Throughout the U.S. tech community this past week, there has been a collective sense of regret over Kimi K3's decision. Why didn't such an outstanding young talent like Yang Zhilin stay in the U.S. back then?
This topic garnered over five million views on X. Former White House AI leader under Trump, David Sacks, a close confidant of Trump himself, mentioned that he has now switched a large part of his work from Claude to Kimi, stating, "It's much more enjoyable because it gets the job done directly without lecturing you." Meanwhile, Silicon Valley veteran investor Vinod Khosla attributed this to immigration policies, stating that the U.S. is driving outstanding talent away.
As speculation grows, Yang Zhilin's Ph.D. advisor, Salakhutdinov, has been publicly showcasing his protégé on social media while also clarifying the reason Yang Zhilin didn't stay in the U.S.: "If he didn't even have the courage to attempt entrepreneurship, Yang Zhilin said he would regret it for a lifetime."
On the other side of the ocean, there is another name from Guangdong who is well-known in the Chinese AI community, Liang Wenfeng.
Seven years older than Yang Zhilin, Liang Wenfeng was born in Wuchuan, Zhanjiang. A programmer who has been in quantitative trading for fifteen years, he has scarcely been interviewed, has no social media accounts, and colleagues describe his only personality trait as "having no hobbies other than programming." After the release of his DeepSeek R1 in January 2025, NVIDIA's market capitalization evaporated by nearly $600 billion in a single day, which Silicon Valley dubbed the Sputnik moment. While the whole world was looking for him, he retreated to his hometown to play football for a few days.
Two individuals from Guangdong, one born in Wuchuan and the other in Shantou, separated by the Leizhou Peninsula. Their companies are on the same track, following nearly identical trajectories, yet each branching step actually stems from their true characters.
A Radio and a Band
Liang Wenfeng was born in Mililing Village, Tanba Town in 1985. Both his parents were teachers at the local primary school. With few toys at home, the most important item during his childhood was a Leap brand radio, which he dismantled and reassembled countless times.
This quiet child showed his uniqueness at an early age. His junior high school homeroom teacher recalls that he was not a bookworm, nor did he necessarily study harder than others, but he had already self-taught himself high school-level math in junior high and started reading university textbooks, as if he didn't need to spend a lot of time to excel in each subject.
During the 2002 National College Entrance Examination in China, he scored 806 points, becoming the top scorer in Zhanjiang City. A photo of him receiving the award can still be found today: wearing a maroon short-sleeved shirt with a large red flower pinned to the chest, his expression stiff, clearly pushed onto the stage by a teacher. That autumn, he entered Zhejiang University to study Electronic Information Engineering. However, over the next twenty years, this individual never appeared in any photo-worthy occasions again.
Seven years later in Shantou, another young person from Guangdong was growing up. Yang Zhilin, born in 1992, graduated from Tsinghua University's Computer Science program with the highest GPA for four years and published over twenty papers. Achieving such academic success at Tsinghua was not particularly surprising, but what was truly remarkable was that, in addition to his academic achievements, he formed a rock band called Splay, named after a data structure, and served as the drummer.
He later explained: "At that time, I felt there were many things I wanted to express, including the pressure from reality and the absurdity of the environment." They wrote a song about a daydream of overnight entrepreneurial success and sudden wealth, "half out of empathy, and half as a reminder not to become too utilitarian."
Many years later, this band would unexpectedly reappear in the story in a way no one could have imagined: the bandmate, Zhou Xinyu, became a co-founder of the Dark Side of the Moon.
Outside the office of the Dark Side of the Moon, there is a white Yamaha electric piano with Pink Floyd's 1973 album "The Dark Side of the Moon" on top. The company drew its name from this album.
One dismantled a radio because something needed to be seen, while the other played the drums because there was something to express. Almost every decision these two individuals made in the following twenty years stemmed from this.
Rental Housing in Chengdu and Corridors of CMU
After graduating from Zhejiang University, Liang Wenfeng did not join a tech giant to receive a prestigious technical business card. Instead, he ran to Chengdu, hiding in a cheap rental house to experiment with various algorithms, attempting to integrate AI into traditional industries, but all in vain. The founding teams of China's top quant funds mostly have a gilded resume from overseas hedge funds, while Liang Wenfeng carved his own path in a rental house.
In 2015, he co-founded Magic Square with his classmates from Zhejiang University. What followed was a series of actions that were almost incomprehensible at the time: in 2019, he spent nearly $200 million to build a self-owned cluster with 1,100 GPUs; in 2021, he added another $1 billion investment to hoard about ten thousand A100 cards.
