Vol.224: A Conversation with Chaos Founder 李善友 in the Tide of the Times

Summary

• The defining keyword for entrepreneurship in the AI era has changed. 李善友’s core call is that “cognitive upgrading,” the keyword of the mobile internet’s first decade, is now merely the floor; “your ceiling is not determined by cognition, but by philosophy.” The entrepreneurs of the future will be driven along 2 dimensions—cognition and philosophy—where philosophy means “the great vow deep in your soul,” a calling expressed “through me,” not a mission, vision, or values statement. He calls this an “evolution in consciousness,” and believes the Philosophy World course that “flowed out of him” during the pandemic “was made for the AI era.”
• “Will the AI era produce a new 张一鸣?” His answer is an unequivocal yes. The progression across 3 generations: the PC generation practiced “copy to China” with little original innovation; the mobile internet generation—张一鸣, 黄峥, 王兴—built application innovations from 1 to 10, with TikTok going global and breaking into the US; the AI generation may produce original 0-to-1 innovation. 梁文锋 “pierced the paper window that had hidden the truth from an entire generation,” people born in 96, 97, and 1998—the 08 Olympics generation—carry confidence “in their bones,” scientists are becoming entrepreneurs, and companies go global from day 1. “I am highly bullish on China-born entrepreneurs in the coming AI era.”
• BAT’s failure to internationalize was a clash across cognitive levels. His explanation: the PC generation was rooted in experience and intuition, corresponding to agrarian civilization; the mobile internet generation stood on logic and rationality, corresponding to industrial civilization. “It works in theory and was bound to happen in practice,” creating a power gap over the prior generation. BAT was attacking uphill against a world dominated by industrial civilization, while ByteDance and Pinduoduo could move across on level ground. 李翔 added that BAT’s experience of defeating global giants at home led them to model their odds abroad as unfavorable.
• The test for a new paradigm is whether it is native. The new AI-era paradigm is “definitely not 豆包 or 千问,” but native players such as DeepSeek and Kimi—just as 张一鸣 was native to mobile internet, while “the two Mas” migrated from PC. 李翔 joked: “When 张一鸣 retires, he will simply be moving his physical body from mobile internet to AI.” The warning for new-paradigm founders: if they build a to C agent, “they may well be operating inside someone else’s comfort zone.”
• Entrepreneurship is slowing, but innovation is accelerating. That decade was “the best era for Chinese merchants in the past 2,000 years”; today, “who talks about fundraising at Starbucks anymore? The angel-investing industry has practically disappeared.” Entrepreneurship is no longer running at 10x, but innovation is; the funnel is narrower and the bar higher, so founders increasingly need technology and scientific innovation to break through. He never joins the bubble-watch chorus: “My underlying disposition is still optimism.” Two signals in 2025: people who quickly emerged from pessimism and reset their entrepreneurial state are doing well, while young AI founders “bloom at the first bit of sunshine.”
• AI and humanity follow the logic of “termites don’t eat.” Termites do not eat living trees: “If your roots are still pure rationality and logic, AI will eat you,” because AI itself is built on rationality and logic. If you move up to the levels of spirit, consciousness, and creativity, “AI becomes your tool”; otherwise, “you are the person in The Matrix supplying raw material to AI.” The AI industry iterates “by the week,” and Silicon Valley workers now aspire merely to “get in good physical shape,” but “no one is asking about AI’s first principles anymore.” He intends to be part of that minority.
• The line between speculation and innovation—and between fraud and failed innovation—lies in motive. Speculation means “doing whatever is hot”; innovation means “I was here when it was hot, and I was here when it wasn’t,” as in 梁文锋’s case. A fraudster’s ego disguises itself as a great vow; an innovator has the feeling of being a conduit, alongside the energy of humility. The success rate of innovation is “only a little greater than zero,” as in Elon Musk’s chopstick rocket catch. “I am not cheering for the successful; I am cheering for innovators”—which is why 徐正’s portrait still hangs on Chaos’s homepage. “Success is ratified after the fact.”

Deep dive

1. Three chapters: maximizing the employee track, a brutal startup, and business education

• 李翔 opened by tracing the résumé: a mathematics graduate from Nankai, he started at Motorola, a Fortune 500 company, then joined Sohu in 2000 and became editor-in-chief and senior vice president. He founded Ku6 in 2006, entering a battle with Tudou and Youku whose brutality was no less than the group-buying wars, food delivery, and bike-sharing wars that followed. According to the opening introduction, Ku6 should have been the first Chinese video site to list on Nasdaq.
• The third chapter was business education. Starting with the mobile internet generation of founders, “the overwhelming majority of people have heard his classes”; he introduced and popularized concepts now repeated everywhere, including internet thinking, the second curve, and first principles. Outside Chaos’s own classes and events, 李善友 rarely gives interviews. This was an unusually long conversation.

2. Connecting the dots: “feeding one mouth with another” at 26

• Using the framework of Steve Jobs’ Stanford commencement speech, 李善友 looked back: “When you’re in the middle of it, you can’t see the direction; looking back, many dots connect into a line.” On the surface, the sequence was chaotic—mathematics, HR, editor-in-chief, entrepreneurship, professor—but “there really was something fated in it”: he was class monitor in every class as a child; when no one wanted to lead Motorola’s new-hire orientation, he volunteered, and “scored higher than many professionals and bosses.”
• The strongest evidence was a 1998 article he wrote at age 26 titled “Feeding One Mouth with Another.” He aspired to become a highly prestigious corporate trainer, then believed the idea that “some things you need to say and express really can come true.” The side path gradually became the main road: at Sohu, he discovered that after sitting through any course 3 times, he could perhaps teach it better than the instructor—“the only exception was 王东岳’s course. I couldn’t teach that one.”

3. The collapse of Chinese-classics management theory: “It was useless, so the startup failed”

• Starting in 2004, he became obsessed with Chinese classics, blogging and publishing the book Managing Through Inner Sageliness and Outer Kingliness in an attempt to fuse classical Chinese thought with management. He disliked people who “read 1 book and immediately went out to teach it”: without personal practice, “what you say has no power; it feels fake.” So he applied the theory while serving as Sohu’s editor-in-chief and treated its implementation as a parallel track during his startup.
• He eventually owned the mistake: “In fact, this was nonsense. The theory was completely useless, so the startup failed.” But the failure left 2 things behind: a complete experience of foreign companies, the internet, and entrepreneurship, plus the desire to teach. “Those 2 things seemed to have been preparing me.”

4. Selling Ku6: financially free, but “a complete failure”

• He sold Ku6 after its 2011 listing. The outside view was a successful exit and financial freedom, but “deep down, I always felt the whole thing was a complete failure.” His body also broke down—“like a pot of herbal medicine boiled dry into dregs.” He was working well beyond 996, drinking to an “unbearable” degree, and going through an extremely dark period.
• After the sale, he had no idea what to do. The conventional options were serial entrepreneurship or investing, but neither made sense to him. He went to the park every day to read 王阳明’s dense books. That vacuum became the prelude to his next turn.

5. God’s phone call: joining CEIBS at 39, a wormhole that opened and closed

• He describes it as “receiving a phone call from God.” CEIBS’s new dean, 奎尔奇—a former Harvard associate dean and former London Business School dean—wanted to replicate London’s model by bringing back an entrepreneurial alumnus to run an entrepreneurship center. 李善友 had been the class representative during his CEIBS EMBA and fit the profile perfectly. Professor 周东生’s phone call arranged the meeting. 李善友 made an audacious request: “Could you make me a professor?” The dean agreed, despite his lack of a PhD.
• He went on to launch the CEIBS entrepreneurship program, earlier than the 2013-14 “mass entrepreneurship and innovation” push and among China’s earliest entrepreneurship-education efforts. The strange footnote: he was the first—and the last—CEIBS graduate to return as a professor. “History opened a slit in spacetime, a wormhole. Once I passed through, it closed. Was it destiny or chance? I don’t know.”
• He also acknowledged the role of timing. At 39, he was in his darkest period: “If my career had been booming then, and that call had come, I might even have hesitated.” In retrospect, he says, building video in the PC era was simply the wrong move; none of the companies ultimately made it.

