Hyperstructures Redux

As Interlocutor-at-Large, the deputy editor of e-flux Architecture and head of Architectural Design at the Gerrit Rietveld Academie annotates, responds to, and unpacks Jacob Horne’s seminal text.

Text Jacob Horne and Nick Axel
Published 21 Jul 2023

Originally Published January 16, 2022.

After spending two years working on Zora, I’ve come to realize how distinct our approach is. This difference comes from a fundamentally new mental model of what’s possible (and already happening) with crypto protocols.

The nature of crypto protocols opens up radically expansive new models to create invaluable public infrastructure. This view is so distinct that I think it warrants new language.

In this essay I outline the concept of Hyperstructures: what they are, how they work and why they will be the basis of the internet for the next decades to come.

Paolo Soleri, City in the Image of Man, 1969, Rendering for Arcology Research¹


Hyperstructure: Crypto protocols that can run for free and forever, without maintenance, interruption or intermediaries.²

Hyperstructures take the form of protocols³ that run on blockchains. Something can be considered a Hyperstructure if it is:

  • Unstoppable: the protocol cannot be stopped by anyone. It runs for as long as the underlying blockchain exists.
  • Free: there is a 0% protocol-wide fee and the protocol runs exactly at gas cost.
  • Valuable: accrues value which is accessible and exitable by the owners.
  • Expansive: there are built-in incentives for participants in the protocol.
  • Permissionless: universally accessible and censorship-resistant. Builders and users cannot be deplatformed.
  • Positive sum: it creates a win-win environment for participants to utilize the same infrastructure.
  • Credibly neutral: the protocol is user-agnostic. (4)

(1) “Why might we understand an experimental planned settlement in Arizona as an arcology, but NEOM, the Line, and Masdar are always already failed future cities? After all, what do utopian cities leave behind? Blurry hypertrophic scars on Google Maps, some urban ruins, some tchotchkes—all I remember of Arcosanti were the carefully patinated bells—and hopefully no messianic cults or massacres. One or a few big ideas at best.” Rahel Aima, “The Khaleeji Ideology,” Horizons (e-flux Architecture, 2022).

(2) The mess of a place is often what makes it real. And is it possible for anything to be truly without maintenance, without interruption, and without intermediaries? Maybe the protocol itself does not need to be maintained, but the infrastructure that allows individuals to use it does.

(3) What (else) can protocols be used for? We should know what we build a protocol for before building it. See: Aristide Antonas, Archipelago of Protocols (dpr-barcelona, 2016).

(4) This reminds me of Elinor Ostrom’s wonderful principles for designing collective institutions for the governance of common-pool resources, which importantly overcome the twin fallacies of the tragedy of the commons and the prisoner’s dilemma. They are: 1. Clearly defined boundaries. 2. Congruence between appropriation and provision rules and local conditions. 3. Collective choice arrangements. 4. Monitoring. 5. Graduated sanctions. 6. Conflict resolution mechanisms. 7. Minimal recognition of rights to organize. 8. Nested enterprises. While this might be more applicable to what happens when people use the protocol than the protocol itself, it’s interesting to think about what resources are being shared. What is the value of this commons? And how is it governed? See: Elinor Ostrom, Governing the Commons (Cambridge University Press, 1990).

It is worth noting that just because an application is protocol-based does not immediately mean it is a Hyperstructure. For example, the Wyvern protocol that’s (amongst others) used by OpenSea (5) cannot operate without offchain orders being maintained in a privately operated database. This means that if OpenSea or any other platform built on it goes down, the market goes down with it. (6) This is also true of a platform like Coinbase. It facilitates the exchange of cryptocurrencies, but if it goes down, their market goes down with it too. A more nuanced example: if you see a completely onchain protocol with the ability for admins to completely upgrade and change the rules of a protocol, then that is not a Hyperstructure. That is also a platform. This isn’t to say any of these examples are inherently bad; they’re just not Hyperstructures.

Hyperstructures are entirely onchain, and are public (7) goods which create a positive-sum ecosystem for any participants. I’ll go into detail about each of the primary characteristics of a Hyperstructure below.

(5) Is OpenSea the model example of the sort of platform we’d want to run on our protocols?

6) The ontology of contemporary Web3 is modeled on a market-based economy. But are markets not fundamentally capitalistic? See: Michel Callon, Markets in the Making: Rethinking Competition, Goods, and Innovation (Zone Books, 2021).

(7) Who is “the public”? Who is “we”? Who does our design serve? See: Het Nieuwe Instituut, Who is We? (2021).


