* Book: ABCity: Blueprint for the Electric-AI Civilization. Myungsan Jun | March 2026

URL =

This book is

Volume 1: Theory. It lays out the theoretical foundation and critique of technofeudalism.

Volume 2 (Spatial Design of the fully renewable-electric ABCity) and

Volume 3 (CityOS Architecture) are currently in planning and will follow in the coming months.

In Volume 3, we are particularly focused on embedding P2P economic mechanisms as deeply as possible into the CityOS architecture.

Contextual Quote

The Starting Point: Energy as a Civic Right

"AI and robots run on electricity. Without electricity, AI is scrap metal and a data center is a warehouse. Simultaneously, energy is the most fundamental substrate of human civilization. Every act of production, every service, every economic transaction originates with energy. And advances in photovoltaic technology have created, for the first time in history, the conditions under which citizens can directly own the means of energy production.

What if citizens owned the means of energy production — and that energy powered AI and robots — and the value generated by AI flowed back to the citizens who own the energy? A structure in which income is guaranteed even as labor disappears; in which the faster AI accelerates, the greater citizens' income becomes. This is what ABCity calls Universal Basic Energy Equity (UBEE) or Universal Basic Energy Equity (UBED) — energy as a civic right. The ABCity project, through this methodology of energy as a civic right, proposes solutions to the four challenges identified above. This book is that proposal."

- Myungsan Jun

Summary

(summary produced by the author for Michel Bauwens)

Myungsan Jun:

ABCity stands for AI Blockchain City — a proposed prototype city of 100,000 residents built around three integrated principles:

citizen collective ownership of

renewable energy infrastructure,

AI-operated urban governance via an open

operating system (CityOS), and

blockchain-secured direct democracy (RTED).

The name reflects the book's core argument: that AI and blockchain are not technologies to be layered onto existing cities, but the foundational architecture of a new kind of city — one designed from the ground up for the conditions the AI era is producing. We are living through two simultaneous shifts that most analyses treat separately. The first is the AI and robotics revolution. In 2026, the displacement is no longer theoretical. Hundreds of thousands of workers have been eliminated by AI systems. Amazon has reduced over 30,000 positions. Salesforce cut an estimated 4,000 support roles after AI took over more than half its customer inquiries. Samsung has announced that by 2030 all its production facilities will be AI-driven factories. The income architecture that has sustained industrial societies for 250 years — work, wages, consumption, welfare — is under structural strain, and no existing policy framework has produced a durable answer to what replaces it.

The second is the energy transition. China now produces nearly 10,000 TWh of electricity annually — more than the United States, the EU, and India combined — and added 315 GW of new solar capacity in 2025 alone. Electricity is no longer merely an energy source. It is the substrate of AI computation, the determinant of industrial location, and the new axis of geopolitical competition. Who controls the energy infrastructure increasingly controls the AI infrastructure, and through that, the economic future.

The central argument of ABCity is that these two shifts are not separate problems requiring separate solutions. They are one structural condition requiring one architectural response: a city in which citizens own the energy infrastructure that powers AI, govern through a real-time democratic system, and receive as basic income the revenues that ownership generates.

The Proposal: Four Integrated Elements

UBEE — Universal Basic Energy Equity

Every previous attempt at basic income has collapsed at the funding question.

Tax-based transfers are politically fragile and economically procyclical. Robot taxes face global competition constraints. ABCity proposes a different starting point: not redistribution, but ownership.

In ABCity, citizenship is a right of fractional co-ownership in the city's energy production infrastructure. Solar installations — covering dedicated zones, rooftops, parking structures, and facades — generate electricity sold to the grid and to co-located data centers under RE100 long-term contracts. The revenues are distributed as basic income. Not as welfare. As a property right, constitutionally protected, deriving from ownership of a productive asset rather than from the political will of any government.

The reference design: a city of 100,000 residents on 30 km², with approximately 3 TWh of annual electricity generation (perovskite tandem bifacial panels, Powered Cell data center structures, full BIPV deployment). City self-consumption: approximately 0.5 TWh, consistent with passive house construction standards and 100% electric transportation. Surplus available for sale: approximately 2.5 TWh annually. Target basic income: approximately ₩500,000 ($357) per person per month — the floor achievable under Korea's constrained land conditions. In land-abundant regions, the same model produces substantially more.

