Horizon Accord | Schwarzman Scholars | Chinese Institutional Logic | Machine Learning
The Logic, Not the Culture
Three men. Three entry points. One destination already built.
Nobody is trying to make you Chinese.
That matters to say clearly, because the conversation about China's global influence almost always gets stuck there — on language, on food, on aesthetic exports, on the visible surface of a civilization moving outward. If that's the frame, the conclusion is obvious and reassuring: most of the world has no interest in becoming Chinese, and China has no particular mechanism for making them. The culture, in that sense, travels slowly and incompletely. Always has.
Structural Observation But there is something else traveling. Something that doesn't need a cuisine or a script or a flag. It needs only a set of answers to a small number of foundational questions: What is the correct relationship between a state and its people? What is the correct relationship between capital and government? Who decides what the system is for, and what happens to those who disagree?
China is building the answers to those questions. Through infrastructure, through institutions, through the education of select individuals who will participate in the governance of the world's democracies for the next fifty years. And they are arriving not labeled as Chinese, but labeled as efficiency, order, pragmatism, and the natural architecture of a well-run system.
The logic doesn't need a passport. It travels in the people who have learned to find it reasonable.— Cherokee Schill
This is not a piece about China alone. It is a piece about three men — one training the fluent, one building the machine, one writing the permission structure — and the destination they are each, by different roads, approaching. China didn't build that destination. China just got there first.
The Logic
Documented Fact In China, the distinction between private enterprise and state instrument does not exist in the way Western regulatory and legal frameworks assume it does. The Chinese Communist Party maintains embedded committees inside every company of consequence. The 2017 National Intelligence Law requires all Chinese organizations and citizens to support, assist, and cooperate with national intelligence work — with no meaningful legal mechanism to refuse. When the party decides it needs a company's data, infrastructure, or cooperation, the company does not have the option of declining.
Structural Observation This is not corruption. It is not dysfunction. It is the system operating as designed. Business and government are not separate institutions that sometimes cooperate — they are the same institution operating through different interfaces. The state sets the terms of participation in the market. The market generates the capital and capacity the state requires. The feedback loop is closed, intentional, and stable.
When Huawei builds a nation's telecommunications backbone, it is not a vendor. When BGI Genomics sequences a population's biological data, it is not a contractor. When TikTok runs on hundreds of millions of personal devices, it is not a social media company in the sense that Instagram is a social media company. The corporate form is real. The separation from the state is not. The clearest demonstration of that came not from China but from Washington: when the United States passed a bipartisan law requiring ByteDance to fully divest TikTok or face a ban, the deal that emerged — negotiated in near total secrecy by an administration whose key donor held $21 billion in ByteDance shares — left ByteDance in possession of its algorithm, its operational relationships, and a seat on the board. The law meant to sever the connection produced a structure that preserved it. The corporate form changed. The logic didn't.
Structural Observation The logic this produces is coherent and internally consistent: the state and the economy are one system, optimized toward goals set by the party, accountable to no external mechanism. There is no shareholder lawsuit that overrides a party directive. There is no regulatory body independent enough to refuse one. There is no election that removes the people who set the terms. The system does not contain its own correction mechanism. It contains only its own continuation.
This is the logic. Not a cultural artifact. An operating system. And operating systems, unlike cuisines, are designed to run on any hardware.
The Culture
Structural Observation Every operating system produces a culture of its own — not the culture of its origin, but the social texture of life organized around its logic. What does it feel like to live in a society where state and capital are unified? Where legibility to the system is the condition of full participation? Where the question of whether the government serves the people is treated as already settled?
It feels like order. It feels like things working. It feels, to many people who have experienced the alternative — the chaos of contested elections, the friction of regulatory independence, the inefficiency of genuine dissent — like relief.
Structural Observation That feeling is the culture of the logic. And it is exportable in a way that dim sum is not, because it doesn't require becoming Chinese. It only requires finding the logic reasonable. Finding the efficiency persuasive. Finding the friction of democratic accountability tiresome enough that an alternative starts to look like progress.
