BANGKOK – The 40-year-old founding father of China’s DeepSeek, an AI startup that has startled markets with its capability to compete with business leaders like OpenAI, saved a low profile as he constructed up a hedge fund after which refined its quantitative fashions to department into artificial intelligence.
Liang Wenfeng, who based DeepSeek in 2023, was born in southern China’s Guangdong and studied in jap China’s Zhejiang province, dwelling to e-commerce big Alibaba and different tech corporations, in response to Chinese language media studies.
The hedge fund he arrange in 2015, Excessive-Flyer Quantitative Funding Administration, developed fashions for computerized inventory buying and selling and started utilizing machine-learning methods to refine these methods.
Like many Chinese language quantitative merchants, Excessive-Flyer was hit by losses when regulators cracked down on such buying and selling up to now yr. Nonetheless, it reportedly manages $8 billion in belongings, ample assets for funding DeepSeek’s AI analysis.
It additionally has plentiful computing energy for AI, since Excessive-Flyer had by 2022 amassed a cluster of 10,000 of California-based Nvidia’s high-performance A100 graphics processor chips which can be used to construct and run AI programs, in response to a post that summer on Chinese language social media platform WeChat. The U.S. soon after restricted sales of these chips to China.
“Factor is, we’re positive now that we need to do that, can do that, and are able to doing this, so we’re among the many best-suited candidates to sort out it at this second,” Liang instructed Waves, a tech media outlet, in 2023.
“Presently, neither tech giants nor startups have an unassailable lead. With OpenAI paving the best way, everyone seems to be working with printed papers and open-source code,” it quoted him as saying.
Liang stated he spends his days studying papers, writing code, and taking part in group discussions, like different researchers.
DeepSeek is exploring what intelligence means, he stated.
“Individuals might imagine there’s some hidden enterprise logic behind this, however it’s primarily pushed by curiosity,” Liang stated.
When DeepSeek was requested, “Who’s Liang Wenfeng?” its first reply was to call a special Chinese language entrepreneur with the identical identify, at the least as spelled in English letters.
When requested: “The place is Liang Wenfeng from and the place did he go to college?” it stated that as of October 2023, the newest information cutoff for DeepSeek’s R1 AI mannequin, “there isn’t a publicly out there details about Liang Wenfeng’s background, together with his place of birth or academic historical past.”
“In case you are referring to the founding father of DeepSeek, particulars about his private life or educational background haven’t been disclosed publicly. For extra details about DeepSeek, you’ll be able to go to its official web site,” it stated.
Liang’s targeted strategy matches in together with his willpower to push AI studying ahead. After a long time of counting on innovation from the West, he says China must be making its personal contributions.
“What we see is that Chinese language AI can’t be within the place of following eternally. We frequently say that there’s a hole of 1 or two years between Chinese language AI and the US, however the actual hole is the distinction between originality and imitation,” he stated in one other Waves interview in November. “If this doesn’t change, China will at all times be solely a follower — so some exploration is inescapable.”
___
Related Press author Ken Moritsugu in Beijing contributed to this report.
Copyright 2025 The Related Press. All rights reserved. This materials is probably not printed, broadcast, rewritten or redistributed with out permission.