6. The engine of Season 1: desire, fear, and “staying in a senior-year-of-high-school state for life”

• He divides his life into 2 seasons. “Season 1 is when a person uses his ego to build a company.” The drivers were not a sense of mission but desire and fear: desire for more fame, influence, and money; fear of “being free from poverty.” The roots went back to childhood. He came from the countryside, “came close to hunger and experienced a little of it,” and carried intense scarcity and insecurity. “There was only 1 path that could change me: go to university.”
• A childhood friend’s assessment became a signature line: “You have stayed in the mindset of a high-school senior your entire life.” After teaching a major class, he still reviewed and studied the next day. “If I don’t work hard for 1 day, I feel a little guilty.” In reality, he says, this was fear, scarcity, and insecurity driving him—a constant, exhausting need to prove himself. 李翔 connected it to the previous generation’s worship of extreme grind and wolfishness: “There was fear underneath it.”

7. More frightening than failing is completing the goal: achievement as a bigger demon

• More dangerous than missing a goal is achieving it: “The moment you complete it, you think, ‘Is that all?’ Then you set another goal and enter a state of torturing yourself.”
• “The demon of having is bigger than the demon of losing.” Once you have achieved something, it proves you were right, so you rationalize parts of yourself and push the problem down. That is ultimately more dangerous. The attribution mechanism is decisive: success makes people say, “This must be right”; failure can be constructive because it prompts reflection—“Maybe this approach was wrong.” This thread runs through his later self-criticism of periods of expansion.

8. The wild wolf and the wilderness: using an entrepreneurial mindset to teach entrepreneurship

• A widely circulated 2015 essay began: “I have always felt that I am a wild wolf. Even in comfortable surroundings, I will eventually return to the wilderness.” 李翔 asked what wolves and wilderness had to do with education. His answer was that the spirit of entrepreneurship is his underlying character—“as my wife puts it, I like to stir things up, doing it once every few years and making her worry once every few years.”
• His methodological self-awareness was the key. While building the CEIBS entrepreneurship program, “I used an entrepreneurial mindset to build the project”: he personally interviewed and selected students, recruited speakers from industry, and devoted all his energy to it 24 hours a day. “If I had used the mindset of a business school or a professional manager, I couldn’t have done it. I wouldn’t have been on the same frequency as the students.” Business education is fundamentally management education, and its rhythm is different from entrepreneurship education.

9. The case-study dilemma: Southwest Airlines, taught 30 years later

• He was always an outlier in business schools. “The Southwest Airlines case happened in the US 30 years ago, and 30 years later it is still being taught in China—by strategy professors, HR professors, and finance professors. Deep down, I felt something was strange.”
• His starting point was “internet thinking.” The term did not even exist in the US, yet appeared in China and became an obsession. The best-known formulation was 雷军’s “focus, excellence, reputation, speed.” He began with the public feud between 雷军 and 周鸿祎, but quickly found the spectacle boring: “This is still a phenomenon, a story. There should be a theory underneath it. Is there some invariant theory?” That question defined everything that followed.

10. Christensen laid the foundation: “I wanted to slap myself”

• When 周鸿祎 was starting a company, he recommended The Innovator’s Dilemma. 李善友 bought it but never read it: “I couldn’t get into it.” After becoming a professor, he tried again and “suddenly it clicked. I wanted to slap myself: I should have read this book properly back then.” Having founded and failed a company, he now found echoes of his own experience throughout it.
• That led to his first course, disruptive innovation. “I think I was one of the few people in China who taught this course best.” One choice still mystifies him: the program was called the CEIBS entrepreneurship camp, but “from day 1, I didn’t teach entrepreneurship. I taught innovation. If you ask me why, I don’t know.”

11. 王东岳 and first causes: first principles born from dissatisfaction with induction

• The real intellectual turning point was 王东岳. Beginning in 2013, 李善友 attended his lectures for 3 consecutive years and read The General Theory of Evolution. “He brought me into a philosophical world.” The core lesson was a Greek-style pursuit of the ultimate: “Don’t look for causes at the level of appearances. There has to be a first cause; behind everything is a simple law operating at a broad, cross-disciplinary scale.”
• With that standard in hand, he became dissatisfied even with disruptive innovation: “It was an induction from several stories. You induce a theory, but if there is nothing foundational underneath it, no simple law, then it is just induction. It doesn’t hold.”

12. Studying at Stanford: skipping business-school classes for philosophy and physics

• The family moved to Stanford for 1 year of visiting study, “a little like 玄奘 going west for Buddhist scriptures,” in search of “the extremely real simple law underneath entrepreneurship and innovation.” In the first semester he attended business-school classes, each taught by a Silicon Valley entrepreneur and a professor walking through the entrepreneur’s own case. “Good, but still stories. No one was looking for the invariant underneath.” In the second semester he stopped attending and switched to philosophy, physics, biology, and cosmology. His classmates were undergraduates. He “tore off every label from the past and immersed himself in the ocean of knowledge.”
• 2 details stayed with him: he cycled once around his home every day before heading to campus; and the original impulse for Chaos University—“a university without walls”—took shape at Stanford.

13. Three sources for the first-principles course—and the mocked “philosophy-and-science mindset”

• His second course, First Principles, came together from 3 sources: Elon Musk, the first person in the entrepreneurial world to talk about first principles, who “gave me an entry point”; 张守成, whose Stanford class “The West Wind Behind 2012” explained first principles through physics and was “the single class that helped me most”; and 王东岳 and Aristotle, the intellectual sources whose meaning he was pursuing—“if you chase the roots of first principles, you eventually reach Aristotle.”
• He sees his own contribution as the graft: “No one in China was explaining philosophy to entrepreneurs, so I kept believing, tirelessly, that a philosophy-and-science mindset illuminates innovation.” A business-school colleague once mocked him on social media, saying the 2 concepts “simply did not go together.” “I was thin-skinned then and got angry.” But today, “no one fails to talk about first principles, and no one fails to talk about cognition.”
• He invoked Whitehead to defend the value of articulation: before understanding something, express it first; expression helps understanding, and before expressing it, you must feel the thing’s importance. “I have a feeling for whether every word matters.”

14. Chaos Online: 500 classes in 10 years, a first-person record of an era

• Discussing Chaos’s historical role, he first separated Chaos Online from the courses he taught himself. He is “quite proud” of the online platform: it replicated his pattern by bringing in entrepreneurs to explain the cognition they acquired while building companies—not to tell stories or boast. It coincided with the decade of mass entrepreneurship and innovation in mobile internet. “50 classes a year, 500-600 people over 10 years.”
• His role was that of an era’s recorder: documenting “the best 10 years for Chinese merchants in several thousand years of history,” in the first person. “Look at the classes 左晖 came to teach—my God.” To 李翔 he said: “That’s why I like your The Detailed Interview. We share the same instinct: it is both communication and record-keeping.” The slogans “accompany the people with the biggest dreams of this era” and “help you understand this complicated world half a step early” became more precious to him over time.