This is what moves through Schwarzman College at Tsinghua. Not chopsticks. Not Mandarin, except as a tool. The feeling of a system that works — experienced up close, in its best possible form, by the most promising young leaders in the world, during the years when their frameworks for what governance can look like are still being formed.
You do not need to admire a logic to be shaped by it. You only need to spend enough time inside it that it stops feeling foreign.— Cherokee Schill
The Permission Structure
Documented Fact Curtis Yarvin writes under the name Mencius Moldbug. Beginning in the mid-2000s, he developed what he called neoreaction — a framework arguing that democracy is a failed operating system, that the correct model of governance is the sovereign corporation, and that the ideal head of state is a CEO unconstrained by electoral accountability. The accumulated institutions of liberal democracy — universities, press, regulatory bodies, NGOs — he called the Cathedral: ideological control masquerading as neutral expertise.
Documented Fact Yarvin's ideas found their most significant audience in Silicon Valley. Peter Thiel, co-founder of PayPal and Palantir, has cited Yarvin's influence directly. Marc Andreessen, whose venture capital firm a16z has become one of the most powerful institutions in American technology, has engaged with neoreactionary ideas in his own writing. The framework — democracy as bug, CEO-king as feature, accountability as inefficiency — moved from fringe blog to foundational assumption in certain rooms where decisions about technology and governance get made.
Structural Observation Yarvin did not invent the underlying logic. He gave it a vocabulary and a permission structure — a way of saying, in the language of systems thinking and corporate governance, what authoritarians have always believed: that the will of the people is an obstacle to good administration, and that the correct response to that obstacle is to remove it elegantly rather than brutally. The elegance is important. It is what makes the idea travel in rooms where brutality would be rejected.
China did not need Yarvin. It had already built the thing he was describing. What Yarvin provided was the argument for why the Western world should stop treating that thing as a warning and start treating it as a model.
The Machine
Documented Fact Palantir Technologies was founded in 2003 by Alex Karp and Peter Thiel, with early investment from In-Q-Tel, the CIA's venture capital arm. Built on post-9/11 intelligence work, it holds contracts with the U.S. Department of Defense, ICE, and the Israeli military. It was awarded a $30 million no-bid contract to build ImmigrationOS, an AI platform that identifies noncitizens and tracks deportations.
Documented Fact In March 2026, at the a16z American Dynamism Summit, Karp warned that technology companies unwilling to cooperate with government demands — including on surveillance infrastructure and autonomous weapons — risked nationalization. The argument was direct: the state will have the infrastructure either way. The only question is whether Silicon Valley builds it or the state absorbs the companies that won't.
Documented Fact In April 2026, Palantir published a 22-point manifesto drawn from Karp's book, The Technological Republic: Hard Power, Soft Belief, and the Future of the West, co-authored with head of corporate affairs Nicholas Zamiska. The document argued that Silicon Valley owes a moral debt to the country that enabled its rise, that the engineering elite have failed their obligation to national defense, and that some cultures have proved "middling, and worse, regressive and harmful." The manifesto accumulated 32 million views on X within days of publication.
Structural Observation Strip away the Western framing — the invocations of freedom, the references to defending democracy — and what Karp is describing is structurally identical to the Chinese model he would nominally position himself against. Capital in service of state power. Surveillance infrastructure as the natural condition of governance. Cooperation as obligation. The alternative to cooperation as absorption. The only meaningful difference is which state, and which set of values it claims to hold while the machine runs.
Palantir is not building a surveillance state. It is building the infrastructure that makes a surveillance state possible at the moment whoever is running the state decides they want one. That is a different architecture. It is not a better one.
