appConfigEntityId

bodyType

customContents

entityId

entityUuid

isVideo

articlePageQuery

article

articleSchemaContent

content

hooksArticlePageAdvertisementConfigArticle

hooksArticlePageFloatingButtonConfigArticle

appConfig

filter

json

liveArticle

Coco Feng

In an era dominated by aggressive tech founders who chase billion-dollar valuations and maximal profits while curating loud public profiles, Liang Wenfeng stands out for his insistence on staying in the background.
With only a couple of photographs of him circulating online, the founder of Chinese artificial intelligence start-up DeepSeek and quantitative hedge fund High-Flyer Quant has long been a reclusive figure.
But last week, a leaked transcript of a closed-door meeting with potential investors in May offered an unprecedented look into Liang’s strategy: one rooted in a philosophy of corporate restraint, low-margin pricing and open-source models.
On mainland Chinese social media, transcripts of the nearly four-hour discussion went viral, resonating deeply in a culture that traditionally values humility, discipline and long-term ambition over fast commercial cashouts.
While DeepSeek did not officially confirm the authenticity of the document, the leaked remarks gave the industry its clearest glimpse yet into the mind driving one of China’s most intriguing AI contenders.
Calculated, long-term sacrifice
Throughout the meeting, Liang repeatedly emphasised the concept of “restraint” as a core company vision, a principle reflected in DeepSeek’s open-source model and affordable pricing.
“We set our prices to earn only a reasonable profit, not to maximise income,” Liang said in the meeting, according to the transcripts.
He added that keeping costs low for users increased the chance of achieving artificial general intelligence (AGI) – the breakthrough where AI systems match or surpass human cognitive abilities.
I have no doubt that AGI will have tremendous commercial value. With that in mind, my priority is not to gain more market share
Liang Wenfeng, DeepSeek
While many global tech ventures prioritise immediate profits and rapid market-share capture, Liang framed his low-margin approach as a calculated, long-term sacrifice in service of a greater objective.
“I have no doubt that AGI will have tremendous commercial value,” Liang said. “With that in mind, my priority is not to gain more market share [now], but to increase the probability of achieving [AGI].”
Ballooning valuation
Liang’s low profile contrasts sharply with his burgeoning personal fortune.
According to the Bloomberg Billionaires Index, Liang’s net worth stands at nearly US$38 billion, driven largely by his controlling stake in DeepSeek and his hedge fund empire. The estimate cements his status as one of the wealthiest founders among pure-play AI model developers globally.
Despite Liang’s admission that DeepSeek remains far from fully commercialised, investor demand during the May funding round – the firm’s first external capital raise – was intense.

DeepSeek was originally spun off from High-Flyer Quant in 2023 to pursue AGI research. While the start-up was not in urgent need of capital, the funding round was intended to establish a market benchmark for employee stock options and maintain team stability, according to previous reports by the South China Morning Post.
The company’s valuation climbed from an initial US$10 billion to about US$59 billion following the round’s conclusion in June. To ensure he retains strategic direction after bringing in outside investors, Liang personally invested nearly US$3 billion, becoming the largest backer in the round.
But it has paused the second funding round, aiming to raise at least 10 billion yuan (US$1.48 billion), after the leaked meeting records, Bloomberg reported on Saturday.
Contrasting cultures
Industry analysts said that Liang’s remarks highlighted a distinct cultural divide between Chinese AI players and their Western counterparts.
Compared with leading American AI labs like Anthropic and OpenAI, DeepSeek’s “approach and corporate culture differ significantly” because of “the founders’ distinctly different core philosophies”, said Guo Tao, a member of the Chinese Association for Artificial Intelligence.
Guo said that while Anthropic and OpenAI focused primarily on commercial implementation and ecosystem expansion, Liang “does not seek short-term commercial gains but focuses solely on ultimate technological breakthroughs, thereby fostering a unique, long-term, technology-driven corporate culture”.
I prefer the narrative that ‘a group of ordinary people did extraordinary things’ rather than ‘geniuses did extraordinary things’
Liang Wenfeng, DeepSeek
Qian Wenying, dean of the Magnet Institute of AI and Business Innovation in Shanghai, echoed this view, noting that OpenAI “believes in large-scale deployment and market feedback”, while Anthropic “places greater emphasis on institutionalised safety constraints”. DeepSeek, by comparison, “chases technical efficiency, engineer autonomy, and an open ecosystem”, Qian said.
Addressing the broader geopolitical landscape, Liang told investors that the gap between China and the US did not lie in talent, but in computing resources.
Chinese companies have been barred by the US government from accessing the most advanced chips from Nvidia, the main workhorses of AI training and inference.
Liang said it was necessary for DeepSeek to build its own computing clusters, but noted that the company had not decided if it needed to develop proprietary chips.
Closing the meeting, Liang described DeepSeek as “a group of very ordinary people”, challenging the notion of elite exceptionalism that often characterises Silicon Valley narratives.
“I prefer the narrative that ‘a group of ordinary people did extraordinary things’ rather than ‘geniuses did extraordinary things’,” he said.