Traditional infrastructure (8) like power grids and social media platforms require a trusted intermediary to maintain and operate it—without it the infrastructure will degrade and/ or stop operating altogether. These operators are organizations, either privately owned or state-controlled—and they require profits, labor, and/or subsidies to serve their purpose over time. (9)

By nature of running on a blockchain, Hyperstructures can run forever with no ability to stop them. (10) They can continuously function without a maintainer or operator, (11) and they can run for as long as the underlying blockchain is running—which can be at the very least a decade. (12) Hyperstructures are unstoppable, and this is a superpower that we have no good prior example of. You can have a one-time creation cost, and then once deployed it will run exactly as designed and without degradation. There is no further labor or capital required to sustain the Hyperstructure. If the core team or platform built on top of the Hyperstructure were to disappear, it would still run exactly as designed and be fully operational for decades. Unstoppability (13) is a net new capability that is possible because of blockchains, and it’s a capability that I believe changes the economic nature of infrastructure.

If we were to zoom out one level of abstraction, we could call the blockchain the trusted trustless (yes you read that right) operator of the Hyperstructure.

(8) Much work is being done to develop more resilient, decentralized, low-impact infrastructure. This has also always been the case in many parts of the world, and remains that way in places that haven’t been subject to total development. For historical cases, see: Ernst Fritz Schumacher, Small is Beautiful: A Study of Economics as if People Mattered (Blond & Briggs, 1973) and Ivan Illich, Tools for Conviviality (Harper and Row, 1973); For further elaboration on the meaning and consequences of what I refer to as “total development,” see: Amitav Ghosh, The Great Derangement: Climate Change and the Unthinkable (University of Chicago Press, 2016).

(9) The unfortunate reality is that infrastructure does not work for most people on this planet. That’s not to say it doesn’t work, but it is not designed for—nor does it reach— the vast majority of the human population. Neoliberal, privatized economies tend to weaponize, appropriate, and coopt infrastructure, and it’s often designed against those who are serviced by it. “Trusted intermediaries,” then, are not only rare, but often explicitly untrusted. On the relationship between infrastructure and power, see Power: Infrastructure in America (2016–2021), a project from Columbia University's Temple Hoyne Buell Center for the Study of American Architecture.

(10) This doesn’t mean they can’t die.

(11) Is it really “without,” or is it simply that we, the Hyperstructure’s users, become the maintainer or operator?

(12) I guess that’s the nice thing about bricks and mortar: they last!

(13) I wonder how this could apply in reverse, to traditional infrastructures.


Now that we’ve established unstoppability as a net new feature, I’ll make my most speculative and contentious proposition: Hyperstructures can be simultaneously free forever to utilize and extremely valuable to own. (14) This is possible because of their unstoppability. There is no cost to maintain and keep the protocol operational forever. Once deployed, it will work exactly as it’s designed, with no degradation. For example, if the Uniswap team and website disappeared today, the protocol would run in perpetuity. This is something that simply hasn’t been possible before. It is worth stating that while the Hyperstructure runs exactly at cost, someone still has to pay the gas cost to operate it at that point in time.

Jacob Horne, Untitled, 2021, GIF. Courtesy of the internet

(14) Excuse my French, but no shit. Infrastructure literally creates value. The entire idea of wealth is inconceivable without infrastructure. See: Timothy Mitchell, “Infrastructures Work on Time,” New Silk Roads (e-flux Architecture, 2020).


A Hyperstructure can simultaneously be free to use and also extremely valuable to own and govern. (15) This is a familiar value model that we observe for NFTs: the media can be universally free to consume, yet valuable to own and control (16) as an individual or group.

But what does ownership mean in this context? The presence (and control) of a fee switch that can be turned on across the protocol. This creates a dynamic called the “threat of the fee.” (17) This means that owners of the Hyperstructure have the right to turn that fee on across the protocol at the base level at any time via a vote. (18) It’s the threat of the fee, because it is long-term value-destructive to ever turn it on. Turning the fee on is a value-destructive action because it would immediately lead to an incentivized fork, since there’s now a clear reason for new entrants to do it themselves.

This right to destroy is an ownership right, (19) in the same way that owners of an NFT have the right to burn. A rational actor won’t do it, (20) but they can if they want to. In my opinion, this right to destroy creates a natural market force to ascribe true value, since people will want to protect that value and the utility being offered at large.

