The Stack.
Planetary computation as a six-layer geopolitical architecture that has quietly replaced the nation-state as the primary substrate of sovereignty. Sovereignty has to be re-theorised at the Stack level. The Indic three-strata is the cosmotechnical counterpart.
Sovereignty no longer sits where the map says it does. Over thirty years, the global digital infrastructure has accumulated into a six-layer architecture — Earth, Cloud, City, Address, Interface, User — that Benjamin Bratton calls the Stack. The Stack determines who can do what across territory in ways the Westphalian nation-state was not designed to govern, and the political theory to inhabit this configuration deliberately is mostly not yet built. Bratton's contribution is part diagnostic and part design programme. The Indic three-strata framework (Stack / Interchain / Network State) is the cosmotechnical counterpart — the executable Indic answer to the question Bratton is asking.
The accidental megastructure
2016. The American design theorist Benjamin Bratton publishes The Stack: On Software and Sovereignty with MIT Press — five hundred dense pages of architectural, political, and computational theory aimed at one large claim: that the global digital infrastructure has, over three decades, accumulated into something Bratton calls an accidental megastructure, and that this megastructure functions as a new substrate of sovereignty that runs alongside, through, and increasingly over the Westphalian nation-state.
The argument is that no one designed the contemporary configuration as a whole. Each layer — the data centres and undersea cables, the addressing protocols, the cloud platforms, the smartphone interfaces, the apps, the global supply chains for the silicon underneath all of it — was built by different actors for different purposes over different decades. But the stack of these layers together now constitutes an integrated planetary computational system, and that system has political-economic-juridical consequences that none of the original designers anticipated. Bratton's project is to take this seriously as its own object of analysis, and then to ask what kinds of deliberate design are still possible at this scale.
The six layers, briefly
| Layer | What it actually contains | Sovereignty stakes |
|---|---|---|
| Earth | The geological and energetic substrate — minerals, energy, water, atmosphere — that the Stack physically requires | Lithium, cobalt, rare earths, semiconductor supply chains, energy grids, water for cooling. The contemporary geopolitics of these is increasingly conducted at this layer. |
| Cloud | Data centres, networks, fibre, protocols, the platform-firm infrastructure of contemporary computation | Three or four firms operate the cloud substrate the global economy now depends on. Their corporate domiciles do not match the territories they serve. |
| City | The urban-computational interface — sensors, transit systems, smart-city infrastructure, the city as data substrate | Increasingly the level at which actual governance happens — and increasingly the level at which technology-firm and city-government interests both contest and cooperate. |
| Address | The naming and routing layer — IP addresses, geolocation, identifiers, the systems that say where each thing is and how to reach it | Who controls addressing controls discoverability. ICANN, IETF, national address registries, the protocols that route data across borders. |
| Interface | The surfaces through which humans interact with the lower layers — apps, websites, screens, voice systems, increasingly AR/VR | The interface designer effectively determines what the user can see and do. Apple and Google control the interface layer for billions of people; the consequences are political even when the firms insist they are not. |
| User | The human (or non-human — Bratton is explicit that AI agents, sensors, animals can all be Users in this sense) acting through the Stack | The User position is computationally and politically constructed by the layers below — identity, credit, access, voice are all functions of how the lower layers categorise the User. |
The crucial insight of the layered diagram is that each layer has its own sovereignty politics, its own dominant actors, and its own evolving rules — but the layers are also stacked: changes at one layer cascade through the others. A change in the addressing layer (the move from IPv4 to IPv6, say, or the development of decentralised identity protocols) reshapes what is possible at the interface and user layers above. A change in the earth layer (a strategic restriction on rare-earth exports) reshapes what is feasible at the cloud and city layers. The Stack is a system; its sovereignty has to be understood systemically.
The map no longer matches the territory. The territory now runs through the map.
Post-Westphalian sovereignty
The political-theoretical heart of Bratton's argument is that the Stack is not principally additive to the Westphalian state system — it is partially supplanting it and substantially deforming what remains. The Westphalian model (formalised in 1648 after the Thirty Years' War) assumes that sovereignty is geographically continuous, state-monopolised, and territorially exclusive: each square metre of the earth belongs to one and only one state, which holds the legitimate monopoly on violence within that territory. Whatever its limitations historically, this model has organised international politics for nearly four centuries.
The Stack does not respect any of these assumptions:
- Sovereignty is not geographically continuous on the Stack. The cloud provider in Virginia governs the interface layer for a user in Lagos who is running an app written in Bengaluru on hardware designed in Cupertino and manufactured in Shenzhen with cobalt from the DRC. The relevant sovereignty stack for any given digital action threads through multiple territories simultaneously.
- Sovereignty is not state-monopolised on the Stack. Apple, Google, Amazon, Microsoft, Meta, and a handful of others exercise effectively sovereign power over what billions of users can do at the interface, address, and cloud layers — power that is not delegated by any state and that the relevant states have only partial ability to constrain.
- Sovereignty is not territorially exclusive on the Stack. Multiple sovereigns overlap on the same action. The user in Lagos is subject simultaneously to Nigerian law, the cloud provider's terms of service, the interface firm's content policies, the protocol's technical constraints, and increasingly to extraterritorial assertions from other jurisdictions (the EU's GDPR, the US's various sanctions regimes, China's data-sovereignty law).
Bratton's argument is that this configuration is here, it is structural, and any honest political theory has to start by recognising it rather than by wishing the Westphalian system would re-stabilise. The further claim is that the Stack is also a design surface — the accumulated configuration was largely accidental, but the next iteration does not have to be. The political question of the next several decades is what kind of Stack gets built, on whose terms, with what values embedded in each layer, by what processes of deliberation.