The critical structural feature: as AI and robotics accelerate, energy demand rises. A city whose citizens own the energy infrastructure sees its basic income grow with automation, not against it. The economic vitality and the social purpose are, by design, the same thing.

CityOS — The AI Operating System for Democratic Cities

Every city already runs on an operating system. The difference is whether that system is visible, accountable, and subject to democratic control — or invisible, proprietary, and managed by whoever controls the infrastructure. CityOS makes the city's operating system explicit. It does not propose to automate more — cities are already automated. It proposes to make that automation governable.

Two predecessors are worth naming. Conventional smart city projects — Songdo, Sidewalk Labs, Singapore's Smart Nation — automate specific domains in isolation: traffic here, energy there, permits elsewhere. The systems do not talk to each other, the data does not flow between them, and no unified governance layer exists. The result is partial and fragmented automation dressed as intelligence.

China's City Brain - Hangzhou, now in its third iteration in Hangzhou, goes further. It is the closest existing approximation to a genuine city operating system: a unified AI platform that manages traffic, emergency response, public services, and urban administration in real time across a city of 12 million people. City Brain 3.0 (launched March 2025) deploys domain-specific AI agents — a virtual police officer, a mental health specialist agent handling 17,000 daily consultations, a medical insurance processing agent — coordinated through a central platform powered by DeepSeek-R1. It is, in the author's assessment, an early prototype of what an AI-governed city can look like operationally.

What City Brain cannot answer is the question citizens are entitled to ask: on whose behalf is this system optimizing, and who decided that? The answer — the state's behalf, and the state decided — is the answer of a system that manages citizens rather than serves them. The data is centralized, the vendor is a single corporation (Alibaba), the governance layer is invisible, and the revenues flow to the state and the platform, not to the residents whose data and whose city generate the value. CityOS begins from the opposite premise. Three technical layers distinguish it from every existing smart city platform.

The first is an ontology — a formally defined layer in which every concept relevant to city operations (energy loads, traffic patterns, citizen votes, policy rules) is precisely specified in machine-readable form, so that AI cannot drift from the meaning citizens intend. The second is a four-layer blockchain architecture — recording every AI decision, every citizen vote, and every policy execution as an immutable, publicly auditable ledger. The third is a sovereignty mechanism: citizens are not in the loop of AI decisions but on the loop, holding constitutional authority to redirect, override, or redesign any automated process at any time.

There is a further implication that will be familiar to readers of your work. The matching and data functions currently monopolized by large platform companies — ride-hailing, delivery, short-term rental, labor markets, local commerce — are not technically complex. They are institutionally entrenched. A city-scale AI operating system with full access to real-time urban data can replicate those matching functions without the platform intermediary. When a city owns its own operating system, the rent that currently flows to platform owners can instead remain within the commons — distributed to the citizens whose activity generates the value.

CityOS, in this sense, is the infrastructure for city-level peer-to-peer production and exchange: direct matching between residents, workers, producers, and consumers, governed by the community itself, without the extraction layer that platforms impose.

The partner state role Bauwens describes — the public authority that actively builds the infrastructure for commons governance rather than merely regulating markets — finds a concrete institutional expression here: a city that owns its operating system, opens its data as a public resource, and routes the value generated by citizen activity back to citizens rather than to platform shareholders. This is not a secondary feature of the design. It is one of the primary economic arguments for building the city's operating system as citizen-owned public infrastructure rather than licensing it from a private vendor.

The contrast with China's City Brain is deliberate. City Brain is technically sophisticated. What it cannot answer is the question citizens are entitled to ask: on whose behalf is this system optimizing, and who decided that? CityOS begins from the opposite premise.

RTED — Real-Time Executable Democracy

The critics of democracy are not wrong about the speed problem. China makes large-scale infrastructure decisions in months; democratic nations spend years in permitting and litigation. But the slowness of democracy is not an inherent feature of democratic governance. It is the consequence of running eighteenth-century institutional architecture on twenty-first-century problems.

