Recently, I’ve been reading some anthropology. The economic canon postulates a primitive economy of barter, where at some point money emerged to solve the problem of double-coincidence-of-wants. This assumption is taken for granted in many places (including, to some degree, in the original Anoma vision paper), but when you consult the historical record, as David Graeber does, it is clearly bunk. Early societies - and small societies today - didn’t swap cows for chickens, at least not most of the time, and they didn’t invent coins to solve double-coincidence-of-wants because they didn’t need to. Rather, they used credit. Credit solves double-coincidence-of-wants in a beautifully elegant way: by integrating over time. If I’m the butcher, and you’re the baker, I may not need bread right now, but I’ll definitely need some in the future. If we live in the same town, and I expect you to be around for awhile, we can simply keep track of who gave who what when, and settle accounts periodically - this requires, of course, likelihood of repeat interactions and sufficient trust, but in small communities these are both present. Still, comparing goods is often helpful, so such communities often picked a particular good to serve as the unit of account and measurement (but with the actual exchange taking place with specific goods in question). The “store of value” function of money, as it were, was mostly virtual, not physical: although some farmers may have had more cattle or grain than others, the most important asset held by residents was usually the trust of the community, with which they could obtain what they needed when they needed it and could more easily handle shocks to their own supplies. This trust is a form of credit, which anyone could issue whenever they wanted to - but if they started taking too much without providing in return, their neighbors may not want to accept their credit anymore. This kind of credit accounting is virtual, tracked not in exact precision with spreadsheets and central banks, but roughly through observation and gossip, and it is scale-free, in that anyone at any scale of operation - both individuals and institutions (such as the bakery) can issue credit - there is no monopoly on doing so.

Most of us don’t live in a world of scale-free credit money anymore, for a simple reason: in a world where money is physical, this kind of trust accounting doesn’t scale. When there are more people interacting in wide economic networks but still finite time, most interactions are with strangers who one never expects to see again. Correspondingly, and thus perhaps unsurprisingly, we now live in a world of fiat credit money (for the remainder of this post I will just say “fiat money”). In a world of fiat money, money is issued by only a few (hopefully) trustworthy institutions, such as governments and banks, and people in everyday interactions trade not personal credit but rather obligations to those institutions. This solves the trust accounting problem between strangers because strangers need merely trust the same institution and the accuracy of the accounting mechanism - they need not trust each other.

But fiat money, as a coordination mechanism, has two fatal flaws.

First, fiat money centralises trust and thus loses fault tolerance. Because of the network effects of unit of account, store of value, and medium of exchange, the difficulty of creating appropriate accounting mechanisms, and the tendency of warmongering states to create silly laws, the issuance of money is limited to a very small number of institutions, and control over those institutions becomes a central point subject to fierce competition. Even a small number of actors prioritising personal gain over the public good can seize control over the mechanisms of money supply and redirect parts of it for their own private ends. Even if they don’t succeed in doing so, the negative externalities to public sensemaking created by elites competing for control over the money supply can so pollute discursive spheres with “alternative facts” that regular mechanisms of social feedback and coordination no longer function at all. Only with the decentralisation of trust can fault-tolerance be achieved.

Second, fiat credit money relies on measurement in the present. In order for me to pay you using a debt instrument, when we don’t trust each other, we must agree on the trusted third party and the exact amount of payment such that we are both willing to walk away after the transaction and never expect recompense from each other again. If the primary benefit of the good paid for is easily estimable, limited to the party purchasing it, and itself in the present (such as a sandwich), this isn’t too bad an approximation to future value, but if the primary benefit of the good paid for accrues over time, also to other parties, and in the future (such as knowledge), this is an absolutely awful way to measure it.

Many of the dystopian elements in our world today can be traced to these two fatal flaws, I think. Wars, climate change, nuclear proliferation, lack of public education, pollution of the information commons, and similar phenomena source their causes in large part from awful decisions made by elites competing for government positions or propaganda so created (such as convincing their citizens to pay for nuclear weapons), a result of this centralisation. Failures of public-goods provisioning, in turn, are a result of financial measurement happening almost solely in the present.

