rather than a research lab. He co-founded investment firms in the early 2010s before establishing High-Flyer, a quantitative hedge fund, in 2016, using machine learning techniques to drive trading strategies and building substantial in-house computing infrastructure in the process. That infrastructure, including large stockpiles of GPUs accumulated before later US export controls tightened, became the foundation for High-Flyer's AI research arm, founded in 2019, and ultimately for DeepSeek, launched as a dedicated AI company in May 2023.
DeepSeek's ambition, stated openly by Liang, has been nothing less than artificial general intelligence, pursued with a research culture that eschews the long working hours typical of Chinese technology firms; Liang has said publicly that DeepSeek staff do not work mandatory overtime or operate under strict KPIs, an unusual stance in China's tech sector. The company released a rapid succession of increasingly capable models, culminating in DeepSeek-V3 in December 2024 and the R1 reasoning model shortly after, both offered with openly published weights and training details at a fraction of the compute cost claimed by leading US labs.
The market reaction was immediate and severe: DeepSeek's efficiency claims wiped billions of dollars off US technology and chipmaker valuations in a single trading session and forced OpenAI, Google and Anthropic to publicly reassess their own cost assumptions. Liang, previously an obscure figure even within Chinese tech circles, was thrust into the international spotlight, later meeting Chinese premier Li Qiang and appearing on lists including Time's 100 Most Influential People in AI and Fortune's most powerful people in business.
Liang has remained notably private and rarely gives interviews, preferring to let DeepSeek's models and research papers speak for the company. His continued backing of High-Flyer's capital and refusal, at least initially, to take outside venture funding has let him keep tight control over DeepSeek's direction, cementing his position as one of the most consequential and least understood figures shaping the global AI race from China.
Career timeline
- 1985Born in Wuchuan, Guangdong, China
- 2010Completes master's degree in information and communication engineering, Zhejiang University
- 2013Founds early quantitative investment venture
- 2016Co-founds High-Flyer quantitative hedge fund
- 2019Establishes High-Flyer AI research division
- 2023Launches DeepSeek, declaring pursuit of AGI
- 2024Releases DeepSeek-V3, followed swiftly by the R1 reasoning model
- 2025DeepSeek's efficiency claims trigger a global tech stock sell-off