15. One instructor across 7 modules: Chaos exists to ask questions

• Chaos School’s format was unique: it admitted only 60 top-tier entrepreneurs a year, and all 7 modules—21 days in total—were taught by him alone. “You won’t see that at any business school.” The downside was uniformity and that he was easy to attack; the upside was the ability to build a fully self-consistent system.
• Recently he asked, “Who exactly is Chaos?” His answer: “Chaos exists to ask questions.” For 15 years, it has done 1 thing—pursue the first principles behind entrepreneurship. The answer in the first 10 years was cognitive upgrading. Entering the AI era, “I am asking again: how is this era different from the previous one?” 李翔 asked whether first principles change with technological paradigms. He was emphatic: “If they change, then they were not first principles.”

16. Will the AI era produce a new 张一鸣?—“It has to be yes”

• The first question he plans to ask at the start of the new term is: “Will the AI era produce a new 张一鸣?” The mechanism is that “most people believe only what they can see, and what they can see are the giants.” A decade ago, everyone asked how to escape BAT’s gravitational black hole and what to do if 马云 was your competitor. Then the mobile internet produced 张一鸣, who did not compete directly inside PC; he entered a new era after a paradigm shift.
• His 2 reasons: first, a theoretical conviction—he does not know who the person will be, but there will definitely be one; second, faith. “What I worship is innovation. I am not cheering for the giants; I am cheering for innovators in a new era. That is my mission.” 李翔 compared it with the question 曾鸣 used to ask: Which industry will produce the next $100B company? The same question in another form.

17. Only 3 courses in 15 years, each with a figure to honor

• “In these 15 years, I have taught only 3 courses.” Disruptive Innovation honors Christensen: “The architecture of the course is his; the cases are mine. That’s nothing special.” First Principles honors 王东岳 and Aristotle, with Musk providing the entry point. “This course contains my own creation: bringing philosophy and business—2 unrelated things—together.” Philosophy World, created after the pandemic, honors Plato and Siddhartha by combining Western philosophical thinking with Eastern inwardness and expressing both rationally.
• He also described the shift in his aesthetic focus: while teaching first principles he was obsessed with Aristotle; when he began teaching the world of ideas, he moved toward Plato. Now, “Plato’s position is simply too high. Philosophy has still not moved beyond the questions Plato discussed.”

18. The birth of the world of ideas: “This course’s consciousness frequency is higher than mine”

• In the first 3 months of the pandemic in early 2020, he read books and watched films without any instrumental purpose for the first time. Previously, he was fear-driven and read only to prepare classes. After randomly consuming biographies, books on the mind, and science-fiction films for 3 months, “the structure of a course suddenly appeared.”
• He never claims to have created it. “The level of this course is higher than mine—much higher. It could not have been created by me. The course actually elevated me.” His favorite phrase is “through me”: “Something larger was expressing itself through me. The feeling was so strong that I even experienced serious self-doubt.” He tested the material with an internal group for 3 days, and everyone reacted: “Wow, something new has emerged. It is completely unlike my old style.”
• 李翔 connected this to artists’ experience. John Lennon said God whispered softly in his ear and he was responsible for writing it down. Michael Jackson, floating in a swimming pool, said he was seriously receiving God’s signals: “If I’m not ready, God may find someone else—perhaps Prince.” 李善友 said he had once believed this belonged only to artists, but every profession is an art: create work in a state of flow; work itself is art. That is where the course’s concept of a “beautiful work” came from.

19. Only in hindsight: the course was made for the AI era

• After he began teaching Philosophy World, the audience changed. Previously, entrepreneurs came; suddenly couples and parents with children began attending. He deeply questioned himself: “Was I moving into the mind-body-spirit world?” The course had “completely flowed out of me”; it was supply-side creation, created without thinking about whom it was for.
• Only recently did he understand it: “This course was made for the AI era, to prepare for the next 10 years.” First Principles had prepared people for the previous decade’s cognitive upgrading. “The core difference between AI-era entrepreneurs and mobile internet entrepreneurs is no longer cognitive upgrading. I call it an evolution in consciousness, or the evolution of consciousness. Strange, right? I didn’t know that while I was making it.”

20. “Teaching with my life” is not a figure of speech: student zero, structural arrogance, and 3-hour conversations

• Whether he is speaking to a small class of 60 or a large audience of 3,000-5,000, his expression is identical. “Every time I teach, I am student zero—I feel as if I am teaching myself. I don’t care how many people are there. What matters is pushing my current understanding to 101%, across the threshold, holding nothing back.” “I teach with my life”—for him, that is not a figure of speech.
• He is wary of a professional hazard. People who lecture constantly become addicted to teaching others and develop “structural arrogance”: “No matter how humble you are, the structure of your expression itself contains structural arrogance. The profession is stabbing you.” Teaching himself gives him a sense of growth and youth rather than making him increasingly arrogant.
• He has recently added a third mode: one-on-one long conversations, lasting at least 3 hours and sometimes 6. “After 3 hours, it is no longer an exchange of minds but an interaction of souls. It is nourishing.” He says he is not a good questioner but a listener. 李翔 disagreed: “I think you are a good questioner.”

21. Professor or founder? Chaos could never IPO, and in 2019 returned investors’ money

• Forced to choose, he picks professor and scholar. “Chaos is related to my career, but scholarship is related to my mission.” He is blunt about Chaos’s commercial ceiling: “Chaos can never IPO. When it was expanding, I considered it briefly, then quickly realized something was wrong, so in 2019 I returned the money to investors. I knew it was not scalable.” Besides, “I have already been the CEO of a listed company. I have had that satisfaction; repeating it is not very interesting.”
• His academic standard comes from 王东岳: “To do scholarship, you first need to accumulate some understanding. After teaching for a lifetime, you should at least understand something, and you have to bring that understanding out.” He cares more about the research itself than the front-stage act of lecturing.

22. Season 2’s keyword: personal growth, and the regret of being “too serious”

• The identity that matters most to him now is neither professor nor founder but “a personal growth practitioner—not growth in external business or theory, but growth inside myself, and how to become a complete person.” He contrasted himself with the author of In Search of Excellence, who described himself as a player and wanted “player” on his gravestone. The thing he is worst at is play: “As a rural child, I studied constantly—primary school in the village, middle school in the township, high school in the county. I never played as a child. That is a huge regret, so I became too serious.”
• 李翔 countered with Susan Sontag: “Can you imagine telling Shakespeare to relax a little?” 李善友 accepted the point but still wants to change. “I hope to become more relaxed and softer. Luckily, even at this age I am still growing and changing.”

23. Two shocks from 王东岳: the feeling of ignorance, and the purity of “planning to die after finishing the book”

• The first shock was ignorance. At the first 东岳 study workshop in 2013, there were only 30-some people. Half fell asleep during the lecture, and 李善友 understood little. “How can there be so much that I don’t know? I felt so stupid.” Yet he sat in the front row and nodded throughout all 3 years. 王东岳 later told him that his persistence was because 李善友 was in the front row.
• The second shock reordered his life. 王东岳 was “one of the purest people I have ever met”; “he planned to kill himself after finishing that book.” He had expressed the idea that what he was doing mattered more than his life. “That cracked open my understanding.” From then on, “teaching with my life” meant that the course mattered more than his life. Without 王东岳, he would have thought entrepreneurship and Chaos were most important; because of him, scholarship became most important.

24. Mission as one: not occupying that position would be the greatest harm to the institution

• Asked about the relationship between his personal mission and Chaos’s mission, he admitted to wrestling with it. Saying that Chaos’s mission was 李善友’s mission felt too ego-driven to say aloud. But the conclusion reversed: “The farther you go, the more you see that I have the greatest influence on Chaos’s character. My refusal to occupy that position because I am afraid people will think I am egotistical is actually the greatest harm to this institution.”
• The mechanism is that organizations need someone willing to make a certain degree of sacrifice, and that sacrifice may mean exposing parts of one’s ego. “Chaos and I are becoming increasingly one.”