For quant trading, you don't need that many cards. Liang Wenfeng himself admitted that even for trading, a small number of cards would suffice. Those who had interacted with him in the early years recalled that when they saw him hoarding cards to train models, they only thought of him as a tech enthusiast with a terrible haircut burning money.
However, what he did was exactly the opposite of what everyone else understood. He wasn't using AI to reduce costs and increase efficiency in finance; instead, he was using finance to provide funding for AI research. The typical sequence for most Chinese AI companies is to first secure funding, then develop a product, and finally establish cash flow. But he reversed the order entirely: first, he created a cash flow machine, and then used it to acquire the freedom to conduct research.
Yang Zhilin chose a different path, one filled with flowers and thunder.
After graduating from Tsinghua University, he went to CMU (Carnegie Mellon University) for his Ph.D. During this time, he conducted research at both Google and Meta's AI labs. In 2017, he invested all his efforts in language models, later stating that it was the "only important problem." During his Ph.D., he published two papers: one teaching AI to remember longer contexts and another that outperformed Google's strongest model in 20 tests, with a combined citation count of nearly twenty thousand.
Years later, Kimi became well-known for his ability to input long texts. Many people thought it was a differentiated selling point concocted on the spot in 2023, but it was actually a direction he had decided on during his Ph.D.
In 2016, while still pursuing his Ph.D., he co-founded Recurrent Intelligence, focusing on sales call analysis. Sequoia Capital and GGV Capital were both shareholders. This entrepreneurial experience allowed him to witness the roughness of technology implementation early on, but it also planted a landmine under his feet that would only detonate eight years later.
In 2019, after completing his Ph.D., his mentor arranged for him to be introduced to a senior Apple executive to whom Yang Zhilin could report. It was inquired whether Yang Zhilin was willing to join Apple, possibly even in Apple's China office. However, Yang Zhilin declined all emails from Apple and offers from Silicon Valley, choosing to return to China.
At that time, while Liang Wenfeng was hoarding cards in Hangzhou, Yang Zhilin was waiting for an opportunity in Beijing. Both of them were unaware of each other's existence, and they had no idea that these two names would represent China's AI industry today.
A Month's Window and a Catfish
On November 30, 2022, ChatGPT was launched, causing a collective insomnia among the tech circles in Silicon Valley. Yang Zhilin recalled that many of his friends were anxious, experiencing FOMO, unable to sleep, and a lot of them turned to entrepreneurship.
“We started the first round of fundraising in February 2023. Delaying it to April was basically impossible. But doing it in December 2022 or January 2023 was also not feasible as there was an epidemic, and everyone was not responsive at that time.” Yang Zhilin seized that one month, and he did not rest a day.
In March 2023, the Dark Side of the Moon was established, co-founded by Tsinghua alumni Zhou Xinyu and Wu Yuxin. It is said to have started with a $60 million fundraising, gathering about 40 AI researchers in three months. Then came the steepest curve in the history of large-scale model financing in China: Sequoia Capital and ZhenFund joined, with Alibaba leading a $1 billion investment, followed by Tencent, Meituan, and Xiaohongshu, pushing it all the way to $3.3 billion. Kimi, relying on a 200,000-word long text, became the first widely used large-scale model product by many Chinese.
During that time, Yang Zhilin lived in a dual state. Externally, he presented the grandest narrative, estimating the probability of the scaling law not working almost close to zero, likening entrepreneurship to driving towards a vast snow-capped mountain, and referring to the first year as having built a rocket prototype and discovered a bit of fuel formula. Internally, he had to focus on the most trivial algorithms: when computing power was tight, one machine was at 260 today, 340 tomorrow, dropping back in a few days, deciding whether to buy or rent, which channel to go through, negotiating and adjusting daily. Half scientist's conviction, half small business owner's shrewdness, all embodied in this 31-year-old.
But everything given by the capital had long been priced. To sustain the growth curve, in October 2024, Kimi injected $220 million in a single month, followed by another $200 million in November, burning more than the entire third quarter in two months. The drummer who wrote songs mocking overnight riches became the most aggressive founder in the industry. It wasn't that he changed; it was the $3.3 billion valuation making decisions for him.
Liang Wenfeng entered the scene in Hangzhou in an almost deliberately antagonistic manner.
In 2023, DeepSeek spun off from Huansquare without any external investment. The team had less than 140 people, almost no returnees, all fresh graduates from domestic universities and young people who had graduated for a few years. There were no KPIs, no levels, ideas could be implemented directly.
