The rise of Chinese artificial intelligence (AI) firm DeepSeek caused a meltdown in US tech stocks on Monday (Jan 27).
Nvidia on Monday suffered a wipeout of nearly US$600 billion in its market value – the most for any stock ever on Wall Street.
DeepSeek soared to the top of Apple Store’s download charts, as an alternative that is said to be able to perform on a par or even outperform Open AI and Meta’s models – and able to be built more cheaply and take up less power.
The DeepSeek-R1 model is 20 to 50 times cheaper to use than the OpenAI o1 model, depending on the task, according to a post on DeepSeek’s official WeChat account.
Who is the founder and entity behind DeepSeek
Who is behind this Chinese startup that has shaken up US tech titans and other global stocks that are part of the AI supply chain?
DeepSeek is based in Hangzhou and its controlling shareholder is Liang Wenfeng. The 40-year-old graduated from Zhejiang University with a degree in AI, said Chinese media reports.
BT in your inbox
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
He co-founded quantitative hedge fund High-Flyer Quant in 2015, which became known for innovatively using AI-driven trading strategies, according to Chinese media publication CGTN.
The fund announced in 2023 that it was going beyond trading to focus on creating a new research arm to explore “artificial general intelligence”, Reuters reported.
High-Flyer also owns patents related to chip clusters used to train AI models, according to Reuters, citing Chinese corporate records.
Liang began buying thousands of Nvidia graphics processors in 2021 as an AI side project, the Financial Times reported. That was before the Joe Biden administration began limiting US exports of AI chips to China.
High-Flyer’s AI unit had said it owns and operates a cluster of 10,000 A100 chips as at 2022, Reuters reported.
In a July 2024 interview with Waves republished in domestic publication China Academy, Liang said, on his ambitions for DeepSeek and China’s overall AI strategy: “For years, Chinese companies have been accustomed to leveraging technological innovations developed elsewhere and monetising them through applications.
“But this isn’t sustainable. This time, our goal isn’t quick profits but advancing the technological frontier to drive ecosystem growth.”
China’s top leadership has taken note of DeepSeek. On Jan 20, before the DeepSeek-R1 model was released, Liang attended a closed-door symposium for businessmen and others in the industry hosted by Chinese Premier Li Qiang, said Chinese state media.
Could DeepSeek suffer any US “crossfire”?
There’s “low imminent risk” for DeepSeek to be penalised by the US government, said Morningstar in a Tuesday note.
The research firm noted that DeepSeek has recently restricted new registrations to Chinese numbers, hence limiting the number of US users.
“They have not released any commercial versions or monetisation plans, so there’s no direct monetary impact even if new restrictions are rolled out. They have not accepted outside investors (all funding comes from founder’s quant fund) as there’s no pressure for anyone to divest,” said Morningstar analyst Phelix Lee.
As DeepSeek is open source, the most probable scenario is that US tech companies could leverage the public code to refine their own models, he added. That could result in lower AI cloud and computation costs globally, Lee noted.