25. Accompanying innovators of the era: 梁文锋, “a little greater than zero,” and a reprimanded remark

• 梁文锋 made him truly understand the word “innovation,” which he had used for years. “The probability of 梁文锋’s path succeeding may be only a little greater than zero.” The formulation came from Musk: asked about the odds of catching a rocket with chopsticks, Musk laughed and said they were only a little greater than zero. “I am not cheering for successful people. Media companies are willing to cheer for winners because it is safe. I cheer for innovators not because they succeeded, but because success is so difficult and they kept going.”
• AI and 梁文锋 also gave him a sense of national responsibility. Last year he told colleagues something they found hard to accept: “If this class has no meaning for China, it has no meaning for Chaos. You can criticize me for saying that—how much money can this class make? But if I accompany innovators through an especially difficult period, and in the future you can shoulder China’s AI destiny, then if my companionship has meant something to you, I would find that beautiful.”

26. Missing 张一鸣: the only person who asked whether the course was “root” enough

• There were “many, many” disappointments in admissions. The biggest regret: 张一鸣 applied to Chaos in 2019. “一鸣 kept asking me, ‘Professor, is what Chaos teaches root enough?’ He was the only person to ask whether it was root enough and whether it included foundational disciplines. Everyone else asked who their classmates would be and what problems it could solve for them. You can see how well matched he was with me.” He did not come that year for several reasons. “That miss is still a real regret.”
• Another category of loss was self-inflicted. Because of the interview process, “our own arrogance caused us to miss many people. We rejected them without understanding them. That was deeply arrogant.”

27. Cognition is the floor, philosophy the ceiling: the line between speculation and innovation

• His verdict on the AI era is clear: “The mobile internet era was about cognition, but cognition alone is no longer enough. Cognition is the floor; what determines your ceiling is philosophy.” And philosophy is not a mission, vision, or values statement in your head; it is “the great vow deep in your soul.” The next generation of founders will be driven by both cognition and philosophy.
• He repeatedly cites 梁文锋’s distinction: “Speculation is doing whatever is hot. Innovation is that when something becomes hot, I was already here—I was here when it was hot, and I was here when it wasn’t.” 李翔 connected it to a story—likely involving a DeepMind founder—about looking down on Altman: “I was pursuing AGI long ago; you came only after it became hot.” 李善友 vouched for himself: “I have been doing this for 15 years. I didn’t start today. After becoming this kind of pure person myself, I will naturally prefer people like that.”

28. Distinguishing fraud from failed innovation: an ego’s great vow or a vow expressed through you

• 李翔 posed the 贾跃亭-style question: how do you distinguish failed innovation from fraud? 李善友 first admitted, “I don’t know how to answer that,” then offered his test. There are 2 possibilities: one is ego-driven and disguises itself as a great vow; the other has a feeling of being a conduit. The vow may sound grand, but the person carries an energy of humility. “The energy behind humility is enormous. It does not come from ego.”
• Musk is the test case. “Musk is not a humble person, but he has that feeling of being a conduit. You almost never see someone else so real—he says exactly what he means, with a kind of sincerity. He may be inflated in the moment, but the inflation itself is his sincerity in that moment.” When discussing SpaceX, Musk says humanity should leave Earth because Earth is our cradle. “His voice trembles a little in that minute. You know the sentence is sincere.” Going to Mars is not entrepreneurship—entrepreneurship requires safety and returns. “This is innovation. The odds are tiny, but the guy truly believes it.” He has stopped teaching Bezos: “I no longer feel anything for this person, so I can’t teach him.”
• He compared 2 remarks from Charlie Munger: first, that he preferred someone with an IQ of 100 who believed it was 80; later, that “sometimes we underestimate people who are extremely arrogant.” 李善友’s verdict: “There are 2 kinds of arrogance: ego’s arrogance, and the arrogance of being a conduit. You have to tell them apart.”

29. Spiritual practice in business: the business Buddha, the philosopher-king, and 王阳明

• The people he admires fall into 2 groups: sincere innovators in business—Jobs and Musk—and figures from the spiritual world—Plato, Buddha, and Laozi. “These 2 groups used to be split apart. I am beginning to bring them together. Spiritual practice in business is the most important phrase for me right now.” Business is full of greed, anger, and delusion; money magnifies everything and makes it easy to lose one’s soul. But “business is actually the instrument of your practice. The company is the dojo.” He rejects the idea that heroes turn to Buddhism only after their careers end: business itself is practice. “I mean ‘is,’ not as a metaphor.”
• He calls the ideal personality the “philosopher-king of the business world,” citing Jobs, 松下幸之助, 稻盛和夫, Buffett, and Munger. The Eastern synthesizer was 王阳明: Confucianism as practice, integrated with achievement and with knowledge and action as one; he simply used government as his vehicle. The term “business Buddha” came from the Korean novel Sangdo, recommended by 梁宁. “Could today’s Buddha or bodhisattva appear as an entrepreneur or innovator? Most people are not naturally suited to practice directly in a temple; they need a vehicle. You think business is your ceiling, but business is your instrument of practice.”

30. Enlightenment does not require giving up ambition: Zen and the Art of Motorcycle Maintenance and “excellence”

• 李翔 relayed Naval’s question: would pursuing inner peace mean losing one’s drive? 李善友 was emphatic: “No. Quite the opposite.” He had just reread Zen and the Art of Motorcycle Maintenance over the May holiday. The book uses rational language to discuss the suprarational, with “quality” as its keyword: “Quality is the Tao,” and some translators render it as the highest good.
• The book pushed his thinking 1 step further. He had traced truth through Socrates, Plato, and Aristotle, who placed “the One” at the center. The book says truth is not that One; the final word is excellence. “Excellence is the One—the highest good, conscience, and beauty. A true practitioner is a person of excellence.” Inner growth in business is not abandoning the pursuit of achievement. “Find your state of excellence, and you are practicing in business.”

31. 黄仁勋 dissolves competition: “the happiest form of laziness”

• The most helpful passage for his second half of life came from 黄仁勋: “Are you doing this to make the world better? If so, and if someone else doing it would make the world better even if you didn’t, why are you doing it? That is first principles.” 黄仁勋 called it “the happiest form of laziness.” 李善友’s reading: this dissolves competition. Find the thing that will not happen if you do not do it—the thing only you can do.
• The next sentence matters just as much: after finding that thing, “in the process of becoming excellent, I will become a better person. That is what practice is.” 李翔 added a line from a friend that made him applaud: one view is “I became Mozart because I am a gifted composer”; but “I more strongly believe that because I did this thing, I became a certain kind of person.” 李善友 approved: “Not ‘because of who I am, I did something.’ That is too shallow.”

32. A secular definition of practice: pursuing ultimate questions

• Concerned that religious language might not resonate, he gave practice an operational definition: “Practice means pursuing ultimate questions.” There are 2: What is the One behind the world? What is the One behind me? “By asking what lies behind the world, I am really trying to answer who I am. The moment you ask who you are, that is practice in the broad sense.”
• He has even seriously considered a creation theory. If the ultimate creator is monistic and does not know who it is, it creates all things to achieve 3 purposes: “know who I am, experience myself, and become myself.” The largest and smallest scales mirror each other: “Why do I build companies? To answer the same 3 questions through creating works. Isn’t that practice?” The course structure changed accordingly: 7 modules, with the first 4 using reason to investigate the first principles of entrepreneurship and the latter 4 asking who I am. The resonance is strong: many people say they had never thought about it, but once he frames it this way, they recognize it immediately. “I think the time has come.”

