Liang Wenfeng, the founder of the Chinese AI startup DeepSeek, has rapidly risen as a notable figure in the artificial intelligence landscape. While OpenAI CEO Sam Altman is widely recognized, Wenfeng’s journey to prominence has remained relatively low-profile outside of China.
Born in the 1980s in a modest “fifth-tier city” in Guangdong, China, Wenfeng was the son of a primary school teacher, according to Business Insider.
His early academic excellence led him to Zhejiang University, one of China’s most prestigious institutions, where he earned both undergraduate and graduate degrees.
From Hedge Funds to AI Innovation
In 2015, Wenfeng co-founded High-Flyer, a quantitative hedge fund, alongside two classmates from Zhejiang University. The fund, which utilized mathematics and artificial intelligence for investment strategies, managed over $10 billion in assets by 2019, according to its website.
Wenfeng’s transition from finance to AI was sparked during his time at High-Flyer. In 2021, he began acquiring GPUs from Nvidia to build a massive 10,000-chip cluster for AI model training. Despite being described by colleagues as a “very nerdy guy with a terrible hairstyle,” Wenfeng’s ambitious vision set the foundation for what would become DeepSeek.
The Birth of DeepSeek
In May 2023, Wenfeng launched DeepSeek as an offshoot of High-Flyer, which continues to fund the AI lab. DeepSeek quickly gained attention with the release of its V3 model in late 2024. In a groundbreaking paper published in December, the company revealed it had trained the model using 2,000 Nvidia H800 chips at a cost of under $6 million, a fraction of what its competitors typically spend.
The R1 Model Shakes Silicon Valley
On January 20, 2025, DeepSeek made headlines again with the launch of its R1 model, further cementing its reputation as an AI innovator. The model’s capabilities and efficiency have intensified global discussions about the growing influence of Chinese tech firms in the AI sector.
