China–US Tech Elites’ Minds: 连长 and 养鸡 in Conversation
Summary
Silicon Valley’s tech elite appear “happy, confident and full of drive” because of more than personal temperament. 连长’s long-form reading of Naval, Elon Musk and A16Z investors—alongside hours-long interviews—left him with the impression that they can discuss politics, history, energy and health, while Chinese founders often speak cautiously only about their companies and products; this “dark forest” silence reduces the chance for ideas and experience to compound positively.
The biggest sampling bias in China’s startup narrative is that failure becomes worth discussing only after a comeback. When 连长 described 连咖啡’s severe losses in its competition with Luckin Coffee while the company was still in the “ICU,” founders in the audience quickly lost interest because the story had “no ending”; once the company recovered, the same experience could become inspirational. “If you don’t make a huge comeback, what’s there to say about your failure?” means public case studies tend to omit the reality of crises as they unfold.
Content is becoming a product, a feedback mechanism and an organizational lever, rather than an attachment to corporate promotion. Learning in public and building in public connect learning, development and user correction into a closed loop, while Howard Schultz’s books at different stages of his company provide a model; 连长’s view is that “content and product cannot be separated,” and tools now let teams handle topic selection, information organization and opening-retention tests from day one instead of bolting on content capabilities later.
The core opportunity in video podcasts is low cost, high commercial value and rapid iteration, but celebrity guests are not a moat. In fragmented consumption, a film costing RMB500M and a low-cost podcast may ultimately compete through one-minute clips: “The whole playing field is clips competing”; if 70% of a show’s value comes from its guests, the host’s distinctiveness, judgment, relationships and shared memories—as well as a clear choice between “interview” and “conversation”—are what can become durable assets.
Chinese tech founders’ intellectual resources are a “stitched-together monster” combining Silicon Valley methodology, revolutionary and struggle philosophy, and homegrown science fiction. 连长 is especially impressed by The Selected Works of Mao as a strategic text: meeting minutes and telegrams could state problems, solutions and methods of observation with extraordinary precision, coordinating large-scale action through writing; meanwhile, Naval’s blunt discussion of “wealth and happiness” exposes the different tolerance levels in Chinese and US culture for “street-level ambition.”
The Doubao phone demonstrates not a product path that has already been proven, but one possible direction for future AI interaction. 李翔 uses AI dozens of times a day instead of search, and directly waking the phone produces the feeling that “this is how it should work”; Agents scrolling through lists and calling apps remain inefficient, but “it’s doing the hard work, not you,” while also exposing redundancy in traditional app flows and product design.
Recommendation feeds are dismantling the infrastructure behind shared hot topics and shared celebrities. One person sees the Nanjing Museum incident while another’s world consists of Jinsha’s proposal and Sniping Butterflies; as “news” ceases to be a distinct category, business figures are increasingly unlikely to achieve the mass recognition once delivered by centralized media, leading 李翔 to make a radical judgment: “Lei Jun is the last celebrity.” Startups also find it harder to build nationwide awareness through centralized media alone.
Deep dive
1. “Happy, confident and full of drive” is first an ecosystem-of-expression issue
连长’s question after returning from Silicon Valley did not come from a single trip. He has spent years reading A16Z’s long-form articles and watching hours-long interviews, even saying, “I generally don’t open a podcast if it’s under 3 hours,” because when information density is low, total volume still matters.
The overall impression left by this material is that Bill Gates, Elon Musk, Jensen Huang, Naval and Silicon Valley investors can move from business to political systems, democratic evolution, energy, health, history and the humanities, explaining the causal chains behind questions that appear unrelated to their core work.
Comparing this with an interview with a Chinese entrepreneur made 连长 more clearly aware of differences in the scope and manner of expression. But he retained an important doubt: could this simply reflect sample selection, or “me turning my hopes into reality”—projecting his expectations into an apparent China–US difference?
2. What China’s investment circle lacks is not winners, but people willing to define problems publicly
连长 uses Paul Graham as a reference point: his blog is not merely well written; deliberately or otherwise, it defines, guides and even sets the boundaries for how Silicon Valley founders think about problems. 连长 asked multiple investors whether China had an intellectual leader of comparable importance. The answer was essentially “no.”