Former employees recalled to The Washington Post that Liang Wenfeng would delve into the details of training strategies, read papers and write code with researchers, “completely unlike a boss, more like a geek.” He explained why he recruited fresh graduates instead of poaching industry veterans: “Experienced people will tell you how to do it without much thought, but inexperienced people will explore repeatedly.”
In May 2024, DeepSeek-V2 pushed API prices down to one yuan per million tokens, forcing Byte, Alibaba, Baidu, and Tencent to follow suit. The entire industry thought it was a long-planned business war, but his response was: “We didn’t intend to be a catfish; we just accidentally became one.”
Pricing is just slightly above cost, "not losing money, nor making a killing." While internet people talk about market share, entry points, and network effects in a price war, he talks about cost accounting. However, this emotionless price reduction is the most lethal, dragging the large-scale API directly from the high-margin narrative into the pricing logic of infrastructure.
Yang Zhilin and Liang Wenfeng, two completely different business models, have been compared by the outside world for some unknown reason.
The Suspended Yang Zhilin
The landmine planted by Yang Zhilin eight years ago exploded in November 2024.
In Yang Zhilin's previous entrepreneurial project, the five old shareholders of Recycle Intelligence initiated arbitration in Hong Kong, China, accusing him of starting the financing for a new company before obtaining full shareholder exemption.
On December 5, Zhu Xiaohu launched an attack on his WeChat Moments, focusing on Zhang Yutong: the former Jinshan River partner received an initial 14% of 9 million shares for free on the dark side of the moon, exceeding the 9.5% allocated to Recycle Intelligence as the "parent entity." Zhu Xiaohu's solution was almost humiliating: apologize, return shares, or cut ties with Zhang Yutong.
At 9:40 p.m. on December 6, Yang Zhilin posted a 1300-character long article. Instead of cutting ties, he stood his ground: Zhang Yutong is a co-founder, and her shares are the consideration for years of future work, the procedure for leaving Recycle obtained signatures from every director. Those close to the company conveyed the internal attitude: She and the dark side of the moon are already integrated and cannot be separated.
Zhu Xiaohu openly expressed complete incomprehension. In a purely commercial coordinate system, indeed, there is no solution; cutting ties is the only rational option. However, Yang Zhilin had other considerations in the coordinate system for decision-making. Salakhutdinov's later clarification provided a footnote: this is the kind of person who "would regret for a lifetime if not tried," accepts what has been done without looking back, even if the cost is already on the table.
The real hammer fell over forty days later.
On January 20, 2025, R1 was released. Free, open-source, with reasoning capabilities close to OpenAI's o1. The Hangzhou team of over a hundred people turned the global capital market upside down within a week. Carnegie's researcher Matt Sheehan said something interesting: DeepSeek was not the company originally selected by China; its explosion in popularity even surprised China.
Meanwhile, Liang Wenfeng was celebrating the Lunar New Year in Wuchuan. On the afternoon of January 27, he played football with his middle school classmates in the village. The village entrance was crowded with tourists checking in, but the focus was on the football field.
For Yang Zhilin, this was a double kill. Just as the arbitration ended, R1 directly sentenced him to the death penalty for his past year's strategy: users acquired through paid acquisition were worthless in the face of a free and stronger competitor. Public opinion reversed, with an article titled "Yang Zhilin, the Suspension of a Post-90s Idealist." Suspension here means not touching the ground. When he was writing songs, he was worried about becoming utilitarian. Now the whole world says he is both utilitarian and a failure.
On the dark side of the moon in early 2025, there was still money in the bank, but very little say.
The Reversal Against Gravity
The next year is a crucial year for Yang Zhilin.
Yang Zhilin almost completely negated himself from the past year. Stopping ad placements, cutting redundant business lines, retreating to the basic model, and turning to open source. In a conversation at GeekPark, he said the organization's inertia is to want to do more and more things, "We have to fight against this gravity."
This statement sounds light, but in practice, it means admitting the wrong direction, laying off people he hired, and bowing to the opponent who almost killed him. Most 33-year-old founders cannot get over this hurdle. Yang Zhilin crossed it exceptionally decisively, perhaps because open source and long-termism were his default settings, and closed-source paid acquisition was just the outfit capital put on him, which he has now taken off.
In July 2025, the trillion-parameter K2 went open source. In November, K2 Thinking outperformed GPT-5 on several of the toughest Agent benchmarks, and Hugging Face co-founder Thomas Wolf asked on Twitter: Is this another DeepSeek moment?