By contrast, scale-free credit money decentralises trust and shifts measurement to the future. My credit is valuable to you if and only if you expect me to be able to pay you back in some form then, since I can offer nothing now. While parties who have exchanged debt can just walk away, parties who have exchanged (perhaps heterogeneous) credit have a mutual interest in each others’ future success. If I teach you something and you pay me in debt, I don’t care whether what I teach you is right or wrong, I just care about convincing you to pay me more money. If I teach you something and you pay me in credit, I want to teach you something right, and useful, so that your credit will be valuable to me in the future.

In our world today, the world of fiat credit money, trust and money are misaligned, so misaligned that they have inverted - at least personally, on average, I don’t trust people with more money or institutions which can issue money, more - I trust them less. To re-align them, we must re-align control over the issuance of money with trust, and return to a world of scale-free credit money. But, you might ask, why would this work now, when it hasn’t so far?

Now comes the point in the story where I offer you a pill - but, I promise, the pill is purely conceptual - a thought experiment: what would a world of scale-free credit money look like?

What would a world of scale-free credit money look like? In our world, the issuance of money is strictly controlled, generally only available to governments and specific entities to which they delegate the ability such as banks. If money is to be credit, re-aligned with trust, then these restrictions don’t make any sense, since trust is distributed, and credit is personal, so let’s change a few basic assumptions. Assume that everyone can print money, whenever they want, as much as they want (though they could voluntarily limit their ability to do this), and send it whomever they want. Assume that individuals and institutions everywhere, now and in the future, are constantly creating new denominations of money. Let’s also assume that denominations of money are content-addressed: money is defined by who (cryptographically) currently holds it, who (cryptographically) can issue it, and how much and under what conditions they can issue it (in case of self-limitation). Local name-systems and consensus algorithms handle human-readable mapping and continuity in time.

In this hypothetical world, so far, money isn’t actually useful for any sort of coordination, because everyone is using a different money. How could such heterogeneous instruments function as a store of value, a unit of account, or a medium of exchange? To be fair to economists, they aren’t always wrong - those things are useful!

Let’s add an ingredient. The agents printing money aren’t just individuals - they may also be institutions which want to provide functions such as store of value, unit of account, and medium of exchange for their constituents in some area of space (be it digital or physical). But in this world of scale-free credit money, money is very competitive, since anyone can switch which money they use at any time, so institutions wishing to issue money which will get used must create an initial distribution and issuance schedule which its potential users consider to be fair (and thus select) - and this problem, in essence, is none other than effective retroactive public-goods funding (another entry to add to the list). Money selected by a group in such an area, if the group can agree on a money, can serve as store of value, unit of account, and medium of exchange - but what if the institution starts sending the money somewhere people don’t like? Someone else can simply come along, and issue another money, copying (and perhaps altering) the distribution and changing the recipients.

You might reasonably object - switching costs are non-zero! Imagine having to re-price all goods in the supermarket or re-issue all coinage - these are expensive operations. In a world where money is primarily physical, switching costs are high, but in a world where money is primarily digital they are not. In a world where money is digital, store of value, unit of account, and medium of exchange can be easily disaggregated by automatic price conversion and exchange.

In this world of scale-free credit money, where new money is perpetually created, most of the potential money exists not in the present, but in the future. Competition for value in the present is based not on scarcity and measurement but instead on expected inclusion in potential future retroactive funding distributions. Expected inclusion in future retroactive funding distributions, since money is subject to competitive selection, is based on what value people and institutions in the future think contributors in the past (our present) ended up providing to the present (our future), since it is precisely this allocation which people in the future will consider most fair.

