How Liang Wenfeng’s DeepSeek is Taking on OpenAI

The Maverick Behind China’s AI Challenger

In the world of artificial intelligence, where tech giants like OpenAI, Google, and Microsoft dominate the field with vast resources and thousands of researchers, a 40-year-old Chinese entrepreneur, Liang Wenfeng (梁文鋒), is proving that size isn’t everything. With a team of only 139 researchers, his AI startup DeepSeek is making waves in the industry, challenging the behemoths with an innovative approach that breaks conventional hiring norms and embraces open-source development.

A Radical Approach to Innovation

Liang Wenfeng, known for his unkempt hair and eccentric personality, has always been ahead of the curve. Even in his middle school years, he displayed an exceptional aptitude for mathematics, studying university-level material long before his peers. His belief that "innovation declines when a company falls into inertia" has driven him to adopt an unconventional hiring strategy: he primarily recruits young, inexperienced talent with no overseas education background.

Unlike many AI companies that prioritize experienced researchers, DeepSeek’s workforce consists mainly of fresh graduates and employees with only 1-2 years of experience. In fact, at DeepSeek, an employee with just three to five years of experience is considered a veteran, and those with over eight years of experience are outright excluded from recruitment.

This philosophy stems from Liang’s conviction that youth breeds confidence and creativity, making younger employees more adaptable and willing to take risks. In a recent interview, he stated, "Innovation starts with confidence. Young people tend to be more innovative."

Competing with Giants on a Shoestring Budget

DeepSeek’s workforce of 139 researchers is a mere 11.6% of OpenAI’s 1,200 AI engineers, yet it has achieved comparable breakthroughs. This feat is even more remarkable considering that tech giants like Microsoft and Google have 7,000 and 5,000 AI engineers, respectively.

DeepSeek’s agility and efficiency come from its strong commitment to open-source AI development. Liang believes that open-source software is not just a business practice but a cultural movement. He credits Silicon Valley’s rise in the tech industry to the open-source ecosystem and applies the same principle at DeepSeek.

An AI Prodigy in the Making: Luo Fuli

One of DeepSeek’s rising stars is Luo Fuli (羅福莉, 30), a talented AI researcher who joined the company in 2022. Luo, a graduate of Beijing Normal University and Peking University, quickly gained recognition as an "AI prodigy." Her exceptional contributions even caught the attention of Xiaomi, which reportedly offered her an annual salary of 10 million yuan (approximately $1.4 million USD) to lure her away—an offer she declined.

DeepSeek’s unconventional work environment further fuels its innovation. Unlike traditional corporations that emphasize hierarchy and seniority, DeepSeek operates more like a university research lab than a corporate office. Employees are encouraged to collaborate freely, breaking down rigid structures that often stifle creativity.

From a ‘Nerd’ to a Global AI Visionary

Born in 1985 in Guangdong Province, Liang Wenfeng’s path to AI entrepreneurship was anything but ordinary. After excelling in mathematics during his school years, he entered Zhejiang University in 2002, earning a bachelor’s degree in electronic and information engineering in 2007 and a master’s degree in telecommunications engineering in 2010.

However, his early career was not in AI but quantitative trading. During the 2008 global financial crisis, he became fascinated by statistical modeling and mathematical finance, eventually launching a quant hedge fund called Highflyer in 2015. It was during this time that he started exploring deep learning, leading him to invest 1 billion yuan (~$140 million USD) in AI research, including the creation of Firefly 1, a massive AI training cluster powered by 10,000 Nvidia A100 GPUs.

At the time, only a handful of Chinese tech giants—such as ByteDance and Alibaba—had access to such computing power. His ambitious AI vision was initially met with skepticism, with former colleagues recalling that people "didn’t take his dreams seriously." One even described him as a "nerd with a terrible hairstyle."

But Liang was undeterred. His vision was far greater than just developing better AI models—he wanted to build an artificial general intelligence (AGI) system on par with human cognition.

The Future of DeepSeek: Can It Rival OpenAI?

Liang Wenfeng’s bold approach—focusing on young talent, embracing open-source collaboration, and prioritizing lean operations over scale—has turned DeepSeek into one of China’s most promising AI companies.

As OpenAI, Google, and Microsoft continue to invest billions into AI research, DeepSeek is proving that brainpower and agility can rival deep pockets. While it remains to be seen whether DeepSeek can truly compete with OpenAI in the long run, one thing is certain: Liang Wenfeng is redefining how AI startups operate, and the industry is paying close attention.