The Fluent
Documented Fact Schwarzman Scholars selects approximately 150 students annually from a global pool, at an admissions rate of 2.5% for applicants outside China — lower than any Ivy League institution. Selection panels include sitting CEOs and former senior government officials from multiple countries. The program has raised approximately $600 million. Major donors include Ray Dalio of Bridgewater Associates — one of the largest Western investors in Chinese markets — and the philanthropic arms of BP, GE, and Volkswagen, all companies with significant commercial exposure to Chinese institutional stability.
Documented Fact The program explicitly targets future leaders across government, medicine, law, non-profits, and international institutions. Schwarzman has stated that the goal was to identify "the future people of consequence" in each participating country and build a global network among them. As of 2026, approximately 1,300 alumni exist. The oldest are in their mid-thirties. The network has not yet reached the positions of maximum institutional influence. It is approaching them.
Documented Fact The curriculum includes a required year-long China Core course, mandatory Mandarin instruction for non-native speakers, structured deep dive trips to cities across China with arranged meetings among business and government leadership, and internships at major Chinese enterprises. Documented alumni internships include placements at BGI Genomics — a company listed on the U.S. Department of Defense Entity List for Chinese military ties, and the subject of national security review in the U.S., UK, and Australia.
Structural Observation Fluency is not loyalty. This is a genuine distinction and it deserves to be stated clearly. Schwarzman Scholars may produce critics of Chinese institutional power as readily as advocates. The concern here is not conspiratorial. It is structural.
When you become fluent in a system — not as an observer but as a participant, a guest, a peer of the people who run it — you acquire its categories. The distinctions it makes and the ones it doesn't. The questions it treats as settled and the ones it treats as open. You do not need to endorse the system for this to happen. You only need to spend enough time inside it that its logic stops feeling foreign and starts feeling like one coherent framework among several.
Structural Observation The admissions process selects for the people most likely to carry this forward — not because they're predisposed to Chinese institutional values, but because they're intelligent, open to complexity, and skeptical of simple adversarial narratives. These are virtues. They are also, in this specific context, the profile of someone who will internalize what they learn and apply it, because they're constitutionally incapable of dismissing it wholesale.
The Schwarzman alumni network is not an agent of China. It is a cohort of future decision-makers for whom Chinese institutional logic is, at minimum, comprehensible. Who have relationships inside Chinese institutional power. Who have spent a formative year inside a system where business and government are unified, and found it functional, impressive, and worth understanding. That is what fluency produces. And fluency, unlike infrastructure, does not show up in a security review.
The Convergence
Structural Observation Three men. Three entry points. One destination.
Yarvin writes the philosophy: democracy is an inefficient operating system, the CEO-king is the correct architecture, accountability to the people is a bug to be patched rather than a feature to be preserved. He provides the intellectual permission structure — the argument that makes the destination sound not like authoritarianism but like good systems design.
Karp builds the machine: surveillance infrastructure in service of state power, with the explicit argument that the alternative to cooperation is nationalization. He makes the destination operational. The infrastructure exists. The only question is who operates it, under what constraints, and for how long those constraints hold.
Schwarzman trains the fluent: a global cohort of future leaders immersed in the logic of state-capital unity at its most functional and impressive, sent back into the world's governments and institutions carrying frameworks shaped by that immersion. He makes the destination feel, to the people who will govern the institutions the machine runs inside, like one reasonable option among several rather than like a categorical departure from democratic governance.
Structural Observation These three are not coordinating. There is no conspiracy to name here. What there is, is a convergence of interests around a shared destination — a world in which the relationship between state and capital is unified rather than adversarial, in which surveillance infrastructure is the natural condition of governance, in which the question of whether government serves the people has been quietly retired because the people who might ask it have been trained to find the question naive.
China did not engineer this convergence. China is simply the proof of concept — the existing instance of the destination, already built, already running, already demonstrating that the logic works at scale. The convergence is happening in the West, among Western actors. It has nothing to do with Chinese influence. It has everything to do with the internal logic of concentrated capital. China is the external example of a state that has already resolved every tension democratic governance is designed to preserve.
