 "The icon for the smartphone app DeepSeek is seen on a smartphone screen in Beijing. AP")
China’s innovative AI model, DeepSeek, has caught everyone’s attention.
DeepSeek’s AI assistant soared to the top spot as the most downloaded free app on Apple’s iPhone store on Monday, driven by growing interest in this ChatGPT rival.
ALSO READ | DeepSeek vs ChatGPT, Gemini, Copilot: Which is the most effective AI chatbot?
Liang Wenfeng, the 40-year-old founder of DeepSeek, has made waves in the US tech market, with shares in AI-related companies such as Nvidia, Microsoft, and Meta dropping on Monday.
What sets DeepSeek apart is its team of young, talented individuals who have managed to develop cutting-edge AI technology on a modest budget.
Let’s take a closer look at the team behind this revolutionary technology:
The people behind DeepSeek
DeepSeek’s team comprises mainly Chinese nationals from some of the country’s leading universities, such as Tsinghua University and Peking University, as per a report by South China Morning Post (SCMP).
While DeepSeek predominantly recruits recent graduates from these prestigious institutions, Liang mentioned that the company also brings in PhD candidates and young AI professionals with only a few years of experience.
The technical paper for the R1 model, for example, was co-authored by around 200 researchers, with 18 core contributors.
Luo Fuli: The 29-year-old “AI Prodigy”
One of the standout talents from DeepSeek is Luo Fuli, a former employee who gained massive attention after Xiaomi founder Lei Jun reportedly offered her an annual package of 10 million yuan (US$1.4 million). However, recent media reports suggest that Luo has not yet accepted the offer.
A master’s graduate from Peking University (PKU), Luo has been described as an “AI prodigy” by Chinese media, according to SCMP.
Luo’s journey began at Beijing Normal University, where she initially faced challenges in computer science but eventually thrived. She then secured a position at Peking University’s Institute of Computational Linguistics, where she published eight papers at the prestigious ACL conference in 2019.
In 2022, she joined DeepSeek, and her expertise in natural language processing played a major role in the development of DeepSeek-V2, which is considered a serious contender to ChatGPT.
Gao Huazuo and Zeng Wangding
Among DeepSeek’s impressive talent pool, Gao Huazuo and Zeng Wangding are recognised for their “key innovations in the research of the MLA architecture,” according to the company.
ALSO READ | DeepSeek FAQs: All questions on new Chinese AI tool answered
Gao graduated from Peking University (PKU) in 2017 with a degree in physics, while Zeng began his master’s degree at the AI Institute of Beijing University of Posts and Telecommunications in 2021, according to SCMP.
These profiles reflect how DeepSeek recruits talent, as many local AI startups tend to recruit more experienced, established researchers or overseas-educated PhDs specialising in computer science.
Other key members
Other important members of the team include Guo Daya, who graduated with a PhD from Sun Yat-sen University in 2023, and Zhu Qihao and Dai Damai, both recent PhD graduates from PKU.
ALSO READ | DeepSeek demolishes Nvidia stock, CEO loses $20 billion in a day: Why was the tech giant worst hit?
Leading the team of AI experts at DeepSeek is the company’s founder, Liang. Described as reserved yet possessing a strong intuition and an eye for technical detail, Liang’s leadership has been praised by the team, a former employee told SCMP on the condition of anonymity as they were not authorised to speak publicly.
Liang Wenfeng: The leader
The founder of DeepSeek established the company in 2023. Born in Guangdong in southern China, he studied in Zhejiang province in the east, a region home to tech giants such as Alibaba, according to Chinese media reports.
Liang had maintained a very low profile until January 20, when he was one of nine individuals invited to give a speech at a closed-door symposium hosted by China’s Premier Li Qiang.
Last week, DeepSeek launched a free AI assistant that the company claims uses less data and costs remarkably less than current services, triggering a global sell-off in tech stocks.
Under Liang’s guidance, DeepSeek chose not to focus on app development. Instead, it concentrated its research talent and resources on creating a model that could match or surpass OpenAI’s offerings, with plans to continue focusing on cutting-edge models that other companies can utilise to develop consumer and enterprise-facing AI products.
Liang’s approach is unique in the Chinese tech industry, which has traditionally relied on adapting innovations from abroad - ranging from smartphone apps to electric vehicles - and scaling them up quickly, often at a faster pace than the countries of origin.
“China’s AI can’t be in the position of following forever. We often say that there is a gap of one or two years between China’s AI and the United States, but the real gap is the difference between originality and imitation,” Liang said in an interview with Waves last year.
ALSO READ | Meet Liang Wenfeng, the ‘nerdy guy with a terrible hairstyle’ who founded DeepSeek
Liang grew up in Guangdong, a province that led China’s adoption of market capitalism in the 1980s and 1990s. He noted that at the time, people valued starting businesses more than academic pursuits, but he was more academically inclined.
At the age of 17, he enrolled at Zhejiang University, where he majored in Electronics and Communication Engineering, later pursuing a master’s degree in Information and Communication Engineering, which he completed in 2010.
In 2015, Liang co-founded a quantitative hedge fund that used complex mathematical algorithms for trading rather than human analysis. By the end of 2021, the fund’s portfolio had grown to over 100 billion yuan (US$13.79 billion).
However, in April 2023, the fund announced via its WeChat account that it would broaden its focus beyond the investment sector, directing resources towards “exploring the essence of AGI.” DeepSeek was founded a month later.
DeepSeek has stated that its recent models were built using Nvidia’s lower-performing H800 chips, which are not banned in China, sending a message that the most advanced hardware is not always necessary for cutting-edge AI research.
The company gained attention in the AI industry last month when it released a new AI model, claiming it was on par with similar models from US companies like OpenAI. The model was said to be more cost-effective in its use of Nvidia chips to train on large datasets. The chatbot became more accessible when it was released on Apple and Google app stores earlier this year.
With inputs from agencies