In the early hours after the release, Yang Zhilin, with Zhou Xinyu and Wu Yuxin, held an AMA on Reddit, answering 21 questions in a row. He clarified that the $4.6 million training cost was not an official figure, admitted that the GPU count was not as high as their American counterparts, "but we squeezed the performance of each card to the extreme." When asked how they view OpenAI's burn rate, Zhou Xinyu responded nonchalantly: "We don't know either, only Sam knows, we have our own pace."
Their own pace. The dark side of the moon in 2024 couldn't say these five words. At that time, its pace was that of investors, the pace of ROI from ad spending. Returning these five words to Yang Zhilin was precisely Liang Wenfeng. R1 proved that open source plus algorithmic efficiency could work in China, essentially pitching to his own board for Yang Zhilin. The person who was nearly killed by DeepSeek now had to thank it for his salvation.
The market's response was also very direct. On the last day of 2025, the dark side of the moon announced a $500 million Series C round. Alibaba, Tencent, and Wang Huiwen all participated, valuing the company at $4.3 billion with over $10 billion in cash. In less than 20 days after the release of K2.5, the revenue exceeded that of the entire year 2025, personal subscription orders increased by over 80 times month-on-month, breaking into the top ten of the global Stripe ranking. And then came K3 this week.
Seven years ago, the Ph.D. student who said, "Not trying will lead to a lifetime of regret," has now made the U.S. tech industry begin to doubt life, becoming an example used to criticize the White House.
The Hermit's Bill
But reality never just shows one side of the face. After 2025, it was Liang Wenfeng's turn to foot the bill.
He didn't have a life, but his employees did. Luo Fuli, Wang Bingsuan, Wei Haoran, Ruan Chong, these names were prominent core members within DeepSeek, and starting in 2025, they successively left, with many directly taking on roles as business leaders elsewhere.
A painful rumor was circulating: when colleagues at a similar level can leave and earn so much, why can't I? A research utopia without hierarchy, KPIs, or discussions about money, relying on members' loyalty to the problem itself. However, R1 inflated everyone's market value tenfold, and loyalty faced a priced opponent for the first time.
Money also became an issue. The scale of the magic square management shrunk from its peak to just over $20 billion, and the batteries for love were leaking power. Then, an unimaginable scene occurred: the person who once said "no short-term financing plan" started meeting with investors.
The asking price started at $5 billion for a single investment, later reduced to $1.5 billion. But throughout the entire negotiation, what he repeatedly emphasized was not the valuation or shareholding ratio, but the same condition: no headhunting of DeepSeek's people, no instigating them to leave for entrepreneurship. A funding round ended up as a non-poaching agreement. He could give up shares, but what he couldn't give up was the atmosphere of that laboratory.
By 2026, two paths had led to an unexpected situation.
Yang Zhilin, cutting spending, scrutinizing every H800's performance, and speaking of his own rhythm, was becoming more and more like Liang Wenfeng. Liang Wenfeng, meeting with VCs, worrying about retention, and for the first time being distracted by such mundane organizational issues, had to step out of the lab.
One used to store water, the other used to ride the tide, but now the one who stored water found that the reservoir could also leak, while the one riding the tide finally welcomed their own wave.
Two Cantonese People Rewriting China's AI Landscape
During the most lively period of China's AI entrepreneurship, two names were brought to the forefront: Liang Wenfeng and Yang Zhilin.
They didn't quite resemble the founders familiar to the Chinese internet over the past decade. There were no memorable quotes on posters, no legendary tales from banquets, and no strong desire to shape themselves into legendary entrepreneurs. However, their position placed them closer to the intersection of money, power, and the zeitgeist than most entrepreneurs in the past.
What sets Liang Wenfeng apart is that he seems to have almost no "personal life." In public reports, he appears more like someone engulfed by work: from quantitative investing, to building a computing power cluster, to DeepSeek, his narrative rarely involves family, consumption, hobbies, socializing, and is almost entirely filled with models, architectures, computing power, organization, and originality.
Yang Zhilin's uniqueness lies in the fact that he seems to have no "way out." After the founding of the Dark Side of the Moon, Kimi quickly became one of China's most watched AI applications, with financing, valuation, user growth, model iteration, commercialization, and arbitration by old shareholders all pressing down on this company simultaneously. The faster it runs, the less time it has to stop and explain.
This is where the two individuals are most alike and yet most different.