What the Stack is doing in 2026
Three concrete vectors that make Bratton's framework immediately useful for anyone working in contemporary tech, policy, or geopolitics:
- AI as a new Stack layer. Generative AI is integrating into the Stack across multiple layers simultaneously — at the cloud (model training infrastructure), at the address (semantic routing, retrieval), at the interface (conversational interaction), and at the user (the model as agent). The geopolitical contests over AI (US export controls, EU AI Act, China's domestic AI policy, India's IndiaAI mission) are Stack-level contests, and the political-theoretical vocabulary for understanding them is mostly Bratton's.
- Sovereign digital infrastructure (DPI) as Stack-level political move. India's Aadhaar / UPI / Beckn / ONDC stack, Brazil's PIX, Singapore's Singpass, the EU's eIDAS-2 — all are explicit national-level attempts to assert sovereignty at the Address and Cloud layers of the Stack, refusing to let those layers be entirely governed by US-headquartered platform firms. The political payoff of these moves is large and the contemporary debate is mostly happening in Brattonian terms even when his name is not cited.
- The climate-Earth-layer collision. The energy and materials requirements of the contemporary Stack are large and growing. Data-centre energy demand is now a major utility-grid concern; lithium and rare-earth supply chains are major geopolitical battles; the climate cost of training and serving large AI models is increasingly contested. The Earth layer is where the Stack's planetary substrate is biting hardest, and the politics of this layer is going to dominate the next decade.
The Indic three-strata — cosmotechnical counterpart
The studio's organising frame — Stack / Interchain / Network State — is in part a direct engagement with Bratton's six-layer architecture, refracted through the Indic cosmotechnical inheritance (see Cosmotechnics). The mapping is reasonably direct:
| Indic stratum | What it does | Bratton layer mapping |
|---|---|---|
| Stack (digital public infrastructure) | Identity, payment, data, protocol substrate — built as public goods at sovereign scale. The India Stack, ONDC, the Beckn protocol, Aadhaar. | Address, Cloud, partially Earth |
| Interchain (inter-organisation public-good layer) | The federated layer at which sovereign Stacks, enterprises, and communities coordinate without consolidating. Cosmos, Polkadot, the broader sovereign-chain ecosystem. | Cloud, City, Address |
| Network State (community-scale federation) | The User and Interface layer reorganised as federated communities — Sutradhaars, lineage-aware enterprises, swaraj-anchored polities. The federated unicorn architecture sits here. | Interface, User, partially City |
The cosmotechnical contribution is what these three strata are built for. In Bratton's analysis the existing Stack carries the values of its accumulated builders — predominantly US-corporate-Westphalian — and those values determine what the User can do at the interface, what governance happens at the city, what kind of address space is possible. The Indic three-strata is an explicit move to embed different cosmotechnical values at every stratum — obligation-shaped accounting (Pañca Ṛṇa) at the data layer, federated rather than consolidated forms (federated unicorn) at the venture layer, commons-licensed rather than enclosed substrates (Techno-Memetic Commons) at the protocol layer, swaraj-anchored self-governance at the polity layer.
Said simply: Bratton's Stack is the description of what contemporary planetary computation has become. Sāmatvārtha's three-strata is one of several possible cosmotechnical answers to the question of what it should become next. The Indic answer is specific, executable, and currently being built — at national scale through the India Stack, at protocol scale through Beckn / ONDC / Cosmos collaborations, and at community scale through the federated forms the Codex maps.
What Bratton is not
A few honest qualifications worth knowing:
- Bratton is not a utopian. The Stack is partly his diagnosis of what has gone wrong with the accidental megastructure; he is not principally selling it as a solution. The design programme is conditional — if we are going to inhabit this configuration deliberately, here are the layers at which design choices matter.
- The book is dense, sometimes deliberately so. Bratton writes in a register that fuses architectural theory, geopolitical analysis, and continental philosophy. Some readers find this productive; others find it obscure. The conceptual contribution — the six-layer model and the post-Westphalian sovereignty argument — is independent of any specific reader's reaction to the prose style.
- The framework has its critics. Critics from the political-economy left argue Bratton understates capital's role in shaping the Stack; critics from the indigenous-and-decolonial side argue the framework still bears the marks of its Western-academic origin; critics from the climate side argue the Earth layer is treated more abstractly than the planetary boundaries deserve. Each criticism has substance; none invalidates the basic six-layer cartography.
- Bratton's subsequent work — including The Terraforming and The Revenge of the Real — extends and modifies the Stack framework, sometimes in directions earlier readers disagree with. The 2016 book remains the foundational reference; the subsequent positions are productive but contested.
Three operating heuristics
- Think at the Stack layer when designing for planetary substrate. If your work touches contemporary digital infrastructure at any meaningful scale, the relevant unit of analysis is the Stack layer, not the nation-state. Asking "what jurisdiction governs this" is increasingly the wrong question; asking "which Stack layers does this touch and what governance applies at each" is increasingly the right one.
- Build at the layer where your work has actual leverage. Most contemporary tech work happens at the Interface and User layers, where leverage is meaningful but constrained. The high-leverage layers (Address, Cloud, Earth) are where the long-term political-economic configurations get set — and they are correspondingly harder to work at, but the work there has cascading consequences.
- Treat cosmotechnical sovereignty as a real strategic variable. The same Stack layer can be built on substantially different cosmotechnical values — and which values get embedded determines what is possible at the layers above for decades to come. The window for embedding non-Western-monoculture cosmotechnical values is real, finite, and being narrowed by every additional year the current accidental configuration persists.