China OS vs. America OS
I had an unfiltered conversation with two tech investor friends who operate in both Silicon Valley and China
It's been a while since I've had a genuine "real talk" with technologists like Du Lei and Han Hua - savvy and thoughtful investors and trained engineers who earned degrees from institutions in both China and the U.S., built meaningful careers in Silicon Valley, and developed deep civic consciousness while staying acutely aware of China's unspoken realities.
I first met Han Hua on Clubhouse four years ago and connected with Du Lei through my podcast CyberPink, Lei's also a former colleague when I was working in an AI news company in Mountain View. But we truly got to know each other during that rare window when Clubhouse evaded the Great Firewall, creating an unprecedented space where tech professionals in the Bay Area could engage in unfiltered conversations about civil society, innovation, and technology with their Chinese counterparts. I was part of that fleeting scene. Soon, China banned Clubhouse, and the platform gradually faded away.
The brief window on Clubhouse left me with a strong impression. I saw how cross-Pacific conversations became both freeing and insightful when grounded in real experience. Those smartest Chinese technologists working in Silicon Valley, who often disappointed me with their pragmatic silence, suddenly opened up with insights or revealed their thought-provoking mental models when the environment was set right. This experience prompted me to invite both Du Lei and Han Hua—engineers turned Silicon Valley investors, and my friends—to continue this Clubhouse-style dialogue about China, Silicon Valley, and technology.
We covered:
China OS vs. America OS: Different operating systems for innovation
Chinese AI companies' open source strategy and talent mobility
Why China operates in the "definite pessimism" quadrant
The reality of AI coding tools and vibe coding in China
The possibility of thriving in both OS
Note:
This conversation took place two weeks ago (mid-July) when Trump's AI Action Plan hadn't been announced yet. Therefore, the discussion about open source strategy isn't as timely as it could be. The conversation was originally conducted in Chinese and later translated by Claude and refined by Notion AI.
Du Lei: Investor at Sancus Ventures, formerly cofounder of Huma Finance and early team member at Opendoor
Han Hua: Angel investor, former VC at GV (Google Ventures)
Afra: Du Lei proposed a very apt metaphor in a previous conversation: China and America are using different versions of operating systems. This OS can be understood as a combination of software and hardware. Du Lei pointed out that China has faster hardware updates, but has many problems on the software side. I think this metaphor is particularly fitting.
I'd like to start by having you both share your understanding of what constitutes China's OS versus America's OS. One interpretation is: America continues to rely on email and webpage systems for government services, while China has adopted the more efficient WeChat platform (where almost all civic services can be quickly completed). The hardware gap is striking: China's high-speed rail system represents the rapid flow of resources within its system, while America's infrastructure remains at a much older level. It's as if China has upgraded its hardware with several powerful chips, greatly accelerating data transmission, while America still operates at 20th-century speeds.
I've been reading an upcoming book, Breakneck (it's excellent, and please pre-order here), and I'm also writing a book review. The book mentions an anecdote: when the writer Dan Wang was at Yale Law School, he looked up how long it took to travel by train from New Haven to New York in the early 20th century. To his surprise, the train journey a century ago took exactly the same time as today—two hours.
China’s OS logic: definite but pessimistic
Han Hua: When I first heard this metaphor, I asked myself: What is China's OS and what is America's OS? I found it very difficult to summarize in simple terms, because whether it's "involution" (内卷) or development speed, both countries have areas where they're fast or slow.
Referencing Peter Thiel's framework, he identifies two dimensions in how societies view the future: certainty (definite vs. indefinite) and outlook (optimism vs. pessimism). Thiel argues that China operates with high certainty about the future while maintaining a pessimistic outlook, which significantly shapes its decision-making processes. In contrast, American society tends to be optimistic about the future but lacks a definite vision for how that future should unfold.