Now, you might ask, in this world, with infinite money, how do we track physical, scarce goods? Physical goods are costly to produce and (at least compared to digital goods) tend to offer most of their value to a private party in the near future. Current systems of payment accounting do a decent job of organizing the production of physical goods, so regular old payment at time of receipt seems like a reasonable solution to me. Individuals and organisations producing physical goods can keep operating in this world much as normal, simply accepting credit of parties they trust instead of sovereign debt.

Payments for physical goods also benefit from stability of the unit of account. In this world of scale-free credit money, the requisite stability can be provided by money options with self-imposed issuance controls. Institutions issuing money which they expect to be used by large numbers of individuals can self-impose limits on issuance rates such that issuance cannot exceed more than a few percent a year (comparable to existing targets of central banks), keeping the unit of account pretty stable.

Now, you might come back with a certain gotcha: what about interactions with untrusted parties? After all, these comprise most of our interactions today. Travel to, interactions in, and commerce with faraway places are pretty nice - in such a world, would we need to give these up? This problem, mind you, only comes up for payments in the present - future retroactive payments, by definition, are given only to those who one trusts, based on their work in the past!

Time for the mechanism wizardry to come in handy. Let’s assume some liquidity in credit markets, such that any issued money is freely exchangeable with any other whenever someone wants to create some liquidity. Now, if I want to pay you and we don’t trust each other, all I need to do is find a route in the liquidity graph between us. We need not use the same unit of account, store of value, or means of payment to interact with each other anymore - all we need is a connected path. Of course, not all paths are equal - if a lot of liquidity exists between us, I can pay you a lot without changing the price much, but if little liquidity exists, I can only pay you a little - but this reflects precisely the density (and directionality) of trust!

Retroactive public-goods funding for public benefits, measured in the future, and payments for private benefits, measured in the present - so far, so good, sounds pretty nice to me at least.

But the skeptics among you may have one final objection - I certainly would - this sure sounds like a whole lot of financialisation. Imagine that everyone’s credit was traded - wouldn’t we just compete in an endless game of self-marketing to win the war of usage? I think scale-free credit money substantially reduces the network effects of money compared to today, since it removes the need to come to consensus on what specific money to use in any given interaction, but some network effects will remain. Also, there are clearly a lot of new forms of money at the moment (just look at the list here), and they sure do seem to spend a lot of time, effort, and money (somewhere, somewhere here, there is a trap…) competing with each other, which all seems a bit zero-sum.

Here comes my final bit of mechanism wizardry: commitments to future integral airdrops. Airdrops are a common mechanism in blockchain circles already, and they are already usually used in the attempt to proliferate a new form of money, but as currently deployed they have one fatal flaw of their own: centralisation in time. Airdrops have historically targeted a snapshot of a particular token at a particular point in time, which creates a discontinuity in the incentive space: it is valuable to hold the token right before the airdrop snapshot date, then suddenly less valuable right after. I propose a slight alteration: instead of taking a snapshot in time, take a snapshot over time - an integral of who has held how much from the beginning of time (or, well, the token in question) to now.

Future retroactive funders, by doing integral airdrops, encourage parties who wish to receive such airdrops to purchase the relevant credits (and support the relevant party doing the real work) early on, so the integrated amount over time will be higher even if the prices fluctuate. Integral airdrops can be safely committed to ahead of time without creating odd discontinuities in incentive space, and they can even be performed repeatedly to continuously align incentives. In expectation, complexity is greatly simplified, because if you’re right about what will be valuable, the optimal strategy is just buy-and-hold.

At the moment, in the world today, money and trust are anti-correlated, because control over the issuance of money is controlled by actors who almost nobody trusts. It took me a long time to come to the understanding presented in this blog post for this reason, I think, because I hated dealing with money so much that I was very hesitant to design any sort of system that used the abstraction of money at all (especially when it involved a lot of measurement), and I initially tried to avoid it (bad idea, it turns out that this just leads to way more measurement complexity…). But once you align money and trust, even merely abstractly in the design of a sociotechnological system, dominoes magically start falling into place as if it were all pre-ordained.