One explanation is that the social atmosphere encourages people to suppress certain parts of their nature. Even if they can continue thinking privately, the absence of public expression and feedback can itself inhibit public expression; combined with the widespread pessimism of recent years, many people retreated behind the scenes and stopped discussing even their own work.
养鸡 uses The Three-Body Problem’s “dark forest law” to describe this environment: once someone exposes their coordinates through a 4-hour podcast, as Marc Andreessen or Naval might, they may invite attack. Zhang Lei also faced questions after publishing Value: did he perfectly live up to the values he made public? 连长 said he did not know whether Zhang had, but believed more people should be encouraged to speak publicly.
3. Five thousand years of history are both accumulated capital and potentially a “civilizational time debt”
连长 cites Zhou Hang’s Rethinking Entrepreneurship: the book is not about how to win, but how failure happens and what it leaves behind. Years after publication, readers still repeat particular passages to Zhou in real-life encounters, suggesting that speaking openly about failure can indeed generate long-term compounding returns.
But all 3 participants also acknowledged that East Asian culture often avoids both failure and death. Parents aging and life’s accidents are things everyone will face, yet when either topic arises, the conversation often stops immediately—it is “as if the words won’t come out.” Even the most optimistic and intelligent people often do not know how to discuss them lightly.
The discussion moved forward through Ning Gaoning’s formulation: “the consumption of 5,000 years of history.” An ancient civilization provides resources, but also gives questions of family, clan and mortality a heavy inertia; by comparison, the US, as a younger culture and nation, “is walking without carrying anything on its back.”
连长 remembers that when the internet was rising, founders born around the 1980s—Wang Xing, Zhang Yiming and Huang Zheng among them—felt “refreshingly new”: young, clear, familiar with management theory and full of energy. 李翔 said everyone has gradually become more “mature and steady,” which he believes reflects the combined effects of cultural foundations, institutional conditions and social discipline.
4. Failure is accepted by mainstream audiences only after it is attached to a happy ending
Media organizations have long wanted to tell stories of failure, but repeatedly encounter resistance from both sides: consumers are deeply biased toward success stories, while the people involved face psychological barriers. Unless the failure is followed by “great success,” or the old story helps the next startup, neither side usually has much incentive.
连长 shared a postmortem at 极客公园’s startup camp on 连咖啡’s fierce competition with Luckin Coffee: how the mistakes happened, how major losses sent the company into the “ICU,” and why they had to “cover the wound first and keep it from getting infected,” quietly waiting for the company to recover rather than rushing to prove a comeback.
At the time, the company was still in the ICU. “When there’s no ending, there’s no ending”—only to be continued. The new founders in the audience, initially energized, were visibly losing interest by the second half because the story had not yet turned around. 连长 believes the same story would probably attract far more interest now that the company is running again.
养鸡 proposed a distinction worth preserving: those new founders may only have wanted to learn methods and had not yet entered a real crisis; people already in crisis may instead be more sensitive to this kind of material. The problem is not that failure contains no value, but that the transmitter and receiver are often at different stages.
5. Real startup stories usually become complete only after the storyteller has left the game
李翔 uses The Hard Thing About Hard Things as an example: Ben Horowitz writes about facing investors, partners and colleagues during the company’s most painful moments, as well as making layoffs; the details are “extremely painful, and extremely real.” The company may eventually sell for around $1B—not a Silicon Valley-scale triumph—but that does not make the experience any less useful.
Horowitz also gives a realistic constraint: “When you’re still in the middle of the game, you can’t tell people the whole story.” If a CEO publicly disclosed every difficulty, investors and colleagues would object. Only after selling the company and becoming an investor could the old story stop damaging the outcome at the time.
连长 admits that even though he likes open expression from others, he imposes taboos when it comes to his own problems and discusses only what fits the laws of corporate development. Genuine feelings are not necessarily suitable for real-time publication. “Being inside it” is itself a constraint, and making content becomes another life lesson.
