**Wenfeng Li's Youthful Past: $8,000 Capital, a Fiat, and a Solo Long March**

Original Title: "Li Wenfeng's Youth: $80,000 Initial Capital, a Fiat, and a Personal Long March"
Original Author: Beat of Dyna

January 27, 2025, Monday. Li Wenfeng was playing soccer with some middle school classmates on a field in Wuchuan.

Seven days ago, DeepSeek-R1 was released. Seven days later, news crossed the Pacific and triggered turbulence in the U.S. capital markets. NVIDIA's stock price plummeted over 16% in a single day, causing a market cap loss of nearly $600 billion. American media dug up old terms from the Cold War era, dubbing this day the new "Sputnik Moment."

DeepSeek had already topped the free app charts in both China and the U.S. Overnight, people in Silicon Valley learned to pronounce an unfamiliar Chinese name, and interview requests flooded into Hangzhou from around the world as journalists tried to find out where Li Wenfeng had gone.

No one found him.

That afternoon, he was wearing a soccer jersey, running back and forth on a field in a county-level city in western Guangdong. As the ball rolled to his feet, he caught it, turned, and passed it again.While the world outside the field was in uproar for him, the match on the field continued.

The next day was New Year's Eve. Li Wenfeng returned to Mileling Village, where a banner had been hung at the village entrance with red background and white characters that read: "Warmly welcome Wenfeng, the pride and hope of our hometown."

Tourists came one after another, standing beneath the banner to take photos. Some people bent down, picked up a handful of soil from the roadside, and carefully kept it, as if wanting to take some good luck from this village. Li Wenfeng's high school classmate, Mr. Chen, told reporters that he had promised to return to Wuchuan for the New Year, but after coming back, he planned to find a place to hide.

Children of Mileling Village

In 1985, Li Wenfeng was born in Mileling Village, Tanba Town, Wuchuan City.

Wuchuan is located in western Guangdong, surrounded by rivers on three sides, with the south facing the sea. The Jian River flows down from the north and suddenly turns a bend here before flowing south into the sea. In the past, merchant ships plying the river used to anchor in the bay, with warehouses lining the shore and bustling crowds, giving the county town a rather imposing name, "Little Foshan."

Wuchuan also believed in education. Xiejie Village, the hometown of Qing Dynasty Zhuangyuan Lin Zhaotang, is not far from Mileling Village. The county annals record that there was a zhuangyuan, twenty jinshi, and one hundred sixty-five juren from here. For hundreds of years, generation after generation of young people set off from here, following the same narrow road to the county town, provincial capital, and the imperial capital. The glory of official ranks had long become a thing of the past, but the idea that studying could change one's fate had always remained in this land.

The village of Mileling also has its own story. When the elderly in the village mention their ancestors, they will talk about the Anti-Japanese General Liang Huasheng, as well as Liang Tai Xi, who graduated from Sun Yat-sen University in 1936. According to the village's records, after liberation, nearly a hundred students have been admitted to top universities in China from here.

However, Liang Wenfeng grew up in the 1990s. That was the heyday of money in Guangdong, with factories hiring, businesses thriving, and coastal cities rapidly developing. A young person who was willing to endure hardship and take risks might not necessarily need to spend many years in a classroom; they also had the opportunity to earn money beyond what they had ever imagined. In comparison, studying was too slow, requiring year after year of endurance, and no one could guarantee what one would ultimately achieve.

Many parents in the village believed that studying was useless. Some even went to Liang's house specifically to persuade his father not to let his child continue studying. At that time, the centuries-old fame and glory accumulated in Mileling Village were being overshadowed by the fresh and hot wealth stories of Guangdong.

Many years later, Liang Wenfeng mentioned this past in an interview:

「I grew up in a third-tier city in Guangdong in the 1980s. My father was an elementary school teacher. In the 1990s, there were many money-making opportunities in Guangdong. At that time, many parents came to my house, basically thinking that studying was pointless. But looking back now, their mindset has changed. Because money is not easy to come by anymore, and even the opportunity to drive a taxi may be gone. The times have changed for a generation.」

「The times have changed for a generation.」

This statement refers to the 1990s and also to the year 2024. When he said this, he was engaged in another similarly slow and hard-to-explain endeavor.