One common problem in cryptocurrency systems is that of key recovery. Cryptographic keys are weird things, strings made up out of nowhere, and most people forget them or lose pieces of paper (I certainly do). Designs of social key recovery systems propose that we mark specific combinations of friends who will be allowed to recover our keys, but, while better than no key recovery at all, this solution requires a lot of awkward manual interaction to specify and update this trust graph, and it’s difficult to know exactly how to choose the right people anyways, since whom one trusts changes over time.

Yet if we align keys, trust, and money, a solution naturally emerges. Key recovery requires trust, so we must choose someone - or someones - to trust. Who better to help me recover my keys than whoever holds my credit? Our incentives are very well-aligned - they want me to do well so that the credit which they hold is valuable in the future, and I’m more likely to do well if I can access my account (which holds many other credits, and allows me to issue more)! All we need is a threshold, and straight out of the distributed systems hat we can pull 2/3, which ensures both that the relevant parties can safely come to consensus on my new public key and that I can still recover my key if less than 1/3 are offline - which, in a dense network where key loss is uncorrelated, will almost certainly hold in practice.

Another sought-after pair of hypothetical protocols are those which would enable universal basic income and proof-of-humanity. I mention these as a pair because I think they are hung up on the same problem: what does it mean to be human? It is impossible to design a test that can distinguish a human from anything else, because there is no essence to humanity: I am human only insofar as you consider me to be. At various points in history, laws have classified certain groups of people as sub-human, even assigned them numerical fractions, which to us today seems abhorrent. I think, correspondingly, the animating idea of universal basic income is that of equality, and equality in the eye of the beholder requires two parties who both consent.

These desiderata are two sides to the same coin, because there is no test but equality, and equality on the basis of humanity must be decided by humans. We could each keep a list of other humans’ public keys and all pay each other equal amounts of our own scale-free credit money every second, but this requires too much interaction, doesn’t provide any future predictability (which is perhaps the primary benefit of UBI) and fails to make use of a property we do assume of humans: that they carry information, identity, and cryptographic keys forward in time.

Instead, I propose a slight modification, based on the twin bases of this bilateral test of humanity and continuance of humanity in the future: heterogeneous UBI. We only need one ingredient: trust (and some cryptographic signatures). You and I meet in person, decide that we trust each other, and both cryptographically sign over a commitment to the continuous creation of 1 of each of our respective credit tokens per unit time. These tokens could just be sent to each other, but I think there is a better solution that will immediately create some “trust liquidity” and allow for revocation in the future: deposit both tokens in a multi-signature account which in turns locks them to an automated market maker curve on xy=k (or similar) with a special value k = 1. This immediately allows others to trade through us, and allows our relationship of mutual humanity to balance out other inequities in the network.

Each party can unilaterally sign a message to the multisignature account which will cause it to withdraw the liquidity and burn both credit tokens, so should you decide in the future that you no longer wish to trust me, you can revoke that trust, but if others still trust me, I will still have “trust liquidity” with them.

Of course, anyone can create a non-human cryptographic identity, and start printing money with it, but unless they can convince someone else to trust it, they won’t get any extra liquidity, since all paths in the liquidity graph would have to go through them. No one would want to commit to exchanging their credits with that of a fake identity, since they have no reason to expect anyone else to want those! The attacker could bribe someone else to trust it, but they’d have to bribe enough to make the supply inflation (of the bribed party’s token) worth it, so they’d just end up paying UBI to the bribed party themselves.

From this currency network, we can a derive proof-of-humanity test for any two parties (since, of course, it is relative) by proving the existence of many different separate valid linked-list paths of bilateral signatures over these commitments (with no overlap in the member public keys apart from the beginning and the end), which would not exist for an isolated network subgraph (since, as explained above, they are expensive to create).