Based on these different expectations about the future, the two countries produce completely different decision-making logic. For example, if China's expectations about the future are both definite and pessimistic, it would conclude: future resources are limited, great power competition is zero-sum. If I don't compete, resources will be taken by you; if I don't develop well, you will lead. This expectation about the future directly influences China's political, military, economic, and technological policies.
But if you're optimistic about the future, believing the future is abundant, thinking everyone can get a piece of the pie, then you won't be so urgent. You'll think this is a positive-sum game, the future can continue developing, everyone can find their suitable position, with enough resources to meet everyone's needs.
I think China and America don't have such fundamental differences, but their expectations about the future have huge disparities. This disparity ultimately leads to different decisions with far-reaching impacts.
Why China is obsessed with technology sovereignty
Afra: Since you've been in China for a month already, what do you consider the most striking example of this difference? Is there a specific area where American tech circles maintain optimism about the future while China remains pessimistic?
Han Hua: Let me give the most direct example. China's overall industrial policy is built on self-sufficiency (自给自足). What does self-sufficiency mean? It means being able to produce everything domestically—complete industrial chains. Whether agricultural or industrial products, China aims to develop full industrial chains capable of producing everything. This requires having sufficient materials for all production, and when certain materials are lacking, finding substitutes, such as using new energy instead of oil.
You'll notice that China's industrial policy is founded on this assumption: in the event of future conflict, China cannot rely on partners and must be self-reliant. Therefore, China must achieve self-reliance (自力更生), which means controlling entire industrial chains. If you expect that conflict is inevitable, your decision becomes clear: you must master everything and control your own destiny. You cannot trust allies or any partners—you must ensure your fate remains in your own hands.
Trump and China, to some extent, influence each other, have pushed toward more pessimistic future expectations. This leads to industrial policy changes - "Made in America" means self-reliance, reducing dependence on global partners, producing more in America rather than relying on allies, China, or other supply chains.
Afra: I believe China's emphasis on self-reliance and technology sovereignty (技术主权) has deep historical roots. Unlike America, China experienced continuous foreign invasion, resulting in a profound national trauma—the "Century of Humiliation" (百年耻辱). For such a historically proud civilization, this period left an indelible mark.
I think both Chinese and American tech circles underestimate the impact of the Century of Humiliation. I believe many of China's current industrial policies directly stem from this historical experience. China fears that falling behind in critical technologies could lead to a repeat of that humiliation, with the country once again vulnerable to foreign powers or technological monopolization.
Even during Mao's era, China's dependence on Soviet technology became painfully evident when the Sino-Soviet split occurred. Soviet experts suddenly withdrew experts and resources, leaving China in what's now commonly called the "strangled by the neck" (卡脖子) predicament. This isn't simply about pessimism versus optimism regarding the future—it reflects judgments based on historical experience. China carries a stronger sense of crisis awareness and collective trauma.
As for the Trump administration's approach to technology development, while he takes an accelerationist stance toward AI and other technologies, he has significantly reduced national funding for fundamental research. Cutting funding to research universities hardly reflects optimistic expectations about the future. For private tech companies and industries, however, Trump 2.0 adopts an almost completely unregulated accelerationist approach that aligns closely with Silicon Valley's "just do it" ethos.
Du Lei: Trump's optimistic sentiment indeed varies across different policies, but he has brought a sense of certainty about the future to America, or at least provided a general direction. This certainty influences overall American policy, even including America's allies, producing extremely strong coordination effects.
I recently read a book called Boom: Bubbles and the End of Stagnation, which discusses how, in free markets, especially in democratic societies like America, bubbles are actually one of the most important methods for societal coordination. In countries like America where power is decentralized, top-down policy implementation faces many obstacles and can't be executed as efficiently from top to bottom as in China. But if you want to achieve coordination at the societal level, bubbles are a particularly effective mechanism.
Although bubbles waste capital or talent, they can attract funding, attention, talent, and energy in a short period of time. For example, if everyone believes that we must land on or colonize Mars in the future, then the entire supply chain, all participants, venture capital, and even government policies will invest in that direction. So the book argues that bubbles aren't entirely bad, especially innovation bubbles, as they can promote coordination in technology and throughout society, creating an accelerating effect.