6. Content and product have evolved from separate functions into a single feedback system
连长 and 养鸡 spent months discussing the relaunch of the “Estuary” video account but never actually recorded anything. The new insight from the discussion is that for founders trained as product managers, “content and product cannot be separated”; content is no longer packaging applied after the product is complete.
养鸡 cites Howard Schultz: he wrote a book when founding the company, another when it reached a certain stage, and continued writing during his exit and return—effectively using content to drive the company continuously.
Silicon Valley’s popular practices of learning in public and building in public turn every step of learning and creating small AI products into a public development log. Public exposure brings feedback, and feedback is then used to correct the product; sustained content practice becomes a lever where “1 + 1 is greater than 2.”
连长 defines content directly as leverage: a film company may not receive a high market valuation, yet still leverage enormous influence and resources; Disney has used leverage at every step, from online IP to offline experiences, clubs and theme parks.
7. Tools let content teams be fully equipped from day one
养鸡 stresses that video cannot simply be audio with a camera added. Video viewers need a more focused, more thematic experience, and long-form video must also generate clips; once real cameras are present, people instinctively look for “the more flattering side,” and the sense of performance can end up damaging the conversation.
养鸡 wanted to start with book discussions. 李翔 suggested using tools to track hot topics, organize information and methods, and train the opening 3 or 5 seconds so users do not swipe away: “Come out fully armed,” then adjust the toolset.
Running an account like a commercial account does not mean being obsessed with profit. It means continuously correcting content and production methods against training standards. When performance is poor, creators cannot casually blame “the platform not giving us traffic,” because that answers neither the content problem nor the platform problem.
8. Video podcasts’ cost structure is well suited to fragmented consumption
连长’s view is that video remains the strongest, most mainstream and most commercially valuable medium. For platforms, video can be a relatively low-cost way to produce large volumes of content, creating a sharp contrast with traditional television, film and documentaries.
李翔’s experience working on Tencent Video’s documentary The Age of Surging Currents shows that professional teams can make excellent programs, but production costs are genuinely high. As economic growth slows, fewer people are willing to carry those costs continuously; he believes this is part of the pressure on long-form video platforms and the film industry.
The more fundamental change is on the consumption side. Producers once advanced stories according to a complete narrative design, while audiences now consume them in points and fragments. A film costing RMB500M and a low-cost video podcast may therefore both end up competing through 1-minute or 30-second clips: “The whole playing field is clips competing.”
李翔 recalls that when early technology media monetized through advertorials, executives at large companies asked, “Who would watch this?” The answer was that users watched anyway, because when supply is abundant, they quickly pick out what is useful. Creators can insist on niche choices, but must also accept the question of whether they want to enter the current.
9. The challenge for video podcasts is not booking celebrities, but turning guest traffic into a moat
The most obvious strategy is to invite public figures, but when celebrities promote a film or product, they run through multiple channels and the content is “not very different” from one place to the next. Hosts and guests who do not know each other also struggle to create the relaxed atmosphere a podcast needs, while body language can magnify awkwardness.
The long conversations between Chen Luyu and Dou Wentao offer another path: a guest can appear on many shows, but the pair’s years of shared memories cannot be replicated. The problem is that very few people are both publicly relevant and connected to a host by deep shared memories, making this advantage difficult to reproduce.
Low production costs allow a show to test at greater volume and higher frequency, so “quantity may eventually pile up into quality.” But a professional content team still has to answer a sharp question: if 70% of a show’s importance lies with the guest, what exactly is the host’s moat?
10. “Interview” and “conversation” are 2 different products, and traffic guests push shows toward the former
A veteran who had worked on programs for CCTV and Phoenix TV once divided content into “interview” and “conversation.” An interview has a host and a guest: the interviewee is the protagonist, while the questioner provides contrast and draws things out. A conversation has no protagonist; its value comes from the collision of equals exchanging ideas.
When The Crosstalk invited Jiang Wen, who was rarely interviewed and came to promote a film, the show shifted from “conversation” to “interview,” and its reputation suffered. Both formats can work, but the more a guest resembles a public figure, the more the host is forced to sustain the show through a sequence of questions.