(15) What about to use? I find it unfortunate that the position being taken in here focuses entirely on ownership (see Pierre-Joseph Proudhon, Property is Theft! (AK Press, 2011)) while completely ignoring the question of use. That said, I’m not arguing that this should be extended to say “extremely valuable to own, govern, and use,” because the highest goal of use is not necessarily value. See: Johan Huizinga, Homo Ludens (Beacon Press, 1971).

(16) Here, again, is this problematic omission of usership. The point of making media is not that it is free. Or that it is consumed! We make media because we want to do things. This doesn’t necessarily have anything to do with whether the “consumption” of media takes place in a space of freedom. I would argue, in fact, that we should look at the function, the role, the meaning, the effect, and the potential of media in fundamentally un-free contexts to understand what media can do and what we want to do with it. Looking at the 20th-century history of life under authoritarian regimes, we clearly see that media is more than something to be “consumed.” The entire concept of “consumption” is hugely problematic.

(17) Does this not already presume the form of a market, an economy? Exchange? Must we dream our most utopian dreams on the foundations of commerce?

(18) Or other (either more or less democratic) mechanisms! What guarantees that it will be a vote-based operation?

19) Indeed. Ownership is never a single right, but a bundle of rights.

(20) Never count on rational action. Stickers on fruit are edible because if something is served to people, or bought as food, people will eat it, even if it’s plastic grass or a product label. Same thing with architecture. If it’s possible for people to do something, they will do it. Good design always has to plan for the most diverse and unexpected uses, even unintended and detrimental ones.

The right to sell and transfer the fee switch is another ownership right. (21) This ability to sell and transfer provides a direct medium for the market to exchange and therefore value it. The DAO could also choose to sell the protocol fee switch to another party—this creates a price, and therefore value that can be held.

On top of the right to turn on and receive a protocol fee, other ownership rights can include the right to deploy new versions and governance over a DAO treasury for ecosystem funding and development.

There is a subtle difference between profit and value (22)—and, by extension, profit extraction and value creation. While it is entirely possible to extract profits in the short term, this is likely in direct conflict with long-term success and value creation. I would even go as far as to say that being “for-profit” is a skeuomorphic mode of operation, and that it leads to a local optima that misses the broader opportunities that come from a diverse platform ecosystem on top of your protocol. The nature of the medium of token ownership and unstoppability means that we no longer need to extract profits to realize value creation. (23) We can create “for-public,” (24) with a new underlying value system, which means that the value created in Hyperstructures far exceeds any amount of profits that can be extracted in the immediate term, and Hyperstructures can be recognized for the value they provide to society at large. (25)

(21) Please, as Foucault said, let us not become enamored with power. And it is enamoring, but there is more to life. See: Leopold Lambert, “# FOUCAULT /// EPISODE 2: DO NOT BECOME ENAMORED OF POWER,” (The Funambulist, 2012).

(22) Yes! Very important!

(23) This is great. And sounds emancipatory, from something. But I don’t yet understand why “value creation” is the primary goal of building a protocol.

(24) This is great, but the term “public” is never neutral, never all-inclusive. Publics are constructed. We should be more specific about who we are actually speaking about, who we actually reach.

(25) What bridges are there between the platform ecosystem and “society at large”? How large can the society be that Hyperstructures provide value to?

Marc Ngui, 1914 Paragraph 8 (Methodical Interpretation of Gilles Deleuze and Félix Guattari’s A Thousand Plateaus: Capitalism and Schizophrenia [1980]), 2010. Courtesy of Happy Sleepy

Expansive Fees

Hyperstructures have Expansive Fees—built-in incentives that can be utilized by anyone adding value (26) in a codified manner on top of the protocol. These fees are transparent and available to the entire ecosystem. To effectively earn these fees on top of the Hyperstructure, you need to provide the desired value as defined by the protocol. The presence of these incentives provides a mechanism for community expansion of the Hyperstructure, and good mechanism design here is likely critical for longevity.

A great example of an Expansive Fee is the Uniswap LP fee. LP fees incentivize participants to provide the key resource: liquidity. This fee is paid to anyone who is providing liquidity to a pool; it is not paid to Uniswap. LPs expand the utility of Uniswap. They do not have a monopoly on providing that utility, and an LP only captures value on what they’re creating.

Another example is the Zora Finder’s Fee, which incentivizes (27) participants to provide a key resource: distribution. This fee is paid to anyone who finds the winning bidder or eventual buyer of an NFT. (28) Finders expand the utility of Zora, and they do not have a monopoly on the NFT markets being displayed; a finder only captures the value they’re creating.