33. Jobs’ devastating question, and a view of success that keeps 徐正 on Chaos’s homepage

• 李翔 returned to a Jobs story. A business-school dean invited Jobs to teach, and Jobs asked which innovators and business leaders of the era the school had produced. The dean had no answer. “Then you are a failed institution. Why would I teach at a failed institution?” 李善友 applauded: “That goes to the essence.”
• The conversation moved to success and failure. 徐正’s portrait remains on Chaos’s homepage even though Daily Fresh is conventionally classified as a failure. His position: Chaos’s core keyword is innovator, not winner. Innovation always carries a probability of failure. “If you use business achievement as the standard, you would not dare invite anyone to teach.” The standard should be vitality: if someone has unique cognition in the present and shares it, it no longer belongs only to that person. It is the tide of the times expressing itself through that person. “Success is ratified after the fact,” like a Nobel Prize often recognizing work done 20 years earlier. Has he ever been shaken by a founder who turned out to be a fraud? “No. If anything, the process has made me believe even more that innovation matters.”

34. Three generations of founders: copy to China → 1 to 10 → 0 to 1

• His generational framework: the PC generation—BAT and the 3 major portals—was “copy to China.” Was there real innovation? “Hard to say.” 张朝阳 turned Yahoo into Sohu; 李彦宏 turned Google into Baidu. The impact and significance were large, but there was little innovation. The second generation—张一鸣, 黄峥, 王兴—broke out under pressure from giants whose competitors could not be seen even through a telescope. “TikTok actually went global and broke into the US. The first generation had the money and technology, so why did none of them produce a global product? The second generation clearly had innovation.”
• But the second generation’s innovation was application-led, from 1 to 10; China still had no 0-to-1 innovation. “I believe in innovation more today: China will produce 0-to-1 innovation in the AI era.” 梁文锋 gave him confidence. “I never used to talk about national destiny; it felt too metaphysical. But in teaching the 梁文锋 case, I actually mentioned it on the final page—he is practically a fulcrum of China’s AI destiny.” 李翔 added that even if DeepSeek was less commercially successful than 豆包, it was still a phenomenon and a marker. 李善友: “Exactly. It pierced the paper window that had hidden the truth from an entire generation.”

35. Confidence in the bones: the 08 Olympics generation and the 3 generations of hardware

• 0-to-1 innovation requires confidence among the young. He observes that AI entrepreneurship is now led by people born in 96, 97, and 1998. Their worldview formed around the 2008 Olympics. “Our generation’s confidence is burdened by the suffering since the First Sino-Japanese War. These kids’ confidence is in their bones.” Add the rise of young scientists as founders, and “innovation will become the core keyword; original 0-to-1 innovation will emerge.”
• 李翔 matched this with a 3-generation framework for hardware founders: the first generation—海尔 and 联想—offered cheaper imitation; the second innovated but lacked confidence and competed on value, represented by 雷军’s Xiaomi; the third redefined hardware, with 王星星’s Unitree leading the field. 李善友 said the 2 fields were identical: “Look at DJI—how extraordinary is that?” But he added a caveat: “I am optimistic about innovation. I may not be as optimistic about entrepreneurship as I used to be.”

36. He never calls a bubble: his underlying stance is belief

• 李翔 noted an anomaly. Despite living through the 2000 internet bubble, when “the 3 major portals nearly were delisted by Nasdaq,” 李善友 has rarely joined the many forecasters warning about bubbles.
• His explanation was a single sentence worth recording: “I think my underlying disposition is still belief. My underlying disposition is still optimism.”

37. No one is asking about AI’s first principles anymore: he intends to be the minority

• The tension between pursuing the eternal and chasing change has widened in the AI era. “AI now iterates by the week.” Those who have not entered the field are anxious; those inside are anxious in both directions—excited to be riding the wave, then depressed when the work is replaced 3 months later. An investor friend returning from Silicon Valley told him that many people there now aspire simply to “get in good physical shape.” Even the form of competition has changed: everyone is competing around “lobsters”—the original wording, apparently referring to AI agent products. Before bed, you assign the agent tasks and let an alter ego work while you sleep.
• His judgment: “No one across the industry is asking about AI’s first principles anymore. Everyone is on the wave, on the iteration cycle, competing desperately on speed.” That makes him more determined to be the minority asking: What are the first principles of large models? What is the future of people and AI? Is AI a new industrial revolution, a new technological revolution, or a new human civilization? Once those questions are pursued, people can find their place in the era.

38. The solution is integration: deeper content, a new vehicle

• His practical answer to the tension between the eternal and the changing is to do both at once. First, make the course deeper—from cognition to philosophy, consciousness, and the mind. As change accelerates, do not chase speed; find the invariant more deeply. Second, “the vehicle of expression must be completely new.” His cases now come from AI: OpenAI, DeepSeek, 黄仁勋, and large models.
• The warning is sharp: if he reaches something deeper but teaches old cases, people will think he is outdated. “Even your mobile internet cases have already become Southwest Airlines.” As for how mainstream business schools view him: “They don’t care, I suppose. I haven’t spoken with people at mainstream business schools about these topics in a long time.” 湖畔, 青腾, and Chaos are completely different, as are Tsinghua, Cheung Kong, and CEIBS; he no longer compares himself. 李翔 added the inverse: “I cannot allow someone else to be better than me at making the world a better place. That would be ego—a form of proving myself.”

39. Gödel and crisis: God’s phone call comes through crisis

• He “strongly believes” in Gödel’s incompleteness theorem, as he understands it: once a system becomes complete, a bug must exist outside the system, and that bug sometimes appears as a crisis. He therefore redefined crisis as a signal: “As I have gotten older, I have realized that God’s phone call comes through crisis. A blockage or crisis means you have to cross a threshold. Crisis tells you that the paradigm no longer works and you need to enter a new one.”
• The dividing line is the response. People who feel the crisis, are willing to break the old paradigm, and move into the new one are remarkable. Those who feel the crisis but keep competing inside the old paradigm, trying to make the old thing bigger, are not doing anything meaningful. He sees the second-generation founders still on a path of self-proof: their careers are already successful, but their inner lives have not grown with them. “Watching from the side, you think: what are you doing?”

40. The new AI paradigm: definitely not 豆包 or 千问; native players will win

• He named names: “The new paradigm in the AI era is definitely not 豆包 or 千问. It is clearly more associated with native players—Kimi, and especially DeepSeek. They represent the new paradigm.” He also warned the new players to ensure they are not using old methods to build the new paradigm. 李翔 added that building a consumer-facing agent “may well put you inside someone else’s comfort zone.”
• 张一鸣 is the prototype of native logic. In mobile internet, he was a native founder; “the 2 Mas” brought over products from PC and expanded PC’s boundaries. ByteDance occupies its current position because 张一鸣 had a new paradigm and was native to it. The same will apply to AI: if you stand in mobile internet and try to win in AI, “I don’t think it will work. You have to have your feet in AI.” 李翔 was half joking and half serious: “When 张一鸣 retires, he will be moving his physical body from mobile internet to AI. I’m not joking.”

41. Nankai, 陈省身, and “if not 张一鸣, then 王一鸣”

• A brief alumni aside: he entered Nankai’s mathematics department in the class of 1990 and had already secured a graduate-school recommendation. 陈省身 was running a mathematics institute at Nankai at the time. 李善友 saw him but never interacted with him; they were in applied mathematics, while 陈省身 worked in pure mathematics. “I knew his status was extraordinarily high, but not as high as I would later understand.” Nankai’s most famous entrepreneur today? “张一鸣.”
• That led to the relationship between universities and exceptional founders. “Every era has its own tide. From the individual’s perspective it is chance; in the tide of the era it is inevitable. If he is not 张一鸣, he will be 王一鸣 or 马一鸣.” The second factor is interaction among people: he believes intelligence can emerge from interaction, which is why exceptional people historically appear in clusters. First comes the era, second interaction. The individual’s role is to be ready: then the tide expresses itself through you, as Michael Jackson said—if he was not ready, it would go to someone else.