李翔 believes interviews with public figures must deliver clear public value. If they merely reproduce promotional content, differentiation is weak. The distinctiveness of a conversation, by contrast, can come from the host’s own thinking rather than placing the entire burden on the interviewee.
This discussion directly changed the positioning of Chain Reaction: it would continue to focus on “conversation,” even if that simply meant “finding different friends to come chat with me,” rather than forcing its way into an already crowded interview track.
11. Chinese tech CEOs’ intellectual resources form an effective but complex “stitched-together monster”
李翔 mentioned an article discussing Chinese tech CEOs’ reading lists and spiritual anchors. Its summary broadly matches his observation: business methods and the logic of technological evolution come from Silicon Valley, while Mao Zedong’s methodology, revolutionary and struggle philosophy, and homegrown science fiction such as The Three-Body Problem are layered on top.
While Silicon Valley founders often discuss Dune, Snow Crash, The Hitchhiker’s Guide to the Galaxy and Foundation, Chinese founders once recommended The Three-Body Problem and the dark forest at high frequency. 李翔 calls the final product a “stitched-together monster,” also summarizing it as “making steel through primitive methods, then shocking the world”—and believes it has worked reasonably well.
连长 was struck, after seriously reading The Selected Works of Mao, first by the idea of “governing a country through articles.” Much of the material is essentially meeting minutes or telegrams, yet it can lay out the problem, the solution and the method for observing results in sequence; a single article becomes an executable strategic program.
He is particularly impressed by the precision of the language. Battle plans had to be understood without ambiguity by different armies so they could “move in unison and form an enormous collective force.” The Selected Works also contains few very long pieces and can be opened and read anywhere, which reminded him of The Almanack of Naval Ravikant.
12. Whether “street-level ambition” can be stated plainly shapes the public image of entrepreneurs
连长 believes the best advertising line for The Almanack of Naval Ravikant is its subtitle, “Wealth and Happiness,” because those are the 2 things ordinary people want most. Americans do not avoid talking about wealth and worldly success; in China, even when people feel the same way, saying “my dream is to make money” publicly can seem too vulgar.
Wang Jianlin is therefore an example 连长 particularly admires: “a small goal” connected wealth and the individual through a nationwide internet meme. Even when the context was clipped, Wang himself was indeed direct enough. 连长 also mentioned the view that Wang had openly expressed his ambition for wealth before various rankings elevated him to the status of China’s richest man.
The discussion then corrected the simplistic China–US binary. American entrepreneurs also face limits on what they can say, though observers may not feel them; and the company’s stage changes the task of communication. When a company remains in a life-or-death competition, it is entirely reasonable for its CEO to focus cautiously on the company and product.
The difference still appears in the language of competition. Phil Knight could say that he woke up wanting to “throw rocks at Adidas,” while Chinese entrepreneurs usually call competitors “friendly peers” despite fierce competition behind closed doors. The discussion also mentioned “dancing in shackles” and the line that “the audience doesn’t care whether you’re wearing shackles; it only cares whether the dance looks good.” 连长 once asked whether the latter was a quote from director Zhang Yimou, but the conversation did not establish its exact attribution.
13. Entrepreneurial circles favor history and nonfiction, while fiction is still treated as low-efficiency input
The most important intellectual resource among the entrepreneurs around 连长 is reading, with history taking up a particularly large share. Different people know different periods—Ming and Qing, pre-Qin, ancient China or Egypt—and historical material occupies a large portion of their conversations.
The rest of their input is concentrated in science, energy technology, the history of science, health and wellness, and philosophy. 连长 also places art within his broadly defined category of “natural science” or nonfiction. If 10 people read 100 books in total over a year, he estimates that fewer than 5 are likely to be novels, because many still instinctively believe that “reading fiction is a waste of time and has no use.”
连长 increased his fiction reading over the past few years, then partially returned to his earlier mix. The books 养鸡 recommended or gave away most often in a year were The Almanack of Naval Ravikant and Ted Chiang’s Stories of Your Life and Others and Exhalation—the former for its thinking on wealth and happiness, and the latter 2 as his especially recommended examples of fiction.
