The Viscous Age
When progress thickens, not stops
A followup to when institutions learn, and when they don’t
Spring in Kunming, another city, another model in this case for biotech that I’ll cover one day!
Prelude — From Energy to Learning
In Metabolic Wars we looked at the world through the flows of fuel and power — who moves energy, who controls the pipes, and how those flows define geopolitics.
But there’s another kind of flow that decides whether societies actually learn from what they build: the circulation of information, feedback, and trust.
This is where the story shifts — from how power moves, to how meaning moves.
China is entering what might be called the viscous age: a stage of development where speed is no longer the main constraint. The problem isn’t that things don’t move; it’s that everything moves through layers of bureaucracy, infrastructure, and culture that slow, redirect, or absorb momentum. The more data and intelligence the system gathers, the thicker it becomes. Yet that thickness — that viscosity — is also what keeps it from tearing apart.
This essay traces how China is learning to manage viscosity rather than fight it: from its central laws that make regulation part of the software, to its cities that treat friction as a design principle.
1. The Central Layer — When the State Learns to Flow
In most countries, regulation sits on top of industry. In China, regulation has become part of the industry’s wiring.
In September 2025, Beijing’s new law on AI-generated content quietly came into force. The rule sounds simple: every piece of AI-generated material must carry both a visible and hidden label. The hidden tag is baked into the file itself — a kind of genetic code for digital truth.
The system that enforces it, known as “full-chain responsibility,” pushes accountability through three levels:
AI companies must build the labeling tool into their models.
Platforms like WeChat or Douyin must detect those labels and remove unlabeled content.
Users are forbidden from tampering with them.
Instead of policing misconduct after the fact, Beijing has embedded compliance into the technology itself. Law becomes a feedback circuit. The rule produces data; that data improves the rule.
It’s a subtle but profound shift. Where Europe regulates through risk categories and the U.S. relies on market correction, China treats governance as infrastructure. The same principle underlies its national AI registry, which by late 2025 had catalogued more than 800 approved models. Regulation functions like an operating system that keeps the entire machine running within tolerances.
This central layer — the plumbing of rules, registries, and responsibilities — allows provinces to experiment without chaos. Hefei can behave like a venture fund, Shenzhen like a testbed for airspace policy, and Guiyang like the country’s data-cooling reservoir, because the state has already coded a common rhythm for how learning flows.
2. Hefei — The Bureaucracy That Invests
Hefei, once dismissed as a provincial backwater, has turned itself into China’s capital of applied patience.
The local government acts like a venture capitalist, taking temporary equity stakes in high-tech firms — from BOE’s display panels to CATL’s batteries — and selling them once the companies can attract private investors. The profits then seed the next round of strategic bets.
This creates a learning loop:
state capital → firm growth → exit → reinvestment.
Each cycle teaches officials how to judge risk, time exits, and build industrial ecosystems. Bureaucracy becomes a form of collective intelligence.
Hefei’s success rests on another feedback layer — its connection with the University of Science and Technology of China. Civil servants work side by side with scientists, translating patents into products. It’s not glamorous work, but it shortens the feedback path between research and revenue.
In Hefei, slowness is not inefficiency; it’s incubation. The city learns by tying capital to coordination, making policy an iterative craft rather than a decree.
3. Hangzhou — Improvisation as Intelligence
If Hefei is the bureaucracy that learns through planning, Hangzhou is the city that learns through accident.
In September 2025, two unrelated announcements, five days apart, revealed how Chinese innovation actually moves.
First, Alibaba declared a US $53 billion investment to turn itself into the “Android of the AI era.” Days later, a small local lab called DeepSeek released a new programming language, TileLang, designed to make Chinese chips run as smoothly as Nvidia’s CUDA. Within hours, Huawei, Cambricon, and Hygon pledged support.
No one issued a directive. There was no central plan. Yet the moves clicked together like tiles in a pattern that only made sense after the fact.
That is Hangzhou’s genius: coordination without a coordinator.
The city grew up around accidents that stuck. Alibaba never left. Zhejiang University kept feeding engineers. Merchant families supplied capital. The local government learned to be a gardener, not a gatekeeper—watering when needed, pruning rarely.
Its innovation funds act less like command tools and more like tuning knobs. When a robotics startup ran out of cash, a municipal fund transferred twenty million yuan within a week. When the “Six Dragons” of Hangzhou’s tech scene drew floods of visitors, officials built shared demo halls so founders could return to their labs instead of performing for delegations.
This is feedback as choreography: a rhythm of give and take between state, capital, and talent. Hangzhou’s entrepreneurs call it mutual predictability—each actor trusting that the others will adjust rather than obstruct.
It’s also a reminder that viscosity doesn’t only live in paperwork; it lives in culture. Hangzhou’s famous civility—the art of waiting one breath longer before acting—keeps the whole machine from overheating.
Where Hefei’s learning is structured, Hangzhou’s is musical. Together they form a duet of Chinese modernity: one composed, one improvised, both tuned to feedback.
4. Guiyang and Gui’an — The New “Third Front”
Head southwest to Guizhou province and you find another kind of loop: between security and infrastructure.
Half a century ago, Mao moved China’s critical industries inland to protect them from attack — the “third front.” Today, that instinct has returned in digital form. Gui’an New Area, outside Guiyang, has become one of China’s largest clusters of data centers. Huawei, Tencent, and Apple have all built massive facilities here, drawn by cool mountain air and abundant hydropower.