Scale-free credit money and heterogeneous UBI can be issued with existing protocol primitives roughly as follows: smart contract accounts for each issuer (since they might still want keys on multiple devices with different spending limits so key recovery is only invoked when absolutely necessary), smart contract accounts for the bilateral humanity test liquidity locking relationship, Uniswap-style AMMs to facilitate exchange, multi-hop swap routing to find paths through the credit liquidity graph (a la Circles UBI), a blockchain to order transactions and prevent double-spends, and recursive ZKPs for retroactive integral airdrops.

It’s worth noting that privacy is essential for scale-free credit money. If trust isn’t private, it’s possible to threaten someone for trusting someone else, but if trust is private, it isn’t anymore. To provide the requisite privacy, all of this must be implemented on a fully private substrate, likely with ZKPs for the individual accounts and some threshold FHE for batch swaps, liquidity provisioning, and trust-minimised private bridging.

While the details remain to be worked out, I don’t think anything here is too difficult to understand or implement. Scaling to a liquidity graph of 8 billion people will take some time, but it doesn’t have to all happen on one blockchain - only the ordering of spends from each individual account needs to be total (no double-spends), everything else can be partial - and ordering is expensive, better to only do it when necessary. Besides, one “megablockchain” would have to pick one security model and one currency which controls it - and that sure seems like a point of trust centralisation to me. We can all recursively verify state transition & consensus history anyways.

To somewhat misappropriate Foucault, let’s call this world of scale-free credit money heterotopia. To Foucault, a heterotopia is a place outside all places, a place which is real but a place where regular operations of society and culture are represented, contested, and inverted - cemeteries, zoos, and fairgrounds are all heterotopias. The sense of heterotopia I mean is not quite this one, which conceptually demarcates precisely those places which offer a temporary interstice from everyday cultural rules of place. Rather, my sense is of a heterotopia which is at once total and fragmentary. Consider Foucault, from the end:

Brothels and colonies are two extreme types of heterotopia, and if we think, after all, that the boat is a floating piece of space, a place without a place, that exists by itself, that is closed in on itself and at the same time is given over to the infinity of the sea and that, from port to port, from tack to tack, from brothel to brothel, it goes as far as the colonies in search of the most precious treasures they conceal in their gardens, you will understand why the boat has not only been for our civilization, from the sixteenth century until the present, the great instrument of economic development (I have not been speaking of that today), but has been simultaneously the greatest reserve of the imagination. The ship is the heterotopia par excellence. In civilizations without boats, dreams dry up, espionage takes the place of adventure, and the police take the place of pirates.

Modernity no longer possesses any boats - and not only because there are fewer treasures to plunder - those former “heterotopias par excellence” have been instrumentalised into costs per kilogram-kilometre and transport APIs, organised and regulated on the basis of the totalising dollar. The heterotopia I mean is a heterotopia of value, tracked and organised in a purely virtual space, itself fragmented into fractal Venn diagrams of partially overlapping sub-spaces. Foucault’s heterotopia implicates the existence of a dominant set of cultural practices and dominant set of spaces to which it is is semantically opposed, but a heterotopia of value presumes no particular ordering of spaces, merely a plurality of difference.

We don’t live in heterotopia right now - we live in a world trending towards dystopia. Heterotopia isn’t utopia - people will still disagree, accidents will still happen, broken hearts will still ache - but I think that it’s better compared to this world, because it correctly alters the cultural and technological substrates of money to match the interests of humans in the future and the corresponding shared interests of humans today. Heterotopia is not merely a matter of the mechanisms of money - money should be a small and insignificant component of the societies, activities, and traditions of a culture or a space - but as it is a marker of the failure of our current form of money that it is not, so I shall focus here on the mechanisms of monetary transition.