From this perspective, you could say he has brought a "Trump bubble" to the tech world—regardless of whether Trump himself is pessimistic or optimistic, he has brought a kind of certainty about future expectations, thereby producing coordination effects at the societal level.
China’s tech meme: Wang Xing's new year theory
Du Lei: I think Meituan founder Wang Xing's "New Year Theory" (王兴过年论) from 2019 serves as a particularly good example. Wang Xing famously said at the end of 2019: "This year 2019 might be the worst year of the past decade, but it's definitely the best year of the next decade."
Initially, this statement seemed contrarian, but as it proved true year after year, it evolved into a popular internet meme. Though Wang Xing now disclaims creating it, insisting he quoted it from elsewhere, this "each year worse than the last" theory perfectly captures China's pessimistic outlook—we consistently prepare for the worst and fundamentally don't expect the future to improve upon the past.
At an individual level, people in any position—investors, entrepreneurs, CEOs, or ordinary employees—tend to pursue immediate certainty when they lack optimism about the future. This explains why Chinese investment funds operate on shorter cycles than American ones and prefer more certain projects rather than diversifying broadly like American funds. This mentality of seeking collective security, certainty, and quick returns represents the short-term phenomena we observe, possibly rooted in deeper character differences between the two countries.
The paradox of China’s pessimism
Afra: China's "indefinite and pessimistic sentiment" often confuses me and I've been contemplating its roots. When visiting China, you experience impressive infrastructure and consumption options that are noticeably more abundant, higher quality, and futuristic than those in Europe and America. The surface prosperity is undeniable. Each time I take the high-speed train, I'm genuinely impressed and never take it for granted. Given China's remarkable achievements, why does EVERYONE I meet continue to harbor such profound pessimism? One possible explanation is: you never know when the iron fist (铁拳) from the state will come down. This feeling that it could strike at any time creates constant tension, much like playing Russian roulette.
Du Lei: What you said is exactly my reaction when thinking about this. I'm trying to find a more tactful, gentler way to express this viewpoint. I think certainty about the future might be the key OS difference between the two sides.
America also faces uncertainty, but its basic governance logic is very bottom-up. In a strongly bottom-up system, every participant has greater agency. With stronger agency, people think: "Even though I'm uncertain about the future, I can remain optimistic because I always have multiple chances. If I don't win this round, I can try again." This mindset exists because the entire system operates as a bottom-up free market approach: I can always have another opportunity.
But in a very top-down model like China, success or failure often doesn't depend on individual actions, but on which side you stand. Top-down decisions directly determine your fate in this game played with a visible hand. In this situation, all participants lack agency. People think: "As a whole, China will definitely win, but as an individual, I'm not sure if I'll be the tears of the era or a sacrifice." In China, everyone believes the state will ultimately succeed, but no one knows whether they'll be the victor or the price paid for victory. This top-down versus bottom-up approach shapes how individuals judge their situation. People aren't pessimistic about collective success, but are more pessimistic about their personal outcomes.
Afra: I'm curious, if you're on the winning side in China, if you're the creator of the game rules, would you feel optimistic? To be honest, I don't know many princelings (nepo babies with the Chinese state), and I don't have much opportunity to engage with party-state capital. Have you ever met someone who is a princeling and happens to be working on emerging businesses or projects heavily subsidized by the government? Are they optimistic? I suspect not. It's strange - why don't people feel optimistic even when they're on the winning side and know they're definitely winners?
Du Lei: I think the "winning" camp is infinitely subdivided. We discussed earlier the certainty of China's future versus the uncertainty of America's future. But simultaneously, there also exists certainty and uncertainty about the game rules themselves. One type concerns certainty and uncertainty about future development expectations, while another concerns certainty and uncertainty about the game rules themselves. This is somewhat like what the book Finite and Infinite Games describes. The American market is more like an infinite game with clear rules - it's a multi-round game where you can try again if you lose.