Liang Wenfeng's father taught at Meicu Primary School and later became the Deputy Headmaster. His mother was also a teacher at Bianling Elementary School. This family had little to do with scientific research, but the attic was always filled with books. According to local media reports, there were over a thousand books at one point.

The father set two rules for the family. At the dinner table, they did not discuss studying but only talked about the family's life; after dinner, each family member would read a book. Liang Wenfeng's academic performance was rarely inquired about. What his father often said was that solving problems was more important than scores.

The most valuable appliance in the Liang household was a Leap brand radio. It had a black case, metal knobs, and an antenna that could be extended section by section. When you turned the knob, a crackling electric sound would first come out of the speaker. After tuning the frequency, voices from distant places would slowly emerge from the noise.

When Liang Wenfeng was in the fourth grade, he disassembled this radio. He unscrewed the screws, opened the case, and exposed the densely packed wires and components on the table. He disassembled and reassembled it, reportedly going back and forth thirty-seven times.

His father never taught him about circuits, nor did he stop him. He simply handed him a screwdriver and said, "Remember to put it back together after disassembling."

Later, when the family got a television, Liang Wenfeng started to dismantle it. One time, he broke the TV. His father didn't scold him or take away his tools. He just looked at the TV that would never light up again and asked him, "How can you make sure you don't break my TV next time?"

His kindergarten was in Meiling Elementary School, and his elementary school was in Meizhi Elementary School. His elementary school teacher, Mr. Li, remembered that he rarely daydreamed in class, especially focused when facing difficult problems, and good at summarizing methods. In sixth grade, he won the national second prize in a math competition. It was also that year that he was admitted to Wuchuan No.1 Middle School.

During his three years in junior high school, his overall ranking was around the top 100 in the grade, not outstanding, but he often ranked first in math competitions. In the 1999 entrance exam for senior high schools, he scored 819 points, just over 20 points higher than the admission line for Wuchuan No.1 Middle School.

In his freshman year of high school, he jumped into the top ten from around 100 in the grade. His senior-year homeroom teacher and math teacher, Mr. Yang Yahong, said that this child was not flashy; his homework was never a concern for teachers. He was particularly interested in the math group, participating in the provincial math competition every year and winning at least third prize. His junior high school teacher, Ms. Rong, also said that he had self-studied high school-level math during junior high school, started studying university-level math, was quiet but not a bookworm, as if he didn't need to spend a lot of time to excel in each subject. During those junior high school years, he became obsessed with computers, disassembling and assembling them, and even helped the school repair computers.

In the summer of 2002, the college entrance exam results were released. Liang Wenfeng scored 806 out of 900, becoming that year's Jhanjiang City college entrance examination champion.

When filling out his college application, he only wrote Zhejiang University, citing that the Information and Electronic Engineering program here was excellent.

When Mr. Yang Yahong asked him if he was absolutely sure, he responded with confidence.

Later reports mentioned that with his score, he could have gone to Tsinghua University.

The One Who Skipped Classes

In the fall of 2002, Liang Wenfeng went to Hangzhou to study Electronic Information Engineering at Zhejiang University.

He rarely attended classes. Many years later, a classmate from the same year but different class came out on Zhihu to reminisce, saying that most of the time Liang Wenfeng was self-taught and felt that the teachers were slow-paced, wasting time. In his sophomore year, while classmates were rushing to finish assignments, he had already designed a circuit board, written a microcontroller program, created an interface, and developed a software similar to a mini player.

He also modified a regular guitar into an electric guitar, the tone of which could be adjusted by connecting it to a computer. After completing it, he didn't find this invention remarkable. He merely thought the guitar's tuning wasn't accurate enough and said that it would be even better if it could tune automatically.

During the summer vacation of his junior year, he formed a team with two classmates to participate in the National College Student Electronics Design Competition. None of the three were among the top students in their major. During the Zhejiang University on-campus training, many design tasks were almost single-handedly completed by Liang Wenfeng. In the end, they won first place in Zhejiang Province and a national first prize.