His reasons for recommending The Almanack of Naval Ravikant are specific: it is short and can be entered from any point, and much of it was compiled by fans from Twitter posts. Naval improved his writing by repeatedly refining 140-character expressions, while Ted Chiang and an Italian physicist on the book’s reading list overlap strongly with 连长’s own reading.
14. Input channels shape creators; willpower alone cannot bridge media
李翔 summarizes his intellectual nourishment as “reading books and meeting people.” He likes nonfiction and documentary literature written by Chinese authors and is also trying to understand reality and generational differences through sociology and related material. Discussing film explainers and short-form video, he refers to “sex-time ratio”—the return received relative to the time spent.
He once realized that most of the books he read were translations, so he deliberately increased his Chinese-language originals. Re-reading Wang Shuo, Wang Xiaobo and Shi Tiesheng in fiction was not difficult, but in technology and business, the English-speaking world does have a richer ecosystem of expression, creation and information exchange.
YouTube and the content reposted from it to Bilibili offer abundant material on almost every US tech company. 李翔 also mentioned a journalist-written book about OpenAI’s founding, though he was unsure at the time whether its title was AI Empire and had not read it; he also mentioned the documentary The Thinking Game. By contrast, the Chinese-language world still lacks an equally rich documentary about 梁文锋 or narrative about ByteDance.
李翔 worries that he is not a good video consumer and therefore may struggle to become a good video creator directly. Lei Jun advised him, “Love an industry first, then you can work in it, and then you can do it well.” He is considering deliberately watching 1 hour of video each day, just as he once created dedicated podcast-listening situations while riding the subway, flying, taking high-speed rail and walking.
15. The Doubao phone moves AI from an answer tool toward an operational agent
李翔 already uses AI frequently instead of search—“definitely dozens of times a day.” The Doubao phone makes the need clearer: users can wake the phone directly and state a question or task without opening applications layer by layer, materially reducing interaction costs.
Because many applications later stopped working normally, the phone at one point functioned almost like a smart speaker at home. But 李翔 still sees it as a trend. The specific product path “may not necessarily work directly,” but the direction has already been validated by the experience of using it: “After you use it, you feel this is how it should work.”
When an Agent calls an app, it still simulates a human scrolling through 7 or 8, or 10, lists, in a process that is equally inefficient and clumsy. The difference is simply that “it’s doing the hard work, not you”; users no longer have to intervene in every mechanical step.
This machine operation also acts as a reverse audit of apps. If product managers observe how an Agent completes a task, they can more easily spot unreasonable structural requirements and redundant steps, then ask again: “Do you really need the user to operate it this way?”
16. When “news” disappears as a distinct category, shared hot topics and shared celebrities disappear with it
李翔 now reads more nonfiction and documentary writing by Chinese authors to understand reality and generational differences. He mentions “sex-time ratio,” meaning that users measure not only price but how much return they can get per unit of time. The shutdown of 88.7 also made him feel that an entire youth-oriented media ecosystem was disappearing.
When asked where he gets news, 连长 even paused: “It seems there isn’t a news category anymore.” Short-video feeds, WeChat Channels and Douyin can place a dance video directly before a clip about the Russia–Ukraine war, while Xiaohongshu can narrow the frame to a lost dog on a nearby street. Information is no longer stored in the drawers of traditional sections.
Shared hot topics have therefore become unstable. Some people knew that the Nanjing Museum incident was being widely discussed but did not continue following it. The biggest topic 养鸡 saw was Jinsha being proposed to in Iceland by a boyfriend nearly 20 years younger, amplified by the popularity of the age-gap romance series Sniping Butterflies. Recommendation mechanisms show each person the world they want to see, but may also make “your world smaller and smaller.”
李翔 consequently proposed that “Lei Jun is the last celebrity”: nationwide celebrities depended on centralized mass media, and that media has disintegrated. 连长 responded that new celebrities could still emerge, but 王星星’s level of public recognition is far below Lei Jun’s; many people also do not know who the CEOs of Pinduoduo or ByteDance are. The companies remain enormous, but their leaders are no longer automatically public figures.