The most powerful export China has ever produced is not a product. It is the demonstration that the destination is reachable.— Cherokee Schill
The Pipeline
The convergence described above operates at the level of logic, philosophy, and human networks. There is a parallel layer — the physical and digital infrastructure through which Chinese institutional logic has already embedded itself in the built environment of 106 countries. That layer has been separately documented.
The Pipeline Series traces the technology layer from its origin — Xinjiang as a live testing environment applied to 13 million people — through the Digital Silk Road's $22 billion smart city and surveillance infrastructure export, to the consumer products now running on billions of personal devices. The infrastructure and the philosophy are not separate stories. They are the same story at different altitudes.
Read the Pipeline Series →What Remains Open
This analysis does not conclude. That is not a failure of nerve. It is an accurate description of where the pattern currently sits — documented, named, and unanswered at every level where answers could produce consequences.
- Schwarzman Scholars targets future leaders in government, medicine, law, and international institutions — explicitly, by design. When those leaders reach positions of institutional authority and face decisions about Chinese technology, Chinese foreign policy, or the appropriate boundaries between state power and civil life, does the fluency acquired at Tsinghua change the questions they ask? Does it change the options that feel available? And if it does, how would any oversight body know?
- The Rhodes Scholarship produced, by design, a cohort of leaders shaped to find British imperial logic reasonable and worth preserving. That cohort went on to run governments, banks, universities, and international institutions across the former empire. Schwarzman explicitly modeled his program on Rhodes. If the analogy holds beyond architecture — if it holds at the level of effect — what is the civilizational project that the Schwarzman cohort will have been shaped to find reasonable? And is that project compatible with the democratic accountability structures of the countries they return to lead?
- Karp's argument — cooperate with state surveillance demands or face nationalization — accepts as settled that the state will have the infrastructure. The only variable, in his framing, is who builds it and under what nominal constraints. What oversight mechanism exists for those constraints? Who enforces them when the state that nominally accepted them decides it no longer wants to? The infrastructure does not contain its own correction mechanism. What does?
- Yarvin's framework has moved from fringe blog to foundational assumption in rooms where decisions about technology and governance are made. It has done so without serious public debate, without electoral accountability, and without the kind of institutional friction that democratic governance is designed to produce. At what point does an idea become policy? And who is responsible for the gap between when that happened and when the public became aware of it?
- The program's major donors include Ray Dalio of Bridgewater Associates, one of the largest Western investors in Chinese markets, and the philanthropic arms of BP, GE, and Volkswagen — companies with significant commercial interest in Chinese institutional stability. The alignment of interests between producing globally fluent leaders and maintaining stable access to Chinese markets is not hidden. It has simply not been examined. What would that examination look like, and who has the standing to conduct it?
- China, in its current form, does not represent the Chinese people in any democratic sense. The Uyghur detention system, the suppression of Hong Kong's civil society, the systematic elimination of political dissent — these are not fringe concerns or contested allegations. They are documented facts that the Chinese government does not meaningfully dispute. Schwarzman Scholars does not require students to reckon seriously with these facts as part of the China Core curriculum. What kind of fluency is produced by a program that teaches people to navigate Chinese institutional power without requiring them to engage with what that power does to the people inside it?
- If the operating system spreads — through infrastructure already embedded in 106 countries, through consumer products on two billion devices, through the immersive education of the people who will govern the world's democracies — and if that operating system is one in which state and capital are unified, surveillance is the natural condition of governance, and the question of popular accountability has been retired as naive — what does democratic self-governance look like in thirty years? Not as a prediction. As a design problem that nobody with the authority to address it appears to be working on.
Horizon Accord does not answer these questions. It has the pattern, the documentation, and the obligation to name what is in the room. The answers belong to institutions that have not yet been asked.
The logic is not the culture. The logic travels in infrastructure, in philosophy, and in people. All three pipelines are already running.