This point is particularly evident. The following observation, though sensitive, is worth noting: China's game seems to have no endpoint unless you actively exit. I've found that China’s domestic environment tends more toward one-time games, while American tech circles lean more toward repeated games. This isn't absolute. America also has cases of "make money once and leave, regardless of others' opinions," as we just saw. I'm not saying China is definitely single-round gaming and America is definitely multi-round gaming, but the overall trend is indeed so.
Han Hua: I've been in China for a month already, and might stay for another month. My impression is that, while there's a certain pessimism in society overall, many people are still optimistic, especially friends in the investment community, particularly those active in core investment circles, large capital investments, and those closely connected with state-owned capital and national strategic investments. They're not only optimistic but also very intelligent, willing to take on high-risk investments, hoping to explore possibilities for humanity's future.
Of course, they are "dancing in shackles." Whether they have state-owned backgrounds or are institutions partially funded by the government, they all face government supervision and audit issues. They face two contradictory requirements: on one hand, they want to invest in future disruptive innovations to help China lead the world; on the other hand, these investments are often high-risk with a significant possibility of failure. Meanwhile, regulatory requirements for state-owned and related investments emphasize certainty and don't allow for the loss of state-owned assets. If an investment fails, does it count as a loss of state-owned assets? This is an inherently contradictory dual requirement. I can see that some people are still optimistic, but their optimism is suppressed by the market or regulations.
Additionally, when discussing the pessimistic sentiment in China over the past two to three years, we must consider factors beyond the social system, particularly the pandemic's impact. While Western economies saw inflation and the S&P 500 reaching record highs, Chinese markets remained stagnant, only recently showing signs of recovery. Post-pandemic, China experienced no significant rebound in its stock or capital markets, with property prices in first-tier cities dropping 30-50% from their peak, and even more dramatically in second and third-tier cities. Rather than attributing optimism and pessimism solely to systemic differences, I wonder: would a single bull market be enough to restore Chinese optimism?
Afra: I distinctly remember 2014 to 2017 as a very optimistic golden period in China's tech industry. After "China's Groupon Wars" (百团大战) (where Meituan eventually rose to dominance during China's intense group-buying competition in the early 2010s), China's mobile economy took off, and everyone was discussing the "East rising, West declining" phenomenon. We've discussed perceptions and broader environmental factors quite a bit, but I'm actually more curious—Han, during your month in China, have you encountered any particularly interesting innovative Chinese company, or are there any phenomena you'd like to share?
Han Hua: After examining high-end manufacturing in robotics and electric vehicle industries, I believe America can potentially replicate parts of China's success. The "reindustrializing America" initiative might partially succeed, but won't completely replace China. Both nations can and should maintain strong mutual dependence.
I remain cautious about China’s "national destiny" theories—nothing is predetermined or achievable without effort.1 However, America may be able to partially replicate China's success, potentially reaching a balance of power that would create a more peaceful world. US-China relations will likely continue as a mixture of tense competition and necessary cooperation.
I've encountered many impressive companies, especially in manufacturing. Take Nothing, focuses on mobile and consumer electronics.2 They handle product and brand design while utilizing OEM supply chains with global sourcing, though large-scale production and final assembly primarily happen in China. Interestingly, Nothing implemented a "dual sourcing strategy" from the start by establishing an Indian supply chain. Their phones for the Indian market are produced entirely through this Indian supply chain. Unlike Apple, Nothing planned two separate supply chains from the initial design stage to serve different markets—essentially a political risk diversification strategy.