RTED — Real-Time Executable Democracy — is the redesign.

Its core principle: decision equals execution. Existing democracy operates on a three-step delay — citizens decide, government interprets, administration implements. RTED eliminates that delay. When a vote passes, the code deploys. When the code deploys, the city changes. Citizens' collective decisions are encoded as executable smart contracts and automatically enforced on the city's physical and digital infrastructure. A vote to change energy pricing deploys a contract that updates the distribution system. A vote to redesign a traffic pattern updates the signal algorithms.

This is not merely a matter of efficiency. It is the mechanism that keeps AI systems from drifting away from what citizens actually want — because the rules those systems operate under update at the speed of democratic decision, not the speed of legislative procedure. RTED is CityOS's highest layer: the constitutional architecture through which citizens hold sovereign authority over the AI-governed city below them.

Estonia digitized the interface between citizens and government. Taiwan's vTaiwan accelerated civic consensus. Barcelona's Decidim made participatory budgeting routine. None of them closed the loop. RTED attempts to close it — to connect the citizen's vote to the physical world without intermediary delay.

Replication — The Measure of Success

ABCity is not designed to be one city. It is designed to be the first city in a replicable series. If five hundred cities of 100,000 residents are built on this model, the host country produces more than 1,500 TWh of renewable energy annually, generates approximately ₩300 trillion ($214 billion) in basic income per year, and accumulates a library of codified safety policies for AI and robotics governance that itself becomes an exportable asset.

The replication logic is explicit: each city's success reduces the capital and political risk of the next. City One is not the destination. It is the proof of concept that makes City Two, Three, and Ten possible.

Why Korea — and Why Now

This book was written by a South Korean, from a South Korean vantage point. The examples, the institutional references, and the policy experiments it draws on are grounded in Korean experience. But the problems it addresses are not Korean problems.

Korea provides the near-worst-case reference conditions: extreme land scarcity, high population density, an existing energy system dominated by centralized utilities, a regulatory environment that has not yet created space for citizen-owned distributed generation at city scale. If the model works here — and the numbers suggest it can — it works in conditions far more favorable to large-scale renewable deployment. The book is the first of three volumes. Volume Two addresses the physical design — spatial organization, infrastructure layout, construction economics. Volume Three addresses the technical architecture of CityOS — drawing on systematic analysis of existing platforms from Palantir's ontology-based intelligence infrastructure to China's City Brain. The present volume establishes the prior question: why this civilization is necessary, and what its economic and institutional logic must be.

The founding ideal the book proposes is Hongik Ingan (弘益人間) — a five-thousand-year-old Korean founding principle meaning "to broadly benefit the human world." ABCity reinterprets it through a new methodology: energy basic rights + AI automation + direct democracy. Citizens, empowered by direct ownership of energy, become active benefactors of humanity rather than mere survivors of technological displacement."

Excerpt

The Civilization Redesign Imperative

"The preceding sections have described AI as it exists today — systems that accelerate electricity demand, displace labor, and concentrate economic power. But the civilizational challenge this book addresses does not end there.

Design choices made now will need to hold across all three stages of AI development — narrow AI, AGI, and superintelligence — and the book's argument applies to all three. Narrow AI — the systems already operating at scale in 2026 — automates specific tasks, drives explosive growth in data center energy demand, and begins displacing categories of knowledge work. Artificial General Intelligence (AGI) refers to systems capable of matching or exceeding human performance across a broad range of cognitive domains; most leading researchers place this transition within the next decade, though timelines remain contested. Superintelligence describes systems that surpass human cognitive capacity across all domains — a threshold whose implications for labor, governance, and economic distribution are difficult to overstate.

This book is concerned with all three stages, not only the first. The premise is structural rather than predictive: regardless of when AGI or superintelligence arrives, the civilizational architecture required to govern them — citizen ownership of the energy that powers them, AI operating systems accountable to democratic oversight, blockchain-guaranteed transparency — must be designed and tested now, while the window for deliberate choice remains open. A civilization that waits until superintelligence is present to ask who owns the infrastructure it runs on will have waited too long."

Energy Is Civilization: From Fire to Photovoltaics

"Viewed across the broadest historical perspective, the history of human civilization is the history of energy system transitions. The most fundamental variable determining civilizational form is not any particular technology but the type and scale of energy that technology transforms. Energy is defined in physics as "the capacity to do work." In the context of civilization, however, energy is not merely an instrumental resource. It is the precondition of all physical transformation — and therefore the ultimate basis of every change that human beings impose on the world: production, movement, construction, computation. Constructing a building, fabricating a semiconductor, training an AI model — each is, at the physical level, a process of converting energy into a particular form of order. Thermodynamics calls this negentropy: negative entropy."

Distribution "after" labor appears structurally unworkable: Ownership Before Redistribution as the Pre-Labor Principle

"Where can a new distribution logic be found?

The most frequently proposed approach — collecting more taxes and redistributing — is an attempt to create a new distribution logic in the process after labor. Corporations generate profit through AI and robotics; the government taxes a portion of that profit and distributes it to citizens.

This logic has structural weaknesses.

First, the political feasibility of taxation. A handful of symbolic entrepreneurs — Bill Gates, Sam Altman — advocate raising taxes, but whether this is actually achievable is doubtful. Most corporations perpetually struggle under the threat of declining performance. The possibility that a small number of tech giants will accumulate enormous wealth exists, but there is no guarantee it persists for decades; corporations face relentless competition. In such conditions, gaining broad entrepreneurial consent to tax increases is genuinely difficult.

Second, geopolitical competition makes the first problem still harder. The United States and China have engaged in continuous systemic competition, intensifying dramatically in the AI and robotics era. This inter-state competition will constrain the fiscal space available for corporate taxation. China is directing the AI and robotics era through state-centered strategy and resource reallocation; for American tech giants leading this transformation, additional taxes represent a competitive handicap. Under conditions of intensifying systemic competition, which state will implement tax policies that weaken its own AI corporations?

The Big Tech AI arms race is proceeding at unprecedented scale. McKinsey estimates that approximately $7 trillion in data center infrastructure investment will likely be required globally by 2030. Capital expenditure plans for 2026 by just the four hyperscalers — Amazon, Microsoft, Alphabet, Meta — approach $700 billion. Bank of America notes that capex now absorbs 94% of operating cash flow (up from 76% in 2024) — effectively reinvesting nearly all earnings into AI infrastructure, with some firms beginning large-scale borrowing to bridge the gap. Meta and Oracle issued $75 billion in bonds in two months (September–October 2025) to fund AI data center construction. Google (Alphabet) entered the bond market most aggressively: approximately $56 billion in bonds across four currencies from November 2025 to February 2026, including the first century bond ever issued by a technology company — this from a company holding $126.8 billion in cash, because its 2026 AI capex budget of $175–185 billion exceeds available cash flow. Whether these investments will generate returns remains uncertain; analysis suggests that 80% of AI projects fail to generate their projected value. In short, the assumption that Big Tech will pay taxes to fund basic income does not easily hold. Third, the scale problem. Even if additional taxation were possible, whether it could sustain all workers displaced by AI is questionable. The projected scale of AI-induced unemployment is without historical precedent — whether the timeline is five years or ten, society does not have the capacity to absorb unemployment at this scale.

Fourth, the instability of the funding base. Tax-based basic income is vulnerable to economic cycles. When corporate profits decline in recessions, tax revenues decline. Income support is most urgently needed precisely when the economy is worst — yet tax-based basic income is most likely to fail at exactly the moment of greatest need.

From this perspective, distribution "after" labor appears structurally unworkable. Every model that depends on capturing a share of the AI economy's gains after the fact will face the same fundamental vulnerability: the gains can be moved, sheltered, or competed away. What if we look instead to the process before labor? This is precisely the methodology ABCity attempts. To secure the distribution logic in the process before labor — in the very process of producing energy.

Whether voluntarily or involuntarily, the work of converting energy that human beings have performed must now be transferred to machines. Then why not provide human beings, equitably, with the "right to produce energy" — the right that exists at the stage before energy conversion? Energy is an essential element of social maintenance; the future will require not less but more of it. There is no better basis for a new distribution logic."

(via email, March 2026)

AI Blockchain City

Contents