(27) The effect of incentivization is essential. But to reduce the basic dynamics of human action to what I interpret as an economic concept doesn’t feel right. I wish we could think more expansively about this. Incentives presume that we don’t already want to do something and that we have the complete free will to choose whether we do it. This is something that cannot account for structural conditions or dispositions.

(28) What would this text or other Web3 texts look like if we removed all words that signify capitalism? In this sentence: “fee,” “winning,” “bidder,” “buyer.”

(29) This sounds incredibly important. But again, why “value”? Shouldn’t we be focusing on what can be generated, created, or produced using these protocols? That’s what seems to define the value of things. And is there not anything more valuable than “value” itself? Infrastructures have disposition, but value isn’t intrinsic to anything. On the concept of infrastructure and disposition, see: Keller Easterling, Enduring Innocence: Global Architecture and Its Political Masquerades (MIT Press, 2005).


Permissionlessness is a guarantee that is unique to crypto protocols deployed on blockchains. Permissionless means that a Hyperstructure:

  1. Is universally accessible; everyone is able to utilize it completely without prejudice
  2. Is unable to change its core operating functionality
The Uncensored Library in Minecraft that features inaccessible journalism. Courtesy of RSF/DDB

The implication of being permissionless means that everyone can confidently build their own platforms, applications, and economic models on top of the Hyperstructure without the risk of deplatforming. There are no API keys required, and no fear that a platform can suddenly upend your entire project on a whim. (30)

Similarly, this offers censorship-resistance at the individual level, as there is no arbiter or central entity that can decide to restrict access to the utility of the infrastructure.

(30) Sounds like mathematical/physical/scientific principles that can be replicated and instrumentalized at will, like gravity or electricity. But what are the logics of a Hyperstructure? What is the underlying calculus or equation? What are the laws, the “universal principle[s] that [describe] the fundamental nature of something, the universal properties and the relationships between things”? “Law (principle),” Wikipedia.

Credibly Neutral

As a result of being permissionless and unstoppable, Hyperstructures are credibly neutral. (31)

Photo: Pujiang Longsheng Crystal Art & Craft Co., Ltd.

They are user-agnostic. That is, they do not discriminate for or against any specific people and treat everyone fairly. Hyperstructures are universally accessible for anyone to utilize, build on top of, and integrate into their own systems—without any risk of deplatforming or exclusion. Hyperstructures treat every participant fairly, to the extent that it’s possible to treat people fairly in a world where everyone’s capabilities and needs are so different. (32)

(31) This is important, given the power politics of infrastructure as referenced above and in Buell Center, Power (2016– 2021). For a discussion of the inherently constructed nature of neutrality, see: John Palmesino, “Neutrality,” Dictionary of War (2006).

(32) Indeed. And how can Hyperstructures work to address the structural inequalities that often get masked by the veil of difference (but that are perhaps inherent to the idea of “capability”)?

Positive Sum

As a result of being free, expansive, unstoppable, permissionless, and credibly neutral, Hyperstructures create a positive-sum environment.

Giovanni Neglia, Two-person Non-zero Sum Games, 2011. Courtesy of INRIA

By being free forever, there is no incentive for someone to one-for-one replicate the same functionality, since there is no additional value capture or risk minimization that comes from doing so. At a societal level, a Hyperstructure only needs to be built once. This does not mean there isn’t an incentive to innovate: people will still have the incentive to create an alternative with superior and differentiated functionality. But simply forking with the same utility will not win.

By being expansive with fees, Hyperstructures allow for participants to directly profit from the value that they create for the system without needing to compete with the system itself. (33)

By being unstoppable, there is no operational risk for participants relying on the system to function. They can operate on the system without having to rebuild and operate it themselves.

By being permissionless, all of the above conditions are guaranteed to remain unchanged. Which, once again, removes the need for anyone to rebuild the system themselves—you simply create on top of it and capture the unique value that you provide to the system.

By being credibly neutral, there is no group or community that is excluded, which again means there is no reason to rebuild if the system satisfies your needs.

This means that you can have a rich ecosystem of potentially competitive participants organizing on and utilizing the same piece of infrastructure to the net benefit of everyone through shared liquidity and the network effect of integrations.