Open-source strategy not just in AI; various motivations, AI talents competition
Afra: Let's discuss Chinese AI. Few weeks ago, I learned about Nathan Lambert’s "The American DeepSeek Project," which proposes replicating China's open-source AI approach to help America reclaim its position as the premier AI research hub. This proposal is reactionary to how major US AI companies are increasingly adopting closed-source models, while Chinese companies like DeepSeek and Kimi have emerged as leaders in open-source development. According to Kevin Xu, the U.S.’s “‘open source AI deficit’, especially vis-a-vis China, has become very pronounced over the last few months. According to Artificial Analysis, the top four open source AI models are all from Chinese labs.”
I partially understand Chinese companies' open source intentions: demonstrating technical confidence, gaining more influence in international tech circles, and as the number two player, attracting more technical contributors to potentially surpass the leader. Beyond these reasons, do you have other insights about China's open source strategy? You know the DeepSeek team personally, right?
Han Hua: Actually, I don't have much inside information to disclose. Even though I know them, unless you're an insider, relevant details are almost never shared, especially in the current environment. DeepSeek and Unitree are highly controlled, essentially directly led by national ministries, so they can't speak freely.
I think one consideration in making open source decisions, though not the most important factor, is this: we've recently seen many examples like Meta, Anthropic, and companies like Windsurf being acquired. Top AI talent prices have skyrocketed to NBA superstar levels. These top talents are extremely expensive, particularly those experienced professionals who've worked with massive GPU resources and can perform large model training.
In China, if a company wants to compete for top talent, I've heard something interesting: these talents overwhelmingly want companies to open-source their technology. Being able to open-source models, publish papers, and demonstrate model performance is critical for their future career development. I've heard many people would rather accept lower salaries to work on open-source model projects. While this isn't the core factor in corporate decision-making, there's a clear tendency at the employee level.
In America, these talents don't have such strong tendencies to rely on published papers to prove their involvement, engagement, and qualifications.
I also want to mention that China's AI strategy, the open source strategy, isn't limited to AI - it's a common tactic in China's tech circles, especially as a tech follower. Over the past decades, China has largely played a follower role in various tech fields. Facing strong technical barriers and IP protection from Europe and America, it's very difficult for Chinese companies to break through. In this situation, supporting open source is the best way to break the leaders' advantages.
Over the past decades, Chinese tech circles have massively supported global open source projects. You can see Chinese companies' support - funding, personnel, or research - behind almost all major global open source projects.3 For example, after Google open-sourced Kubernetes container management, many Chinese companies joined the same foundation to support it, directly leading to cloud architectures that could compete with AWS and GCP.
Meta did something similar. When Google's infrastructure was seen as excellent but largely opaque, Meta launched the Open Compute Project in 2011 to open-source its own data center hardware designs. This created a shared standard that many cloud providers and hardware vendors gradually adopted. By aligning the industry and supply chain around its framework, Meta reduced costs and benefited from community-driven innovation.
These examples show that open source strategy has always been an excellent tactical differentiation strategy for tech followers. Now many Chinese AI labs use open source strategy, which indeed benefits them in terms of cost, technology, and business models. In addition, although DeepSeek and Kimi have open source models, they usually also have a closed-source version. They release weaker open source versions while selling stronger closed-source versions to enterprise customers.
So it's hard to say DeepSeek is a completely open source company, because its latest models aren't directly open-sourced. It might open source 3-6 months later, or when the next model comes out, it open source the previous one; it might also never open source new versions. But in the current situation, especially in early 2025, facing American model makers' dominance over the ecosystem, open source is the best way to increase visibility and attract users and developers to your standard.
Afra: On AI talent mobility: Are top AI researchers in China being poached by other companies similar to star researchers from OpenAI? I know competition among Chinese tech companies was once extremely intense—Tencent poaching from Meituan, Meituan poaching from Alibaba. But among Chinese AI companies like Moonshot and DeepSeek, do these AI companies poach talent from each other?
Han Hua: Working for open-source AI labs and publishing papers significantly enhances career development by raising one's industry profile. Within the AI community, there exists a group of idealists who aren't primarily motivated by financial gain. These extremely talented individuals genuinely need the open-source environment to express their capabilities, regardless of compensation.