42. Two tides: released entrepreneurial spirit and a thirst for knowledge “seen nowhere else”

• Why does China, despite an education system that suppresses creativity, continue to produce entrepreneurs—马云, 马化腾, 张一鸣, 梁文锋, and 王兴兴, none of whom studied abroad? He began with Jobs’ answer in a 2000 interview with the author of Good Strategy Bad Strategy: “I am waiting.” He then invoked 雷军’s “go with the trend.” “I strongly believe there is a tide of the era. If you are ready, it will express itself through you.”
• China has 2 tides. One is the release of a merchant spirit suppressed for many years. The other is a thirst for knowledge. The evidence is 张守成: after returning to China in a wheelchair to teach at Chaos, he invited 李善友 to dinner. “He had seen so many entrepreneurs in Silicon Valley and later became an investor. He said: ‘The thirst for knowledge in the eyes of Chinese entrepreneurs I saw in the Chaos classroom—I have seen it nowhere else.’” That thirst was expressed in the previous wave through the phrase “internet thinking”; today it is expressed through AGI.

43. 2 signals in 2025: resetters are doing well, and AI has released the 90s generation

• He had also been pessimistic for several years, but 2025 gave him a major lift. First, “people who quickly emerged from pessimism are doing especially well.” They reset their environment to zero, stop complaining, accept conditions as objective, and return to the basics of entrepreneurship, innovation, and learning. Once their state improves, their entrepreneurial performance improves quickly. Those who attribute everything to external factors run into trouble.
• The second signal is more important: a wave of young AI founders emerged last year. “If not for AI, this wave of people born in the 90s would have been suffocated.” Previously, the prominent founders born in the 90s were concentrated in consumer businesses, including the bubble-tea entrepreneurs, who also performed very well. Now AI technology has caught a powerful tailwind. “China’s entrepreneurial soil blooms at the first bit of sunshine.” The momentum is unstoppable.

44. Starting as a union officer: the methodology called three-person teaching

• Why did a mathematics student with a graduate-school recommendation go into HR? “I wasn’t talented enough at mathematics. I didn’t join Motorola because I studied mathematics; I joined because I was student-union president.” His starting point was almost comically low: a union officer. Motorola was a non-union company, but China required foreign companies to establish unions, so it invented an employee service committee. To Chinese executives, it was the union; to visiting Americans, it was employee self-organization. “My professional starting point was an officer in that organization.”
• The foreign company gave him a lesson that shaped his life. A senior vice president taught a learning method called three-person teaching: teach what you learn. Then move “from thick to thin”: reduce a mass of knowledge to mental models and frameworks; and “from thin to thick”: add your own examples and material. “That sentence influenced me for life.” His style was set from then on: he never transmits source material intact, but reduces it to a simple framework and finds new cases. “Find the first principles of the AI era and use the latest cases to teach them. Exactly the same.”

45. The life script: always among the leaders of the field, “I am immensely grateful”

• Another pattern in hindsight: he changed industries repeatedly, but the institutions were always among the top 3 in China at the time—Nankai, one of the best schools north of Tsinghua and Peking University; Motorola, at the height of foreign companies; Sohu, one of the 3 major portals; Ku6, one of the 3 major video sites and, according to the opening introduction, the first to list; CEIBS, one of the best business schools; and Chaos, one of the best in mobile internet paid knowledge. “Now I have entered AI and teach entrepreneurs of the AI era.”
• His conclusion is gratitude, not self-congratulation. “I would never think I was especially impressive. If life has a script, it is an excellent script. So much changed on the surface, but what remained constant was staying with the tide of the era and continuing to grow.” His only regret: he wishes he had attended a liberal-arts college, with its Platonic tradition of reading the classics; he lacked that early training and only began seriously tackling books after joining Sohu. His reading came in 3 phases: Chinese classics; philosophy, physics, and biology after returning from Stanford; and books on inner growth after 2019, such as The New World. Management books had little impact on him—like Drucker, he says, who did not read management books but read Dickens.

46. 张朝阳 as benefactor: from HR to editor-in-chief in 10 months, and a 0.86 stock option

• He joined Sohu in 2000 ostensibly to bring foreign-company HR practices into China. “In fact, HR was just the pretext, the stepping stone. What mattered was entering the internet.” He was not mature enough to analyze growth rates like Bezos; he simply wanted to enter a new field, with compensation “more than double.” He names 3 benefactors, all agents of paradigm shifts who moved him onto a new path: 柯先生 at Motorola, who bridged him from academia to a career; 张朝阳, who moved him from HR into the core business; and 奎尔奇, who moved him from business into academia.
• The layoff years produced 2 memorable details. He cut headcount from 600 to 400, and his English name Kevin Li earned the nickname “Layoff Li”: “Since this Kevin arrived, 3 people leave the company every day.” When the stock fell to 0.86, he advocated issuing options broadly, holding meetings with each department to rally staff. “I said you would become a millionaire in the future—and added one thing: US dollar.” The options later helped many people, including himself, earn their first pot of money. At 28, his energy was extremely high and he felt no fear. Later, laying people off became harder: “Empathy arrived. You no longer see numbers; you see people.”

47. Forty days of nosebleeds: the outsider editor-in-chief’s redesign playbook

• “No one at Sohu dared to take this job.” Sina’s 陈彤 was far ahead, but 张朝阳 gave the role to someone who had spent only 10 months in HR. “I studied mathematics; I didn’t understand Chinese, news, or computers. The key was that I had to dare to take it, and I did.” For 40 days after taking over, he learned headline writing and special-topic production from editors, then had the marketing department take him to visit media professionals. He worked so intensely that “what I thought about in dreams was better than what I thought about during the day.” He completed the redesign in 40 days, and the new energy was visible. “I had nosebleeds every day.” It gave him enormous confidence for the rest of his life: you can enter a field and learn it.
• His strategy had 2 parts. On speed: “Fast wasn’t my invention; Sina was fast. If you are fast, I will be fast too.” He competed head-on and poached Sina’s people. On differentiation: “陈彤 didn’t express opinions; Sina did not express opinions. Media needs opinions.” He learned that from media professionals, and Sohu began building around viewpoints. More important was bringing management into content: using the framework he had learned at 23, each channel made quarterly plans, teams competed publicly, employees voted, and bonuses and options followed results. “If you want the best return, you have to present your best work so others can learn it—internal knowledge flow, healthy competition that builds cohesion. I used management to raise Sohu News. I was never the best media person, but I was a good manager then.”

48. With 陈彤: from “he hated me and I hated him” to the first meal after leaving

• He does not sanitize the hostility of the competitive years. “The relationship was terrible. He hated me, and I hated him. I saw only him; he saw only me. He told others that I chased him so hard ‘the air from my panting had already reached him.’ But I respected him most of all.” After 李善友 left Sohu, 陈彤 was the first person to invite him to dinner. They later became close friends. 李翔 identified the mechanism: “Because you were no longer competing with him.”
• Their later trajectories diverged. 陈彤 barely changed direction and stayed on an extension track; the people who had worked with 李善友 in HR were probably still in similar roles, while he had changed course several times. Compared with the portal editor-in-chief generation—李勇 later went to Yuanfudao—he described his own profile this way: “My strength is management, framework-based management. My weakness is that journalism itself is not my talent or the vehicle for my expression. I knew while doing it that it would not be what I did forever, but I know I will do what I am doing today until I die.”