Liang Wenfeng seems like a technology idealist who emerged from the depths of the capital market. He first proved in Magic Cube Quant that AI could directly change the flow of money, then directed this ability toward large models. DeepSeek's story was not initially driven by financing but by the funds, computing power, and engineering culture accumulated by Magic Cube. What the outside world later remembered was R1, V3, low-cost training, and open-source shaking up the U.S. market, but more crucial was that Liang Wenfeng early on defined the problem as the gap between "originality and imitation," rather than "how Chinese large models can make money."
This makes him seem anti-commercial yet extremely commercial.
The anti-commercial aspect is that he repeatedly played down short-term monetization in public interviews, unwilling to portray DeepSeek as an internet story of grabbing users and market share. The extremely commercial aspect is that every step he took aimed directly at the lowest-level cost structure of the AI industry: training efficiency, inference costs, model architecture, talent density, chip constraints. If a model can approach top-tier capabilities at a lower cost, it changes not just a product list but the entire industry's imagination of capital expenditure.
That's why the global market reacted so strongly when DeepSeek exploded in popularity. It wasn't just another Chinese chatbot but a reminder to investors: the most expensive logic chain in the U.S. AI narrative over the past two years may not be as solid as imagined. Bigger models, more GPUs, higher capital expenditures do not necessarily mean an insurmountable moat.
Yang Zhilin faces a different fate.
The Dark Side of the Moon has been in the spotlight since its inception. The founder's resume is impressive: Tsinghua undergraduate, Carnegie Mellon Ph.D., involved in key research like Transformer-XL, XLNet; the company was founded in 2023, and Kimi, using long context, quickly became an AI product that even ordinary users could perceive. Unlike DeepSeek, which was first discovered by the tech community as a research organization, it entered mass products, capital markets, and industrial narratives earlier.
This has given Yang Zhilin a great advantage, but it has also placed a huge burden on him.
The advantage is that Kimi has a product mindset. Many people seriously used domestic AI for the first time not because they understood a certain technical report, but because Kimi can read long articles, organize materials, and handle workflows. As AI transitions from the lab to the office, it needs an entrance that ordinary people can remember, and Kimi once seized that position.
The burden is that once the product mindset is established, it must be consistently nurtured. Users will wait for a stronger model, investors will wait for higher revenue, the team will wait for greater equity value, and competitors will wait for you to make a mistake. The more funding Moon's Dark Side receives, the higher its valuation, and the less likely Yang Zhilin will retreat to a quiet researcher position.
By 2026, this pressure becomes clearer. Public reports indicate that in May 2026, Moon's Dark Side completed a new round of funding of about $2 billion, with a post-investment valuation surpassing $200 billion; the company's official website also places Kimi K2 and the code, Agent's capability in a core position. The capital market does not give it applause, but rather a bill: you must prove that you are not just the "company with the best product model," but also that you can maintain a dual technological and commercial lead in the sandwiching of DeepSeek, Alibaba, ByteDance, MiniMax, ZhiPu, and others.
What's even more troublesome is the shadow left behind by Yang Zhilin's previous entrepreneurial project. In 2024, media such as the South China Morning Post reported that Moon's Dark Side founder Yang Zhilin and co-founder Zhang Yutao were arbitrated by a former investor in Circular Intelligence in Hong Kong, China; Yang Zhilin claimed to have completed the necessary procedures to leave Circular Intelligence and start a new business, while the investor presented a different view. The core of this dispute is not just a personal dispute of a founder, but a sharper issue in the wave of Chinese AI entrepreneurship: when the value of a new company soars, where exactly is the boundary between the old company, old shareholders, old team, old intellectual property, and new financing?
Yang Zhilin cannot just portray himself as a genius researcher. The scale of financing, commercial revenue, product iteration, IPO expectations, legal disputes, all demand that he become a more complete and ruthless CEO. A researcher can prove himself with a paper, but a CEO must prove himself with an organization. Papers can be signed, but organizations cannot run solely on signatures.
The more low-key a person is, the more the era magnifies them. The more a person wants to move forward, the more the past catches up with them.
In this round of the Chinese AI story, the final winner may not necessarily be the best storyteller. It is more likely to be won by those who can bear three things at the same time: the uncertainty of technology, the patient consumption of capital, and the cost of the founder being both mythologized and judged.
It is not known whether Liang Wenfeng has a normal life. And it is not yet the final answer whether Yang Zhilin has a way out.
But at least for now, they both don't have much room to return to an ordinary person's status.
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia
