By ETHPrague 2026, the network-state thesis (Balaji Srinivasan, 2022) is no longer speculation — it has crystallized into a small but coherent set of concrete experiments. Network School (Malaysia), Zuzalu / popup-city Ethereum-aligned events, Logos circles, and adjacent jurisdictions (Próspera, Próspera-adjacent zones, Estonian e-Residency offshoots) form an emerging ecosystem of opt-in physical communities backed by digital infrastructure. The intellectual backdrop has shifted from “this could work” to “this has to work, because the post-1945 order is over” — the strongest cluster of governance experimentation in 2026.

Key Ideas

History running in reverse — Balaji’s diagnostic

The most-shared frame of the talk: when you plot world-economic mass over the last thousand years, the curve is U-shaped. For nearly a millennium, world economic mass sat in Eurasia (China, India, Middle East, Europe roughly balanced). The Industrial Revolution rocketed Europe to dominance; WWI and WWII destroyed Europe and the colonial empires; 1945–1991 was a uniquely concentrated period of American dominance, sustained by everyone else being bombed or under communism. Since 1991, the curve has accelerated back toward Eurasia — much faster than it left.

Implications:

  • The post-1945 order (US/UK/France/Russia/China as UN Security Council winners) is structurally unfit for the current world.
  • “MAGA is infeasible because 1945 is not coming back.”
  • Empire dissolutions historically produce many new states (post-WWI: 50 new states; post-1991: many; the American empire’s dissolution will likely produce another wave).
  • Kazakhstan’s president: “If self-determination were realized globally, 600 states would exist.” Whether you welcome it or not, that’s the trajectory.

Network State as the alternative to chaos

Balaji’s optimistic framing: if we see the redecentralization wave coming, we can forge order from chaos through opt-in peaceful communities. The math: most UN member states are small (less than 10M people); 38 have fewer than 1M; cryptocurrencies already rank with fiat currencies in market cap; the technological substrate (remote work + crypto + telepresence + WhatsApp) already exists. Going from 1 person → 17K → 100K → 1M is the same pattern as Bitcoin, Facebook, or Google scaling — just applied to a jurisdictional product.

Network School

The operational instance: a startup society that bootstraps other startup societies, located on an island in Malaysia near Singapore. Concrete economics:

  • ~$1,500/month per person (with roommate) covers room, board, gym, food.
  • Net cost ~$18K/yr to participate.
  • 128 → 256 → 1,024 capacity ramp; eventually nodes worldwide.
  • 100+ countries represented in current cohort.
  • Brand: pro-immigrant, pro-capitalist, “centrist democratic capitalist.” Neither MAGA nor woke.
  • Adjacent: Zuzalu, Próspera, Vitalik-attended events; Brian Johnson’s “don’t die” community as another node.

The key claim: a successful network state proves the thesis. Small countries (Tuvalu’s .tv deal with GoDaddy; El Salvador’s Bitcoin adoption; Bhutan mining Bitcoin) have already demonstrated land negotiating with cloud. Network states are the cloud entity in those negotiations.

Pop-up Cities as a category

Jimin Lee’s ETHPrague talk: pop-up cities (Zuzalu being the canonical example) are governance laboratories. They are explicitly time-bound, opt-in, and rule-customizable. The lessons:

  • They should not try to become permanent. The “temporary structure” is the experimental design.
  • They produce reusable governance primitives that permanent communities can adopt — quadratic funding, conviction voting, role-rotation systems.
  • They are immune to most regulatory categories — short-duration “events” rather than persistent institutions.
  • Felix Fritsch’s “From Temporary Gatherings to Durable Institutions” complements this: pop-up cities are the prototype phase, durable institutions are the v2 phase. The transition is the unsolved problem.

Digital economic zones (Caroline Poli)

The pragmatic middle path: rather than founding new countries, work with existing sovereigns to create special economic zones with crypto-native rules. Estonia’s e-Residency is the prototype; Próspera-class zones expand the model. Many countries actively want this (digital sovereignty, tech-talent attraction, capital flows) but lack the design expertise.

A Field Report from Living Inside a Network-State Experiment (David Stancel)

The most empirically grounded talk. Stancel spent meaningful time inside running experiments. Key reported findings:

  • Coordination at scale is governance work, not technology work. The “we’ll just use Ethereum” answer doesn’t operationalize.
  • Cultural cohesion erodes faster than predicted. Initial trust pools deplete in 6–18 months without active reinvestment.
  • Exit rights must be unconditional and trivial. The moment exit becomes friction-loaded, the community becomes hostage-prone.
  • Geographic concentration creates regulatory and cultural single points of failure. The “global mesh” framing is more robust than “we all live in one place.”

Burning Man as 40-year prior art

The Marian Goodell fireside (also covered in DAO Governance Evolution): Burning Man has operated a cultural-movement-with-physical-instance for 4 decades, scaled to 175,000 affiliated event participants and ~1M unique lifetime experiences. The governance lessons (mission > ownership; source-weighted signal; explicit scope of authority; people show up) are directly portable to network-state experiments.

Future Building Starts with Coordination — Yuting Jiang

Jiang’s talk extends the network-state critique: even with the right technology, the question “who decides?” remains. The proposal: shared coordination primitives (intent expression, dispute resolution, resource allocation) become the public goods that any startup society can adopt. Building these primitives is more leveraged than building specific instances.

How blockchain clubs are on-ramping the next generation (Felix Rihacek)

The grassroots variant: university blockchain clubs are the actual recruiting channel for the next cohort. Rihacek’s argument: the recruitment is not through tokenomics or Twitter; it’s through in-person clubs, hackathons, and mentorship pipelines. The pop-up-city pipeline starts in Discord-coordinated university chapters.

Details / Subtopics

How digital economic zones advance the network state vision

Caroline Poli’s ETHPrague talk argues digital economic zones are the cheapest path:

  • No new diplomatic recognition required.
  • Leverage existing legal infrastructure (Estonian e-Residency, Próspera’s Honduras framework, Liberland’s experiments).
  • Crypto-native rule customization within sovereign zoning.
  • Failure mode: the host sovereign can revoke at any time — Estonia’s e-Residency has been tightened under EU pressure.

Veronica’s role / remote presentation

The Balaji talk at ETHPrague was joint with Veronica and delivered remotely from Network School island. The optics are intentional: the new world’s governance leaders are not physically attending old-world conferences. This is itself part of the thesis.

From the Playa to the Protocol cross-talk

The Marian Goodell fireside specifically connects Burning Man with crypto-Ethereum culture (Josef Je as interviewer). The “showing up” pattern Goodell describes for Burning Man (joining the organizing team via volunteer contribution) is structurally identical to Ethereum core contributor onboarding.

Network State, Network School, & Ethereum [remote] — full Balaji content

Balaji’s specific predictions in the talk:

  • Small countries will be the first network-state adopters (already true with El Salvador, Bhutan, Estonia).
  • The U-shaped political coalition: powerless people fleeing failed states + power users fleeing centralized institutions = the same crypto coalition.
  • “Application to country becomes a full stack thing” — you apply to a network state the way you apply to a college.
  • America’s empire is ending; the post-American disorder is the working assumption. Network states are partly the constructive response.

Logos Network as the open-source variant

Vatzlav Pavlín’s Logos (the Neocypherpunk & CROPS-adjacent project) is the cypherpunk-purist version: decentralized storage + decentralized messaging + on-chain governance + physical Logos Circles in 35+ cities. The Network School is centralized; Logos is decentralized. Both share the network-state-as-platform thesis.

Connections

Open Questions

  • The Network School cost of $1,500/month is significantly below US/EU equivalents but well above per-capita income in most developing countries. Is the model genuinely accessible globally, or is it implicitly a Western-passport program with Malaysian infrastructure?
  • Pop-up cities scale to 1,000s of people; network states aim for 1M+. Is the transition between these scales architecturally continuous, or does something fundamental break around 10K–100K participants?
  • Stancel’s “exit rights must be trivial” finding contradicts the path-dependency required for sustained institutional development. Can both be true?
  • Will any major sovereign formally recognize a network state in the next 5 years? Balaji’s prediction: yes, probably a small island nation, probably as a one-line deal (“.net domain rights to network state X”).
  • The Burning Man template (“we don’t read Twitter”) is incompatible with token-weighted DAO governance. If network states adopt Burning Man’s source-weighting, they cannot use the standard crypto-DAO governance stack.