The awards were not announced until October, causing them to miss the deadline for that year's graduate school recommendation. As a result, Liang Wenfeng was delayed by a year in pursuing his master's degree. During the gap year, he continued to work on electronic sensing systems, reportedly related to marine navigation, with hardware, software, and algorithms all being handled by him alone. A undergraduate classmate later commented that any of the electronic systems he had worked on in college could easily pass as a master's thesis in the Department of Electronics.

He also went on a cycling trip. During his university years, he rode a bicycle through several provinces in East China, traveling during the day and sleeping outdoors at night, completing the journey at a minimal cost. This story was told by his electronic design team member on Zhejiang University's campus forum, with the post's title even misspelling "Liang Wenfeng" as "Liang Wenfeng."

In 2007, he began his Master's studies in Information and Communication Engineering at Zhejiang University, focusing on machine vision, under the guidance of his advisor, Xiang Zhiyu. Xiang Zhiyu completed his bachelor's, master's, and doctoral degrees at Zhejiang University, and later conducted postdoctoral research at the University of Aveiro in Portugal and The Ohio State University in the United States, where he remained in the field of computer vision upon returning to China.

Liang Wenfeng's graduation thesis was titled "Research on Target Tracking Algorithm Based on Low-cost PTZ Camera," submitted in May 2010.

The problem he aimed to solve was to enable a moderately priced, controllable camera to continuously track a moving target in an environment with constant changes in lighting, obstruction, and background. With limited equipment and a complex environment, in the acknowledgments of his thesis, he mentioned that Professor Xiang introduced him to the field of machine vision, prepared study plans for him, and often provided line-by-line guidance in coding.

In 2011, this paper was revised and published in the Journal of Zhejiang University (Engineering Science Edition). The author's brief bio at the end consisted of only one line:

Liang Wenfeng, born in 1985, from Zhanjiang, Guangdong, Master's student, engaged in machine vision research.

The file ends here, neat and clean, like a standard good student.

Looking back many years later, that paper already presented a problem he would repeatedly deal with later on: how to identify the truly important target and persist under limited conditions.

$80,000 Principal

In 2008, during the financial crisis, the A-shares plummeted from a high of 6124 points to 1664 points. The trading hall was a sea of red, with consecutive plunges, leaving many people's accounts with only remnants.

Back when he was a second-year graduate student, Liang Wenfeng saw something amidst the chaos. He believed that in this volatile market, there might be an opportunity. He gathered like-minded classmates to explore using machine learning in trading.

According to later media reports, around 2008, a few students at Zhejiang University's Yuquan campus, who were studying machine vision, started contemplating a "cheat stock trading software." Their initial idea was not grandiose; they just wanted to recover the money they had previously lost. At first, their program was barely breaking even in the stock market. It was only after repeated revisions and optimization of their strategy that they started making money. That thing they called a "cheat" had a more formal name in the financial industry: algorithmic trading.

Back then, market data was not as readily available as it is now. They had to extract numbers from market software interfaces, write their own programs, and find a way to connect to trading software. Liang's skills in working with cameras were put to use in monitoring stock prices. He also collected data from various sources, used his connections to obtain data from financial institutions, then conducted modeling, parameter tuning, and wrote code day and night, often staying up until dawn.

Liang Wenfeng's initial capital was only 80,000 RMB. Today, 80,000 RMB would not even be enough to buy a top-of-the-line graphics card. In 2008, it was his entire net worth. He did not explain this gamble to anyone. At an age when others were busy polishing their resumes and looking for internships, he was glued to his screen every day. He rarely explained to anyone what he was actually doing. At that time, algorithmic trading had not yet become a respectable profession in China.

In 2009, he interned at Shanghai Aiqi Information Technology. The company was founded by Zhejiang University alumni Zhou Chaoen. He was appointed as the manager of the new technology department before graduating, with a monthly salary of 10,000 RMB, responsible for AI video image technology. When he left, Zhou Chaoen gave him advice to seek businesses with high gross margins. Liang Wenfeng remembered this advice for many years, and in 2023, when they met again, he brought it up proactively.

However, Aiqi Information Technology was just a side story in his life. In 2009, while enjoying a high salary, he was also preparing for algorithmic trading. In his own life plan, that was the main storyline.

In June 2010, three classmates in the lab graduated with master's degrees. One joined a large company, one chose to start a business, and another went to Chengdu to continue working on things that others could not understand.

Liang Wenfeng was the one who went to Chengdu.

Rental Room in Chengdu, Fiat Heading to Tibet

In Liang Wenfeng's rental room in Chengdu, which was less than ten square meters, dim and humid, there was only a shabby desk and a single bed. When hungry, he ate instant noodles, and when tired, he rested his head on the table. The room was so small that it couldn't fit anything other than a bed and a desk, but all of one person's future was squeezed into this room.

A top-tier graduate, bypassing tech giants and government institutions, chose to face the screen every day. In the context of 2010, this behavior was considered the opposite of the norm. No one knew what he was doing in front of the screen. Reports in the shadows stated that after graduation, he "hid in a low-cost rental in Chengdu, constantly experiencing failures in various scenarios."

On April 16, 2010, the CSI 300 Stock Index Futures were listed, with the main contract opening at 3450 points. For the first time in the Chinese capital market, mature short-selling and hedging tools were introduced, finally opening a door for quantitative trading.

And it was through this door that Liang Wenfeng entered.

According to later reports, in those two years, he spent most of his time in a rental in Chengdu, repeatedly testing his ideas. The prices seemed chaotic on the surface, but there might be underlying patterns. With enough data and an accurate model, perhaps a repeatable explanation could be found within those fluctuations.

This was also the belief passed down by Simmons to future generations. Simmons was originally a mathematician, later founding Renaissance Technologies, engaging in trading using mathematical models and computer programs. Its flagship Medallion Fund has since become one of the most famous funds in the history of quantitative investment. He believed that although market prices seemed disorderly, there were still identifiable and calculable patterns behind them. There must be a way to model prices; the challenge lies in whether that method can be discovered.

During the same period, Liang Wenfeng was surrounded by a group of equally restless young people. Some holed up in Shenzhen's urban villages, researching seemingly unreliable aircraft, and they had invited him to join them. Liang Wenfeng did not go. Later, that group created DJI.

In 2011, he had no idea about his own and his friends' future. In that year, he only had one model that hadn't quite worked out, a rented small house, and a car.

In October of that year, Liang Wenfeng was 26. It was the year when Weibo had just started to become popular, where people shared their lives and also their loneliness. In that autumn, he drove a Fiat to Tibet.

It wasn't a car built for expeditions. Fiat is an old Italian brand of cars, which used to be inexpensive in China, where you could buy one for fifty to sixty thousand yuan. It was a common sight on the streets. The brand later exited the Chinese market, and those cars slowly disappeared from the roads.

Liang Wenfeng called it his "steed."

By the time he set off, the car was falling apart.

All along the way, he was posting on Weibo. If you laid out all those Weibo posts one by one, it would be a diary in pages.

“October 19, Lhasa. Back in Lhasa. Touched~”

Returning to Lhasa is also the name of a song by Zheng Jun. Those who have driven to Tibet before probably had this melody circling in their minds.

“October 24, Yamdrok Lake. Ancient road, western wind, thin horses, sun setting in the west.”

Yamdrok Lake is located at an altitude of 4,441 meters, with a lake area of over 600 square kilometers and a lakeshore of over 200 kilometers. He filmed the lake, the snow-capped mountains, and that Fiat car.

“October 25. Overlooking the mountains.”

“October 26, beside the Yarlung Tsangpo River. Riding across the world.”

“October 27, Himalayas. The Fiat came for a visit. Tibet, Himalayas, on the other side of the snow-capped mountains is India.”

“October 28, by the lake. That lake is actually several kilometers wide, and the snow-capped mountains are dozens of kilometers away.”

Li Wenfeng continued westward.

He passed by Yongbulakang. The palace, situated on the mountaintop, is the oldest palace in Tibetan history, even older than the Potala Palace. He then headed north along the northern foot of the Himalayas towards Mount Everest, capturing photos of the snow-capped mountains to the north and west, passing through the Qomolangma Nature Reserve, and entering Ngari.

With an average altitude of over 4,500 meters, Ngari is called the “Roof of the Roof of the World.” The air there is only about half as thin as that on the plains. The road traverses the wilderness, often hundreds of kilometers without seeing any signs of human habitation.

Li Wenfeng drove the Fiat into it.

Then, he disappeared from Weibo for a week.

When he reappeared, he started posting the photos he took along the way. Two hundred kilometers into the uninhabited area, he photographed Tibetan antelopes; at four hundred kilometers, an eagle soaring above the snow. Beneath the photos, he wrote:

“The wind whispers, the river is cold~”

Throughout the journey, he still spoke for himself with those old verses, until November 7, 2011, when he posted his final Weibo:

“Been stranded in the uninhabited area for a week. Fortunately, I made it out. But my Fiat did not. It will forever remain in that boundless alpine meadow. That’s the END.”

After that, Li Wenfeng never updated his Weibo again. On August 19, 2014, his account was hacked. The young man who once took photos of lakes and snow-capped mountains, and would chat and joke with strangers in the comments section, slowly disappeared from the internet. His later life began on the day he lost his car.

The Final Decision

After the trip to Tibet, Liang Wenfeng stopped updating his Weibo account but did not disappear completely from the timelines of those Zhejiang University friends.

Over the years, they would still interact in the comments section, challenging each other with math problems, discussing games, and inquiring about each other's final decisions. Xu Jin and Zheng Dawei were the two names that appeared most frequently. Years later, they would co-found Jacobie with Liang Wenfeng, becoming the core team of the Magic Square. However, in their 2012 Weibo posts, they were just a few young people standing at the crossroads of life.

On August 23, 2012, Zheng Dawei posted a question on Weibo: an ant can only crawl along the edges of a cube, starting from one vertex and moving to the vertex opposite to it. At each vertex, it randomly chooses the next edge to traverse. What is the mathematical expectation of the final distance traveled?

Zheng Dawei tagged Liang Wenfeng and Xu Jin. Liang Wenfeng replied, "It seems like neither Xu Jin nor I have solved this question, but it definitely has to be greater than 4."

Later, many people recalculated this question, and the answer was 10. He did not provide a proof at that time, only judged that the answer must be greater than 4. The direction was correct.

Underneath this question, he asked Zheng Dawei, "

What is the final decision, to go or stay?"

This sentence had no context or further explanation. Looking back, it might have been just a casual question among friends. Years later, it resembled an early invitation, an unsolved math problem, a group of young people still choosing their paths, and a team without a name yet.

Zheng Dawei later joined the Magic Square and became a core member of DeepSeek. Xu Jin also became part of Liang Wenfeng's entrepreneurial story thereafter.

Zheng Dawei left behind more than four thousand Weibo posts. Reading through them one by one, you can see a young man's journey from student life to society, and how that generation of tech youth once discussed their lives.

He mentioned going through four direct admissions, saving four years and not wasting time on exam preparation. He depicted life as a triangle, with the vertices representing passion, effort, and play. He wanted to create "products that move people" and hoped to drive new technology so that his parents could one day benefit from technological advancements.

Amidst these ideals were a lot of younger, more raw elements. He played Dota and wrote, "Better to have one less person than a pig-like teammate," and also retweeted a widely circulated saying among programmers, implying that our generation's brightest minds are figuring out how to make people click on ads one more time.

Sublimity and jokes appeared side by side on the same timeline. People at that time did not see any contradiction in this. They discussed mathematics and games; they wanted to change the world, but also got angry at a bad teammate in a game. Weibo had not yet become a carefully curated personal showcase, and the words spoken by a person often remained there in such a messy manner.

Xu Jin had almost cleared his Weibo posts. Among the remaining posts, someone asked about being admitted to a university without taking the entrance exams, and he answered very directly: "Because I didn't have to study for the senior year of high school."