This relocation isn’t only about climate or cost. It’s a strategic feedback system: the more China depends on data, the more that data is pulled away from vulnerable coastal regions into the physical heart of the country. Each new server farm strengthens both the national grid and local economies, turning once-poor Guizhou into a node of resilience.
Guiyang teaches a simple lesson about viscosity: decentralize storage, centralize coordination. The cloud is still national, but its body is inland.
5. Shenzhen — Learning to Fly Below 1,000 Metres
Shenzhen’s feedback loops happen in the sky.
The city’s newest frontier is the “low-altitude economy” — the world of drones, flying taxis, and small autonomous aircraft that operate under one kilometre. Delivery companies like Meituan and SF Express already use drones to move parcels across pilot districts.
The trick is that most of China’s airspace still belongs to the military. So every flight path is a negotiation: city planners, private firms, and air-force officers learning to share the same corridor of sky. Each approval, pause, and modification feeds data back into a national framework that will one day govern everything from delivery drones to air taxis.
Here the feedback loop looks like this:
industry innovates → government observes → policy adjusts → industry scales.
What began as chaos — too many players, no clear authority — is becoming a structured conversation between builders and regulators. In Shenzhen, governance literally learns to fly.
6. Xi’an — Rebuilding the Past to Secure the Future
Xi’an, once the capital of the Tang dynasty, offers a very different feedback system: between memory and legitimacy.
Over the past decade, the city has rebuilt vast portions of its ancient imperial layout — from the Daming Palace in the north to the Tang Paradise park in the south. To Western eyes, it can feel theatrical, even fake. But inside China, the reconstruction serves another purpose.
It allows the Party to test how culture can become infrastructure. Each rebuilt site — a wall, a pagoda, a ceremonial avenue — doubles as a sensor for public sentiment. Crowds, tourism flows, social-media trends: all feed back into a national story about heritage and identity.
This is cultural governance as experiment. Nostalgia is not just mood; it’s a planning tool. Xi’an’s revival shows how a city can teach a country how to feel coherent again.
7. Datong — The Cost of Friction
Not every feedback loop is virtuous.
A decade ago, Datong’s mayor, Geng Yanbo, bulldozed half the city to recreate its ancient walls and temples. His plan did bring tourists and cleaner air, but it also displaced hundreds of thousands of residents. A documentary, The Chinese Mayor, captured his obsession: the belief that beauty could justify suffering.
Datong reveals the moral limits of China’s learning model. Feedback, when insulated from dissent, can harden into echo. Yet even this failure teaches something: that progress under viscosity must be measured not just in buildings but in consent.
The reconstructed streets of Datong glitter under LED lanterns, but the city’s true lesson is quieter — that governance needs emotional bandwidth as much as fiscal capital.
8. The Hidden Circuits — Six Loops of Learning
Across these cities, you can trace six intertwined feedback systems that define the viscous age:
Law ↔ Code: Regulation embedded in software, so rules learn from their own enforcement.
Capital ↔ Industry: Local governments acting as investors, converting risk into institutional memory.
Infrastructure ↔ Sovereignty: Data and compute relocated inland, turning geography into security.
Industry ↔ Regulation: Drone policies and logistics standards co-developed through trial and response.
Culture ↔ Legitimacy: Heritage reconstruction as a means of tuning national identity.
Progress ↔ Pain: The social cost of rebuilding and what it teaches about humane governance.
Each loop makes China’s development thicker, slower, and more stable. It’s no longer a race for the fastest model or tallest tower, but for the most conductive society — one that can move knowledge through complexity without breaking apart.
9. The Builder’s Philosophy
All of this reflects a larger cultural shift. For a century, modernisation meant frictionless acceleration: more speed, fewer obstacles. But in China’s current phase, friction is being reinterpreted as intelligence. The state learns through resistance; cities learn through mistakes; industries learn through limits.
This is why the Chinese system, for all its opacity, keeps producing results that surprise outsiders. It doesn’t innovate by perfect planning but by thick feedback: testing, pausing, recalibrating, and folding each lesson back into the next plan.
Viscosity, once seen as a drag, becomes a stabiliser.
10. Toward The Gold Nerve
If The Viscous Age describes how information and learning circulate through China’s body, The Gold Nerve will describe how trust circulates through its veins.
When fiscal loops stretch and global finance fragments, the next challenge is not how to print more money but how to make value credible again — through collateral, compute, and cooperation. Just as viscosity taught the state how to learn, the gold nerve will teach it how to trust.
References
Central Governance and AI Infrastructure
Poe Zhao, “China Sets a New Global Standard: The World’s First Comprehensive AI Content Labeling Law,” Hello China Tech (2025).
Poe Zhao, “China’s AI Playbook: How State Power and Open Source Are Forging a New Order,” Hello China Tech (2025).
Poe Zhao, “China’s Innovation Infrastructure: Governance as Code,” Hello China Tech (2025).
City Studies and Case Reports
Andrew Stokols, “CITY REPORT: Hefei – Local State as Venture Capitalist,” Sinocities (2025).
Andrew Stokols, “CITY REPORT: Guiyang/Gui’an New Area,” Sinocities (2025).
Andrew Stokols, “CITY REPORT: Shenzhen and the ‘Low-Altitude Economy’,” Sinocities (2025).
Andrew Stokols, “CITY REPORT: The ‘Xi’an Model’: Reviving Tang Chang’an and China’s Imperial Imaginaries,” Sinocities (2024).
Andrew Stokols, “CITY REPORT: The ‘Datong Model’ of Heritage Reconstruction Ten Years Later,” Sinocities(2024).