(33) Can we imagine an anarchist Hyperstructure? What would it mean to apply deconstructivist thinking to Web3 protocols? Maybe there is something to learn from the discourse surrounding the etymological relationship between “arche” (the Ancient Greek ἀρχή, or arkhé, meaning “beginning, origin”) and “architecture” that defined architectural deconstructivism. For contemporary interpretations of the architects and architectural projects that were canonically operational at that time, see: Reinhold Martin, Utopia’s Ghost (University of Minnesota Press, 2010) and K. Michael Hays, Architecture’s Desire (MIT Press, 2009); For other, more philosophical engagements, see: early work by Mark Wigley and Giorgio Agamben on ideas of origin, archon, etc.

Jacob Horne, SimCity 2000, Date Unknown, Screengrab. Courtesy of the internet

Building a Hyperstructure

I’ll write more about the specifics of building a Hyperstructure over time—we’re still learning this as we go at Zora—but here are some quick thoughts that are worth mentioning here:

  • Hyperstructures are made to power millions of interfaces, not just one: when creating a mechanism or anything at the protocol level, ensure it’s as generic as possible. (34)
  • Use fees as a way to expand the ecosystem, not extract from it: while it may be possible to extract value in the short term, this might be a local optima that misses the broader opportunities that come from a diverse platform ecosystem on top of your protocol.
  • Take a protocol-first building approach: focus on developer adoption to create as many integrations as possible—an important network effect and cementing the Hyperstructure as the default.
  • Building liquidity: onchain liquidity makes it beneficial for other entrants to join the ecosystem and benefits everyone else. This is a key network effect and also reduces the ability/incentive to fork.
  • Ownership where possible, governance where only necessary: create ownership and governance where only absolutely necessary. Too much of either, and you may find that the Hyperstructure gets skewed incentives or is at risk of attack.
  • There is a long build cycle: deploying protocols in this manner is much more akin to creating hardware than software. There are long design periods, high-friction migrations, and a high bar for deployment.

(34) Like a road, a pipe, a cable, or a switch. Scale is inherent to the thing one designs. See: Keller Easterling, Organization Space: Landscapes, Highways, and Houses in America (MIT Press, 2001).

Holly Herndon, Children of the Future, 2021, Screengrab. Courtesy of Zora.co

Societal Infrastructure

Given their immutable and permanent nature, we have the opportunity to create civilizational infrastructure that can outlast our own lifetimes—all of which is native to the internet. (35)

We’ve never had the tools to create software-based infrastructure that can work as-designed for generations without degradation or falling to the tragedy of the commons to sustain itself, and we are at a once-in-a-generation moment where we are the ones lucky enough to have the privilege to build the first one. (36)

I’d posit that we’ll likely see a singular Hyperstructure emerge for every financial utility: exchanges, marketplaces, lending pools, options, and so on. The same could be said for non-financial utilities that are currently run and operated by Web2 and social media platforms, like domain names, registries, identity, curation, tags, reputation, emojis, read receipts, and so on. The entire internet is being rebuilt in the form of Web3, with crypto and the blockchain at its core. We have the opportunity to restructure the utilities of the internet of the past 40 years outside of private control and create entirely new ones that simply weren’t possible before.

The scale, importance, and ambition of these pieces of infrastructure will eventually be marveled at in the same way we do museums, power grids, canals, dams, and ancient roads.37 Hyperstructures will be at a scale we’ve never witnessed before, and can be more substantial than even the current social media platforms of today. They will also be judged on their availability and utility to the public over a long-time horizon. A long-term view is critical when designing a Hyperstructure, something that’s often typically lacking in the short-term, profit-minded world of Web3.

Crucially, because of their immutable and permanent nature, we must optimistically create them in their purest, most free, and most beautiful forms. These Hyperstructures could very well be used by our grandchildren, so we have the responsibility to build them with them in our minds— and we should do them proud. (38)

It is now possible to create Hyperstructures that run for free, forever—so we should use that logical extreme as our starting point, not something we should avoid or ignore. With new paradigms comes new value systems, and with Hyperstructures we have the opportunity to create public goods that can remain free forever, whilst rewarding the builders and participants for creating and contributing to these invaluable systems that serve society at large for many years to come. That sounds like something that’s worth building towards and finding out.

Jacob Horne, Diagram, 2021. Courtesy of jacob.energy

(37) How will these infrastructures be seen, though? We can all go to museums. We can visit electrical pylons, canals, dams, and ancient roads. But how can we see these infrastructures as such, and not just the things people build on top of and with them?

(38) And how can we design these Hyperstructures to ensure that future generations have more rights, more freedom, and a better life than we do today? See: the project Rights of Future Generations by Adrian Lahoud (Sharjah Architecture Triennale, 2019).

Hyperstructures Redux

Mint – 0 ETH
Share article
Link copied!