His resume was then brought up again by people: Zhejiang University's Zhukezhen College mixed class, doctoral research on robot navigation, and involvement in the "Yutu" lunar rover's visual navigation project. The person who complained on Weibo about the senior year later actually sent a machine to a place where there was no road.

This youthful past does not only belong to Liang Wenfeng. Some of the people who later formed the Magic Square and entered DeepSeek had already met in the same loose network of relationships at that time. Their obsession with technology, impatience with inefficient rules, and almost harsh appreciation of intelligent companions appeared before the company.

The Magic Square did not have a name yet; its original character had already formed during the Weibo era.

In 2013, Liang Wenfeng returned to Hangzhou from Chengdu. He, Xu Jin, and Zheng Dawei decided to continue their "cheating in stock trading" business, starting a business together, with the reason being "earning more money than having a job."

On September 10, 2013, Hangzhou Yaqubi Investment Management Co., Ltd. was registered with a registered capital of 100,000 yuan. Liang Wenfeng held 50.5% of the shares, while Xu Jin, Yao Fengfeng, and Zheng Dawei each held 16.5%. The company was named after the German mathematician Carl Gustav Jacob Jacobi.

Two years later, the market opened another door for them.

In April 2015, the CSI 500 Stock Index Futures were listed, providing quantitative traders with a new hedging tool. Two months later, Hangzhou Magic Square Technology was established. The name was derived from the ancient Chinese Luoshu Nine-Square Diagram, a matrix where the sum is equal both horizontally, vertically, and diagonally. That same year, they moved into the Huijin International Building on Hangzhou Huancheng North Road, which was one of the few newly built high-end office buildings in Hangzhou at the time, with expensive rent, mostly occupied by private equity firms.

In December 2015, there was a recruitment notice on Shuimu Community. The notice did not directly mention Liang Wenfeng's name but told the story of a certain "Mr. L" in the third person:

In 2008, Mr. L started his independent quantitative trading journey with 80,000 yuan. In 2015, after experiencing the seven-year cycle of bear and bull markets, Mr. L, with an annual compound return rate of over 100%, joined the ranks of millionaires.

The notice also stated that China's quantitative trading is about to bid farewell to the era of solo cowboys and enter the era of geek aggregation in private equity funds. Mr. L's dream is to one day sit at the same table with Simmons, the founder of Renaissance Technology.

A person who rarely talks about himself, for the first time, told a stranger about his past, but hid himself behind a single letter. He described himself in the third person, in a calm tone as if organizing another person's resume.

Eighty thousand yuan, seven years, surpassing one billion.

This is his most complete graduation certificate from his youth, and also the first self-confession to the world.

The Tibetan Weibo diary is written to himself, and the recruitment notice on the Water and Wood Community is written to future companions. Between these two texts are four years apart, as well as a Fiat that never left the Ali no-man's zone.

Firefly

On October 21, 2016, the first stock position generated by a deep learning model at Magic Square officially went live in actual trading. It was also from this day that GPUs entered Magic Square's trading system.

Eight years ago, Liang Wenfeng had to capture data from the market software interface, find the API by himself, and modify the rules one by one. Now, what stocks to buy, what to sell, and how much position to take are calculated by the model. People have stepped back to a farther place, responsible for preparing data, designing systems, and then waiting for the machine to provide answers.

In 2019, Liang Wenfeng explained the difference between the two at the Golden Bull Award ceremony. When a person makes investment decisions, they rely on experience and intuition, which is closer to an art form; when a program makes decisions, it faces a problem that can be solved, and there should be a better answer behind it.

From 2008 to 2016, he spent eight years trying to remove himself from every specific judgment. The model does not need excitement, nor panic, and it will not change its mind due to losses the previous day. It only receives data, calculates probabilities, and then executes.

Quantitative practitioners often start with believing in their judgments but end up establishing a system that no longer requires their judgment. Liang Wenfeng has reached a point where he can manage a larger fund and continue to refine this machine more precisely.

But he did not stop at the fund size.