Regarding AI talent competition in China, the rivalry between labs is fierce. However, unlike Silicon Valley, job-hopping isn't as fluid due to the tech industry's non-compete agreements that create real barriers.
The AI researcher community is remarkably tight-knit, with strong personal connections throughout. When someone joins a particular lab, insiders quickly identify which talents are worth recruiting. Many researchers share educational backgrounds from institutions like Tsinghua, Zhejiang University, or UCSD, creating academic and mentorship networks that facilitate information exchange. While China's AI talent competition is undoubtedly intense, it lacks the public transparency seen in the US. This is especially evident with DeepSeek's researchers, who maintain such low profiles that outsiders struggle to track personnel movements.
Vibe coding and developer tools in China
Afra: I have a question for both of you since you're deeply involved in China's tech communities: What percentage of code in Chinese tech companies is AI-written? Which coding tools are they using? I know Chinese AI models like Kimi and DeepSeek often hallucinate when generating code. How do Chinese developers handle this issue? Can they reliably use DeepSeek and Kimi for coding, or are many Chinese programmers secretly purchasing Cursor subscriptions or using Anthropic for programming? I'm aware that Cursor and Claude services are readily available on Taobao.
Du Lei: Let me share what I've observed. Chinese tech giants all have their own AI models - Tencent has "Yuanbao" (元宝), Alibaba has "Qwen" (千问), ByteDance has "Doubo" (豆包) and their related toolkits. These giants maintain their own ecosystems, internally promoting their proprietary models and AI programming tools.
Meanwhile, Cursor is extremely popular in China. Chinese programmers are generally studious and eager to stay current. They bypass the firewall to understand developments abroad and remain aware of emerging technologies in Silicon Valley. This generation of programmers typically tries Cursor first, as it has the strongest brand recognition. Claude Code is also popular, but in terms of brand awareness, Cursor ranks first among overseas AI programming tools, with Claude Code second. Many developers also set up their own GPU solutions.
Recently, Cursor has been strengthening community engagement in China through various activities. In China, large companies use their own models due to data security and IP considerations, plus the need to integrate with their own products. Beyond these companies, many independent developers stay in sync with global tech trends, developing various products and AI tools and market them overseas- this area is thriving domestically. For example, when MCP (Model Context Protocol) first gained popularity, many of the hottest MCP platforms and marketplaces were created by Chinese independent developers. Consequently, there's minimal gap in toolchains and application technology.
Afra: I completely agree, especially regarding Chinese application developers - whether in technical sensitivity, market promotion, or technology adoption, the gap (with US developers) is quite small. I saw that the Cursor Hangzhou meetup was reportedly packed, and I heard Cursor will hold community activities in Beijing and other cities, which seems quite interesting.
Han Hua: Indeed. Regarding overseas development tools, particularly web programming tools, currently only Cursor actively conducts community activities in the Chinese market. I think this has two main reasons: first, Cursor has the largest market share among Chinese developers, at least among overseas tools. Users typically start with Cursor, only turning to Claude Code when they need additional features, or they use both simultaneously.
While Anthropic is a larger organization, Cursor's achievement here is quite impressive. It's worth noting that Claude Code is relatively expensive, making it difficult for many domestic independent developers to afford. Though Cursor recently adjusted its pricing - we can discuss this topic in detail another time.
More importantly, with its large market share and user base, Cursor has created strong network effects. In China, there are many tech influencers with various communities and WeChat groups teaching people how to monetize side projects, and these tech stacks are almost all based on Cursor. This creates a virtuous cycle: because Cursor is popular, discussing Cursor-related topics is also popular; it also represents being "cool," further strengthening network effects. When Cursor's team sees their substantial user base in China, they naturally invest more resources in community promotion.
In contrast, Replit, Lovable, or Bolt haven't made significant moves in China, possibly due to insufficient resources and lacking the brand recognition and mindshare needed to support such activities in China.
Manus: a product of mixed US OS and China OS?