49. He could not see Toutiao: past experience can be a negative for new things

• How did a former portal editor-in-chief view 今日头条? His answer was candid and self-deprecating: “I didn’t see it—not because I looked down on it, but because I couldn’t see it. After I saw it, I couldn’t understand it. Before Toutiao succeeded, I had no idea.” 李翔 added that Toutiao had sent investment offers early on and the portals largely rejected them. “Everyone had their feet planted in the PC internet. 张一鸣 was native to the mobile internet paradigm. Because it was a native paradigm, no one understood it.”
• His characterization of 张一鸣: “From the beginning, this guy knew exactly what he wanted to do.” 张一鸣 said he had thought through the entire plan and drawn the map in the first few months before starting. He was “an extremely rational, cognition-driven entrepreneur.” 张一鸣 was the new paradigm of that era; 梁文锋 is the new paradigm of AI today. 李善友 closed with self-deprecation: “Looking back, I was definitely not the best operator.”

50. Ku6: entering without knowing YouTube, and fighting a video war with portal logic

• The origin story of Ku6 is almost a joke: “I didn’t know YouTube. It was an idea from a partner. Pretty stupid.” More than 300 video sites launched in China in 2006. When he chose the sector, he did not know it was the hottest one; he later found it was too hot, excessively so. “If I had known it was this hot, I would not have chosen the field.” Surviving was a near-death experience. “That period was brutally hard.”
• His strategic error was even clearer. 周鸿祎 had recommended The Innovator’s Dilemma, which teaches asymmetric competition, but he did not read it. Ku6’s slogan was “Whatever you have, I will do better; whatever you don’t have, I have too.” Competing on all content was pure portal logic, and importing it into video was completely wrong. Tudou was the most distinctive of the group and looked more like YouTube; he and 古永锵 (Victor) were both heavily shaped by Sohu’s portal model and ended up looking alike. He later taught asymmetric competition with particular intensity because he had made such a serious mistake. “If only the 李善友 of then had heard the courses taught by the 李善友 of today.”

51. Selling to 盛大: “getting ashore first was ancestral virtue,” but the endgame was doomed

• Ku6 sold to 盛大 in 2009, becoming the first of the 3 major video sites to lose its independence. His review has 2 layers. At the surface, “the essence was a game of money; the core competitive advantage was money—bandwidth and copyrights.” Users had no loyalty after a title was licensed: they came when you had the show and left when you did not. “There was no competitiveness at all. It was painful.” At a deeper level, the paradigms were mismatched: trying to move television onto the internet did not work. “Mobile and video are a natural pair; PC internet and video are not.” Douyin only solved that 10 years later. Almost none of that generation’s companies survived. “I was able to get ashore first through ancestral virtue.”
• Even the counterfactual was blocked off. 陈天桥 wanted UGC short video; 李善友 wanted licensed blockbusters. Both were wrong—UGC did not work until 抖音 and 快手, while every PC-era trick failed. The only path was iQiyi showing movies, but “even being right would have been wrong; it was not a good match.” 李翔 summarized the mirror image: if success was driven by the era, failure may also have been driven by the sector or the era. 李善友 said he had never viewed it that way: even if he had heard the course back then, the best choice would have been to walk away. “Putting something down is not giving up,” and his state of mind would have been better; he would not have had to define himself as a failure.
• He also remembered the pain of the sale. It was still the classical startup era, when founders felt responsible for the company’s fate. Selling was painful, but “leaving at the end was the most painful part”—like losing a piece of one’s own body. His definition of failure: “There were too many things we had no choice but to do. I dislike that feeling of being forced.”

52. The inheritance of failure: the theory was wrong, but believing there was a theory was right

• Ku6 became a philosophical parable in his life. He founded the company to practice his Chinese-classics management theory. The theory was useless and went into “the garbage bin of cognition,” but the mistake did not matter because “believing that there is a theory behind entrepreneurship mattered.” It was like Thales saying water was the source of all things: wrong, but foundational to philosophy because the claim carried a philosophical instinct. “No road is wasted.”
• During the 6-month gap, he dropped everything instrumental and read 王阳明. “Then I received the call—just as in 2020. When you let go of instrumental motives, something emerges.” He turned down investment opportunities twice. “Fortunately I did. I am not good at investing and still wouldn’t be today.” Starting another company would have been like going through primary school and then going through primary school again. “Return to the main line.”

53. The flywheel of the CEIBS entrepreneurship camp: the lecturer became a student

• Professor 周东生 had only “a 1-page idea”; 李善友 made it real. The most interesting mechanism was self-reinforcement. Beginning with the third cohort, speakers became students. “罗振宇 was originally someone I invited to teach. He said, ‘No, I want to come and be your student.’ He said, ‘This represents the trend. I will have to come eventually; earlier is better than later.’” 俞敏洪 would not normally teach at the CEIBS EMBA, but came to teach as a favor and then said he wanted to enroll: “I want to be around young people; I want to learn the internet.” 托布花 came the next year as well, in what was already the final CEIBS cohort before Chaos.
• The format itself was new. Traditional programs were EMBAs and executive classes; there were no dedicated classes for founders. They did not teach old cases, but live cases unfolding in real time, the students’ own cases, taught by the people involved. The atmosphere was rebellious and piratical, and each cohort improved on the last. He distinguishes Chaos from CEIBS: the online to C business was a completely new paradigm, and the financing BP referred to the online business; his class was a continuation of CEIBS. As for the title of “president”: no one called him that. They called him professor, and he had no attachment to being president.

54. That decade, and the transition from an entrepreneurial era to an innovation era

• He gives the previous decade the highest possible ranking: “It was the very best era for Chinese entrepreneurs and merchants in the past 2,000 years.” Without the foundation built then, today’s high-tech and AI entrepreneurship would not exist.
• But the keywords of the 2 eras differ. The core keyword then was entrepreneurship; today it is innovation. “Entrepreneurship is no longer 10x today, but innovation is 10x.” In the past, the café next door would always have people discussing fundraising. Who does that today? The angel-investing industry has practically disappeared. Innovation does not necessarily express itself through starting companies; today’s entrepreneurship increasingly has to be completed through innovation and technology. The bar is higher, the channels narrower, and the odds smaller. Which era does he miss more? “Actually, I miss that era more.”

55. The most puzzling question: cognitive levels explain why BAT never went global

• Looking back, 2 questions still surprise him. Why did 张一鸣 have such a large advantage over the previous generation? Today, why does Alibaba find it so difficult to compete with ByteDance, whose momentum is so strong? And why did BAT fail to internationalize while ByteDance and Pinduoduo did? His framework draws on Kant’s idea that humans legislate nature: the deeper your cognitive level, the larger the world you can see. “Experience and intuitive cognition correspond to agrarian civilization; logic and rational cognition correspond to industrial civilization.” After Newton, the world could be derived from equations—logically valid and bound to occur in practice.
• Applying that framework—he admits it may be unfair—the PC generation was fundamentally experience-based: “There is this model, so I’ll do it here.” 张一鸣, 黄峥, and 王兴 read extensively and believed in logic. “The mobile internet generation stood on logic, which was a dimensional strike against the first generation.” Why did the first generation fail to internationalize? The mainstream world was industrial civilization; attacking uphill from agrarian civilization was too difficult. ByteDance and Pinduoduo were operating on level ground and could move across. The most important tide of the past decade was cognitive upgrading—from experience to rational logic. “I never spoke with 张一鸣, but he practiced cognition; I taught cognition. Both of us were pushed by the tide.”
• 李翔 offered the journalist’s additional explanation: BAT defeated global technology giants at home, which shaped their cognition. When going abroad, they first modeled how to compete with local founders, concluded their odds might be poor, and shifted toward investing in local companies. 李善友 accepted the point but insisted the abstraction came first: the specific reasons were merely projections of the deeper cause. “You use the technology of commercial and industrial civilization, but seem to be wearing an invisible long gown and mandarin jacket.”