Some might worry about states, which in (only) recent history have tightly controlled the issuance of money, and may react violently to the possibility of heterotopia. While I share the fear of state violence, I think this worry is easily overblown. Although the monopoly of the state may appear to be physical, in truth it is purely conceptual: as soon as we stop believing in it, it will simply go away. Heterotopia shatters this monopoly to bits (with nothing but bytes). Who in the future would want money issued by an organisation which rounded up people crossing imaginary lines drawn on a map and sent them to camps, hired armies of consultants to propagandize its own supposed constituents, and kept the world under decades-long threat of nuclear holocaust? Not me, for sure. If they want to survive in heterotopia, states had better stop locking people up and start producing some public goods instead. Some governments may try to prevent heterotopia from ever arriving by exercising coercive force, but in heterotopia, money is merely information, and information is always a moving target, which no bureaucratic mechanism can ever keep up with - indeed, with content-addressed naming systems, no bureaucratic mechanism can even name it. You can fight the laws of physics, but only for so long.

I think that heterotopia is likely. Information systems tend towards more stable states, and our world today is not in a stable state at all, in no small part because money and trust are so misaligned. A state (but not the other sort) which better aligns them is likely to be far more stable. But that doesn’t mean that the transition won’t be turbulent. In particular, existing communications infrastructure, which lacks a sound cryptographic basis of identity and network of trust relationships, is very vulnerable to propaganda, and global sensemaking is deluged by bad-faith noise. “AI” (fancy statistical models) may have excellent uses in artistic creation, but its use in propaganda generation (and, unbelievably, in tools) is making this problem worse quickly.

So, take the rest of this post as contingent on the hypothesis of heterotopia - if it is coming anyways, what can institutions do to ease the turbulence of the transition?

First and foremost, institutions must coordinate the creation of the requisite technological substrate - research, protocols, interface, open-source software and hardware - to make the realisation of this heterotopia vision of scale-free credit money possible. Existing blockchain / cryptocurrency protocol designers & organisations are well-equipped here (excellent candidate examples include Aleo, Anoma, Celestia, Cosmos, Ethereum, Osmosis, Penumbra, and many others), but they need to coordinate with and help out the fine distributed-systems-but-not-blockchain folks working on end-to-end encrypted messaging, properly distributed social media, local-first application design, and generally self-sovereign & privacy-preserving applications (excellent candidate examples include Ink & Switch, Mastodon, Scuttlebutt, Signal, Urbit, and many others). Open-source and verifiable hardware remains quite far away and could perhaps be accelerated by strategic acquisitions followed by application of free software principles similar to those articulated by the FSF to the relevant hardware IP. Cryptocurrency foundations tend to have a lot of capital and should use it to do this instead of throwing money at Uniswap-clone-subsidy grants programs or sponsoring ads on Formula 1 cars. Of course, hardware companies could also pre-emptively do this themselves in hopes of future retroactive funding - which, contingent on the expectation of heterotopia, is an excellent strategy for them!

Second but equally important, institutions must provide stability. Even in the better possible worlds, the path from today to heterotopia will be marked by wildly fluctuating exchange rates, rapid changes in monetary policy, and attempts at the overreach of state power. Institutions can dampen the shocks to their constituents by hedging these risks: holding multiple currencies, committing to inversely adjusting salary payments to inflation rates or adjusting them for real costs-of-living, funding legal defense of individuals targeted by governments, and so forth. Institutions which successfully provide a buffer to these shocks can expect retroactive distributions to include them in the future, so there is reason for them to try. By and large, existing legal structures are already designed to allow institutions to assume risk (“limited liability”) and hold assets, so existing institutions should be able to assume this role easily.

Institutions which are able to can shift future expected value into the present by issuing credit money as outlined above, which can be sold for existing currencies (especially fiat) to fill their institutional coffers and increase the institution’s ability to buffer shocks. The difficulties of doing this within existing legal structures will vary, but institutions which are able to find a way, shift future expected value into the present, and use it correctly to support individuals can similarly expect to create more future value and receive future retroactive funding, so their incentives are well-aligned.