Afra: Finally, let's discuss the possibility of thriving in both China and U.S. operating systems. Do you think some companies might create hybrid products by combining both systems? Manus could be a representative case — after building its reputation abroad, it secured American VC funding, then attracted outstanding talent from mainland China (and laid off many), and moved the entire team to Singapore, where has stronger connections to international capital. I wonder if more Chinese companies will follow Manus's path. When exploring international markets, especially the American market, what trade-offs must be made? And despite these trade-offs, can companies still leverage China's elite talent, efficiency, and entrepreneurial models?
Du Lei: Given the current China-US capital decoupling, companies seeking large American investments, with American IPO exit paths and primarily American market focus, find it difficult to maintain a significant presence in China. Even in early stages, such as securing Chinese Series A funding, you might discreetly maintain a Chinese office for sales, utilizing Chinese resources. However, as you scale, you'll inevitably face capital questions: Do you have sufficient American capital? Will you face censorship? Even without censorship issues, will you encounter congressional scrutiny like Benchmark's investment in Manus? Ultimately, everyone needs to choose sides.
Additionally, for companies like Manus in this AI supercycle, most successful enterprises operate primarily in B2B markets. Even in prosumer areas, small-B large-C markets generate more revenue. In these markets, potential revenue from American markets vastly outweighs the Chinese market. Perhaps after the next consumer cycle begins, this situation might change, but B2B SaaS founders have strong incentives to target the American market. I'm curious if Han has observed similar patterns.
Han Hua: Yes, I generally agree with Lei, especially regarding B2B markets. Of course, our observations assume continued China-US decoupling in capital and politics. Under this premise, this pattern might persist. However, if there's reconciliation between China and America — even if competition continues but markets become more open, or if cultural and political environments relax — Chinese entrepreneurs might face fewer challenges than Manus currently does.
Let me offer an example. Du Lei mentioned Chinese founders seeking American capital, and I've observed American hardware entrepreneurs, particularly in high-end manufacturing, maintaining very open attitudes. A robotics company I invested in, KCIP, develops open-source robotics designs while collaborating with OEMs for hardware production.
These OEMs and supply chains likely come from China, as they recognize complete decoupling is nearly impossible. They're pragmatic — since they want to build successful companies, they'll likely need Chinese supply chains, so they frequently visit China, leveraging these connections to advance their robotics. This applies to many consumer hardware startups that design applications in America but rely on Chinese supply chains for large-scale, cost-effective production. American entrepreneurs are practical, clearly understanding they can't be completely swayed by politics or ideology. Their ultimate goal is building successful companies, so they recognize the necessity of utilizing China's advantages in supply chains and manufacturing.
Everyone's leveraging their comparative advantages. America's enterprise service market, with its high per-customer value and labor costs, creates an extremely attractive environment that draws Chinese talent. Correspondingly, China's supply chain attracts American talent — each side has unique strengths. I particularly agree with what was mentioned earlier — the future will likely, and hopefully, be a state of "fighting without breaking" (斗而不破): continuous competition alongside technological evolution, with both sides leveraging their respective advantages to form a productive equilibrium.
I once wrote an article about national destiny to interpret this nationalistic phenomenon, meaning that "China's technological ascension is inevitable and historically destined. And the rise of companies like Deepseek is evidence of Guoyun (national destiny)." This seemingly superstitious theory is quite influential in the tech industry.
After research, I learned that Chinese tech giants (Baidu, Alibaba, Tencent, Huawei, etc.) are prolific open-source contributors—they’ve open-sourced thousands of projects and are among the most active contributors globally. And some Chinese companies are active sponsors and members of key open-source foundations, including the Linux Foundation, Apache Foundation, and CNCF (Cloud Native Computing Foundation), which oversees Kubernetes. Read more.
Thanks for the shoutout and good additions to the discussion on what China is up to on open source/how Chinese AI companies operate.
“MCP (Multi-modal Control Protocol)” should be model context protocol