56. 3 years of self-doubt: “Did I become someone asking why people don’t eat meat?”

• After the post-pandemic startup boom receded, he underwent severe self-doubt. Founders’ core demand had become survival, while he was teaching cognition. “Had I become someone asking why people don’t eat meat? Teaching first principles—was this dragon-slaying technique?” He seriously questioned his value for 2 or 3 years and suffered deeply. “Was I really supposed to help you with growth? I don’t know how. I can’t do it.”
• He emerged through an external and an internal catalyst. AI rose last year and created momentum for technology entrepreneurship; what had been non-consensus became consensus. Second, he understood that “it was not wrong to explore first principles; I needed to keep going deeper”—from epistemology into ontology, helping people find something more invariant within a faster-moving AI era. He compared it to reform: when problems emerge, some ask whether reform was a mistake; his camp believes reform must continue to solve the problems. The struggle was not wasted. The content explored may have been wrong, but exploration itself was right. Without it, the new tide could not have expressed itself through him.

57. From certainty to doubt: cognition alone no longer works; you also need inner force

• Asked what he once believed without question but now doubts, he pointed to Chaos’s foundation. For the past decade, everyone believed that cognitive upgrading could solve every problem, and nearly everything he taught was about cognition. “Today I feel cognition alone no longer works. After the pandemic I discovered another word: inner force. You can have high cognition, but without inner force you cannot move your legs. Cognition cannot function.”
• He defines inner force directly: “energy of consciousness.” That also explains the destiny of Philosophy World. “I finally understood why I made that course. It came for this thing called inner force.” Cognition is the base; it must be combined with inner force and with elements related to consciousness. “That is my biggest change.”

58. The frozen-pear moment: exposing vulnerability saved the organization

• From 2022 to 2024, he also ran out of inner force. He was going through darkness while still teaching and running the company, and both his physical and psychological health suffered. “I was like a frozen pear from Northeast China—I had frozen solid. I was frozen, the organization was frozen, and the era was frozen too. It was a complete miniature of the era.” In October 2023, he brought 36 core members together for a 3-day “exploration flow” and chose total exposure. “The greatest fear of being a teacher is other people’s judgment. I told them: I am afraid; I have no idea what to do; let’s figure it out together. This saved me and saved the organization.”
• 2 lessons remained. First: “Fear itself is not frightening. Your fear of fear is what is frightening.” It is like waiting for a scorpion to fall from the ceiling—once it falls, you will simply go to sleep. Second: receiving criticism feels like being pierced by thousands of arrows, but afterward it feels good. It pulled him off the altar and broke the structural arrogance. Pretending to be brave while lacking power, with everyone aware of it, is the worst state. “I chose courage.” The experience also taught him how to teach: first be brave, then expose vulnerability. It gave him empathy. In his late 20s, he would see someone without inner force and think there was something wrong with them; after experiencing it himself, “you develop tremendous empathy for people.”

59. Seeing people instead of just courses, and classical thinking in AI use

• Another realization from his darkest period: “I used to think the course was my life, my entire world. I made a huge mistake: I did not see people.” The keyword is now people. The course is a vehicle, and it serves people. He is willing to have long conversations with people even though the ROI is low. “I now think it is the only thing that matters.”
• On AI, he describes himself as a refuser and a classicist on core questions. He thinks independently first and turns to AI only to discuss details while preparing material. When pursuing ultimate questions, he prefers independent thought. “Long, difficult stretches of independent thinking are rewarded.” If something arrives too quickly, it may block the larger current from emerging. 李翔 agreed: people still need friction, resistance, and even pain; he often suspects that anything too smooth is fake.
• This year’s Spring Festival provided evidence. He secluded himself for more than 1 month to prepare a new course, moving from cognition to ontology and then using that foundation to explain large models. He woke at 5 a.m. every day to study philosophy, ontology in the morning, and evolutionary theory alongside large models in the afternoon. His mind was foggy and the structure would not form. On the 4th day of the lunar new year, he flew to Dubai. The poor lighting on the plane made reading impossible, so he tried building the slides. “Suddenly the inspiration surged out, and I produced 100 or 200 pages of PPT at the hotel.” Without the month of difficult, foggy exploration, the later material would not have emerged.

60. From peak to valley: frequency holder and “I’ll accompany you to hell”

• The 2 groups he is most curious about now are scientists and people born in the 2000s. The latter shook him: they are relaxed, playful, and find fun and interest important. “Those words never mattered to me. I was too serious.” One suggested buying a blind box. “That is the beginning of becoming young.” In one-on-one conversations, he listens more. “He can start from the day he was born, and I won’t get bored. As long as it is real, I don’t get bored.” A concrete person’s story is no less valuable than an abstraction. He used to enjoy abstractions more; now he sees beauty in a person’s vulnerable places.
• His language of leadership has changed completely. He once wanted to be a mountain peak, admired from below; now he thinks the valley is better. The Tao Te Ching says, “For learning, add every day; for the Tao, diminish every day,” a passage he never understood before. Everything is frequency, and energy of consciousness is frequency. He is willing to be a “frequency holder”: moving from form to formlessness, because only “nothing” can support others. In deep accompaniment work with students—the Black project, involving 6 or 7 companies where founders bring 6 core executives into workshops and bet on a double spiral of personal and organizational growth—his posture has moved from strict father to compassionate mother. If a student takes on too many things and lacks focus, he used to say they were wrong, intervening too deeply in their causality. Now: “If you want to go to hell, I’ll go to hell with you.”
• On ego, he offers a denominator theory: ego is not the problem; the question is how large the world behind it is. “Musk’s ego is huge, but how large is Musk’s world?” An ego without boundaries can mean working for the universe. A feeling of being a conduit, humility, and enormous capacity to support others is a beautiful state. At this age, becoming more arrogant and taking up more space means failing to understand life.

61. Termites don’t eat: optimism in the AI era

• He is optimistic about the relationship between AI and humanity because of a story from Honduras. Termites arrived and left an area seemingly barren, which looked terrifying. But locals explained: “Termites do not eat living trees. They eat trees only after the trees lose their life.” The insight came like a flash. “That is how I understood AI. I coined the phrase ‘termites don’t eat.’”
• His core warning and promise for the era follow from it: AI is built on rationality and logic. “If you put your roots in rationality and logic, AI will eat you.” If you evolve into something living—with spirit, consciousness, and soul—AI will not eat you; AI becomes your tool. Being “eaten” does not necessarily mean physical destruction. It means you lose your meaning. If you remain on the old scale, you become the person in The Matrix supplying raw material to AI. AI is therefore his new version of God’s phone call. For a small minority, it is a whistleblower saying that a paradigm shift is required: evolution in consciousness. Viewed from inside the system, the problem cannot be solved; move to a larger level and the era looks beautiful. He does not deny that AI may solve organizational problems. When 李翔 said young CEOs could use AI to fill organizational gaps without taking the long Huawei or Alibaba route, he replied: “Of course. The management paradigm in the AI era is facing a massive shock.”

62. The kindest thing: someone always builds a bridge, and the gaze remains sincere

• Asked about the kindest thing anyone had done for him, he gave 2 answers. First was the bridge. “My life has been discontinuous. At every discontinuity, someone appears and helps me turn. Different people have accompanied me through different stages. I am extraordinarily lucky and extraordinarily grateful.”
• The second was tolerance. He never withholds anything in his classes: he gives students all the PPTs and videos and does not claim knowledge rights. But in hindsight, “many things I taught were wrong because my cognition was insufficient.” He is deeply grateful that his students still look at him and listen with the same sincerity. He soon iterates or even repudiates himself, but at the moment he teaches, he believes what he is saying. “If I didn’t believe it, I could not teach it.”