Top Chinese AI models make 10% of OpenAI, Anthropic revenue despite high valuations: report

China to see major shift to Huawei for AI model training in 2027: deputy chair

People’s Daily rejects US claims of malicious AI distillation, warns of countermeasures

DeepSeek AI engineer slams Anthropic, OpenAI over ‘pacing’ calls, invokes Nazi Germany

DeepSeek

Low profile, high AI ambition: what leaked comments reveal about DeepSeek’s Liang Wenfeng

DeepSeek founder Liang Wenfeng says his company sets prices to earn only a reasonable profit. Photo: Weibo

Liang Wenfeng of DeepSeek meets President Xi Jinping at a symposium of tech entrepreneurs on February 17, 2025, in Beijing. Photo: CCTV

How DeepSeek’s landmark funding secures Liang Wenfeng’s grip as China’s AI rivalry heats up

Liang’s decision to lead the round would ensure he still steers the direction of the firm after bringing in external investors. Photo: Shutterstock Images

DeepSeek has raised around 50 billion yuan in a Series A round. Photo: Reuters

Contemporary Amperex Technology Limited (CATL) committed 5 billion yuan to the round. Photo: Eugene Lee

From gifted student to AI hero, DeepSeek’s founder Liang Wenfeng inspires a nation

Illustration: Henry Wong

A stall at Mililing village selling zhuangyuan sugar cane juice. Photo: Coco Feng

The class of 1999 at the junior division of Wuchuan No 1 Middle School. Liang Wenfeng is in the top row, sixth from the right. Photo: Coco Feng

Flowers bloom at the campus of Zhejiang University in Hangzhou. Photo: Xinhua

Visitors take photos outside the house of Liang Wenfeng’s grandparents. Photo: Coco Feng

Liang Wenfeng shakes hands with Chinese President Xi Jinping during a symposium in February. Photo: CCTV

A street sign reads, “I seek deeply in Mililing”, a play on the AI start-up’s name. Photo: Coco Feng

Tech

Tech Trends

Business

China Business

Companies

News

China Future Tech

Tech founders are not known for ‘restraint’. But a transcript shows Liang Wenfeng is chasing an ultimate goal in AI, not short-term gains


Tech founders are not known for ‘restraint’. But a transcript shows Liang Wenfeng is chasing an ultimate goal in AI, not short-term gains.


Low profile, high AI ambition: what leaked comments reveal about DeepSeek’s Liang Wenfeng

.css-1c6uqr6{color:inherit;font-weight:inherit;font-size:inherit;font-family:inherit;line-height:inherit;overflow-wrap:break-word;}Tech founders are not known for ‘restraint’. But a transcript shows Liang Wenfeng is chasing an ultimate goal in AI, not short-term gains

Tech founders are not known for ‘restraint’. But a transcript shows Liang Wenfeng is chasing an ultimate goal in AI, not short-term gains