To make the system smarter, it needs more data, more complex models, and more computing power. Magic Square's GPU count has thus increased along a steep curve. Liang Wenfeng later recalled that there was originally only one card, increasing to a hundred in 2015, reaching a thousand in 2019, and then moving towards ten thousand.

In 2019, Fabletune invested nearly 200 million RMB and built "Firefly-1." This computing cluster has about 1,100 GPUs, covering an area close to a basketball court. Two years later, "Firefly-2" was completed, with an investment of 1 billion RMB, housing around 10,000 NVIDIA A100s. The area of just one data center is close to ten basketball courts.

After making money from quantitative hedge funds, they usually expand their scale, recruit traders, or explore more strategies. Liang Wenfeng converted a large part of the money into graphics cards, server racks, and continuously roaring computing devices. At that time, China's large-scale modeling industry had not truly taken off, and not many people knew what such massive computing power could ultimately be used for.

Someone asked him why he needed to buy so many graphics cards.

Liang Wenfeng replied, "An exciting thing may not be measurable only by funds. It's like buying a piano at home, where one, you can afford it, and two, there are indeed people who want to play it."

In 2018, a financial media outlet visited Fabletune. Liang Wenfeng, holding a thermos cup, wearing a dark blue cotton padded work shirt, looking slim, and appearing a bit reserved, seemed like an engineer who came from the 1990s.

The interview was initially scheduled for half an hour but ended up lasting over two hours. Whenever asked about technology, he almost always had an answer. When encountering areas that were hard to describe accurately, he would pause to think for a while and smile shyly.

The reporter asked, "How does Fabletune evaluate its employees?"

Liang Wenfeng said, "The company does not have specific evaluation metrics. If a person has not made contributions for a long time, the first thing to reflect on is whether the company has placed him in the right position."

The reporter then asked, "What kind of company does Fabletune ultimately want to become?"

After pondering for a while, he said, "I hope that one day, we can operate without performance incentives and management fees." Pausing for a moment, he added, "What I really want to do is create an open-source strategy platform that allows ordinary investors to use."

At that time, Fabletune was still a quantitative hedge fund, and when words like open-source and inclusive came out of his mouth, it seemed somewhat distant from business. However, Liang Wenfeng has never seemed entirely satisfied with just turning technology into an advantage for a few. After making money, he was already thinking about how to dismantle that set of capabilities and make it accessible to more people.

Many of the things he later did at DeepSeek had already hinted at in this interview. Beside the thermos cup in 2018, the engineer who was not good at small talk had already articulated these ideas.

In August 2021, Fabletune's AUM exceeded 100 billion RMB, and it was crowned one of the "Quant Four Heavenly Kings" alongside Nine Realms, Bright Vast, and Soul Equilibrium.

Four months later, the retracement came.

On December 28, 2021, Fantasia issued a public statement acknowledging the company's largest performance retracement to date and expressing deep remorse for it. Subsequently, Fantasia closed its fundraising channel and proactively reduced its management scale. The rapid expansion seen in the past few years in the digital realm began to recede. The once high-speed money machine showed signs of stalling for the first time.

Legends are not always on an upward trajectory.

During Fantasia's years of downsizing, as funds ebbed away, external applause gradually diminished. At the same time, the purchased graphics cards, established data centers, and assembled young engineers were being diverted to an unproven endeavor.

The years of Fantasia's contraction happened to coincide with the years when DeepSeek was conceived. This fact went unnoticed at the time, including by the founder himself.

Epilogue

In January 2023, Fantasia announced its donations for the year 2022. The company donated 220 million yuan, with one employee, signed as "An Ordinary Little Pig," making a personal donation of 138 million yuan. Speculations abound, suggesting that the little pig is none other than Liang Wenfeng.

Three months later, Fantasia announced its entry into the large-scale model field. The poster quoted Truffaut's words to a young director:

"Be sure to embrace ambition madly, and also to be madly sincere."

At that time, the domestic large-scale model entrepreneurship scene was already saturated with internet giants, celebrity entrepreneurs, and capital. Fantasia's abrupt pivot from the quant market seemed somewhat untimely. It had no cloud business, no search entry point, and no massive user base. What it firmly held were only computing power, engineers, and money earned from the quant market.

In July 2023, Hangzhou DeepSeek AI Basic Technology Research Co., Ltd. was established. The team consisted of about 140 people, many of whom were fresh graduates from Tsinghua University and Peking University, with an average age of less than thirty.

Someone asked Liang Wenfeng why he didn't compete for those who were already famous and experienced.

He said that the criteria for choosing people had always been passion and curiosity. The world's top fifty talents might not currently be in China, but China could cultivate such talents on its own.

In May 2024, DeepSeek-V2 was released, inputting 1 yuan for every million tokens and outputting 2 yuan, pushing the price war of domestic large models into a new phase. Liang Wenfeng later described that they did not intend to be a market disruptor, but inadvertently became one.

But the low price is just a result.

When asked why he always had to choose the most challenging path, An Yong replied that over the past few decades, the most familiar route for Chinese technology companies was to wait for the United States to carry out basic innovations, bring back the technology, develop products, expand the market, and scale up. This path was fast and indeed made money.

Li Wenfeng said that with economic development, China must gradually transition from being a technology beneficiary to a contributor and not always rely on others' achievements.

In the past thirty years of the IT revolution, China has hardly participated in core technological innovation. Faster chips, better software, updated architectures—these would always arrive here eighteen months later. Moore's Law was like a monsoon arriving on time. People got used to standing on the shore waiting and gradually forgot where the wind came from.

Li Wenfeng no longer wanted to wait.

Now, it is China's turn to contribute to the world.

On December 26 of the same year, DeepSeek-V3 was released. The official disclosure stated that the final training phase used 2048 NVIDIA H800s, with a computing cost of approximately $5.576 million.

Low cost was his theme. Starting with a capital of 80,000 yuan, an inexpensive PTZ camera, a Fiat worth fifty to sixty thousand yuan, and then a training bill of five million five hundred seventy-six thousand dollars. It seemed that he always found a way in limited conditions. When resources were insufficient, could the method be more precise? When the equipment was not expensive enough, could the system be smarter?

On January 20, 2025, DeepSeek-R1 was released, and its inference capability entered the ranks of the world's top models, open under the MIT license. On the same day, Premier Li Keqiang of the State Council presided over a symposium of representatives from various fields, including experts, entrepreneurs, and education, science, culture, health, and sports. Li Wenfeng was present and spoke about the development of domestic large models.

A week later, DeepSeek caused NVIDIA's market value to evaporate by nearly $600 billion in a single day. That Spring Festival, he still returned to Wuchuan. On the morning of the first day of the lunar new year, he left the village.

On February 17, 2025, a private enterprise symposium was held, with Li Wenfeng sitting in the front row. According to media reports, he did not speak at the meeting. After that, he rarely appeared in public for quite some time. While the world continued to discuss DeepSeek, he once again withdrew from the crowd.

There is a rumor that Li Wenfeng's WeChat signature is "Mathematize the World." This cannot be confirmed, but it sounds like him. Starting from dismantling that Leap brand radio, he has been doing the same thing all his life: opening the shell, identifying the internal structure, and finding the operational rules.

At the age of 44, Simmons founded Renaissance Technologies. By the time he was 40, Liang Wenfeng had spent sixteen years in the quantitative market. Both believed that behind the chaotic prices lay a calculable order. However, as time went on, what Liang Wenfeng attempted to model was no longer just prices.

On November 7, 2011, he wrote on Weibo that the Fiat was forever left on Alibaba's boundless alpine meadow. That was his final diary entry for the internet. After that, he rarely talked about his life.

In Liang Wenfeng's final Weibo post for the internet, the Fiat was forever parked on Alibaba's alpine meadow.

In the years that followed, the wind blew across the wilderness, and the grass fell season after season. People didn't know where the car went afterwards. For those who had read that Weibo post, it seemed to have always stayed there, with its metal slowly rusting away, eventually becoming part of a no man's land.

People walked away from there. Those who walked away never had to look back.

Original Article Link