Founded by hedge fund manager Liang Wenfeng, the Chinese tech startup DeepSeek is a serious threat to the U.S. AI monopoly. The company is based in an AI lab in Hangzhou, eastern China.
Since revealing DeepSeek's R1 model, which generated international attention, the 40-year-old founder Liang has emerged as a Chinese national hero. Chinese Premier Li Qiang even invited him to a symposium in Beijing last week.
Early life and career
Wenfeng was the son of a primary school teacher and was born in a small 'fifth-tier city' in Guangdong, China, in the 1980s, according to Business Insider.
Because of his early academic success, he was accepted to Zhejiang University, one of the most prominent universities in China, where he graduated with a bachelor's and master's degree.
Quantam hedge fund
He co-founded High Flyer, a quantum hedge fund, in 2016, which immediately became well-known for its creative application of AI-driven investing techniques. DeepSeek receives all of its funding from High Flyer.
By 2021, the hedge fund had completely incorporated artificial intelligence (AI) into its business processes, employing machine learning models to forecast market movements and make informed investment choices.
Award-winning mathematicians and physicists, many of whom hold advanced degrees in subjects like statistics, research, and cybernetics, as well as silver and gold medals from academic competitions, are part of the hedge fund's staff, according to its website.
Deepseek's inception
Wenfeng established DeepSeek in May 2023 as a subsidiary of High-Flyer, which still provides funding for the AI lab. When DeepSeek's V3 model was released in late 2024, it immediately became well-known.
The business disclosed in a ground-breaking paper released in December that it had trained the model with 2,000 Nvidia H800 chips for less than $6 million, a small portion of what its rivals usually spend.