In order to coordinate towards heterotopia, institutions can establish bilateral relationships of trust with other institutions. It is important that these relationships of trust are publicly verifiable, as this allows parties operating within or otherwise aligned with those institutions to coordinate more efficiently (e.g. deduplicate work). This can function quite similarly to the proposal for heterogeneous UBI as above, but instead of setting k = 1 (not a natural consensus in this case) and committing to redefining the future issuance schedule, institutions can just periodically agree to mint some of each other’s token and lock it in an xy=k curve at the current value of k, which correlates the value of the two institutions’ expected future retroactive funding and current credit, and can be publicly verified if they choose to make it so.

Institutions operating on the conceptual basis of heterotopia should also selectively extend trust (and money, in whatever forms applicable) to existing legacy institutions. Existing institutions have their conceptual frameworks and reputations embedded deeply in existing societies, and cooperating will likely dampen some of the turbulence of this transition. However, this trust (and money) should not be extended without conditions. Many existing institutions directly or indirectly fund weapons, propaganda, and coercion, and funding should only be extended to these institutions (such as, but not limited to, fiat governments) conditional on their agreement to immediately defund all such activities, adherence to which must be checked periodically. Existing institutions have issued a lot of money but lost a lot of trust, and that trust must be re-earned if they want their money to be valuable in the future. This is incentive-aligned, because scale-free credit money is win-win - it need only oppose those who would oppose others. Existing institutions which cooperate can expect future retroactive funding, and existing institutions who do not can expect none.

Certain existing institutions could easily retool themselves to rapidly accelerate this transition because their skill-sets and assets can act as an incentive shift force multiplier. Venture capitalists, hedge funds, and other private equity firms which retain direct decision-making power over their capital allocations need simply start optimising for provisioning public goods instead of private ones. Optionally, as above, they can issue their own money in expectation of future retroactive funding, but retroactive funding can also be issued to owners of existing stock, equities etc. through interfacing mechanisms, so this is not critical (and may be difficult for some capital allocators operating within existing legal structures).

For existing capital allocators, as soon as they expect heterotopia, this is incentive-compatible, because optimising for private value capture is a terrible strategy from the standpoint of capital efficiency in public-goods provisioning. Public goods, by definition, are non-rivalrous and non-excludable. Existing schemes for conversion of public value to privately capturable value are based around the imposition of an artificial mechanism of exclusion, such as paywalls, token-gating, or IP law. Such exclusion limits the amount of potential future value - and corresponding expected future retroactive funding - because fewer parties are able to access the good, benefit from it, and (potentially) be thankful for it in the future. This exclusion also imposes the requirement to measure each usage (signing into a website with a paywall, paying to view a movie, licensing a theater production) which has a cost of implementation that scales (to first order) linearly with the number of consumers of the good, since each use must be tracked - the more users of the good, the more potential future value, but the more expensive such measurement. Optimal capital efficiency in public-goods provisioning is much more likely to be achieved by measuring only infrequently, as much as is needed to sample demand and coordinate strategic direction of production, but not for each and every interaction. For this reason, in expectation, after the transition to heterotopia, capital allocators which changed their decision-making calculus earlier on can be expected to do better (in terms of retroactive funding) than those who did not, since they will have created more public value. Considering the social dynamics involved, one can expect a sort of preference cascade here (and in general).

Enough for one blog post. Permit me the indulgence to end on a more poetic note. To quote Mao Zedong, who may have not been quite right at the time but said it more eloquently than I can,

所谓革命高潮快要到来的“快要”二字作何解释，这点是许多同志的共同的问题。马克思主义者不是算命先生，未来的发展和变化，只应该也只能说出个大的方向，不应该也不可能机械地规定时日。但我所说的中国革命高潮快要到来，决不是如有些人所谓“有到来之可能”那样完全没有行动意义的、可望而不可即的一种空的东西。它是站在海岸遥望海中已经看得见桅杆尖头了的一只航船，它是立于高山之巅远看东方已见光芒四射喷薄欲出的一轮朝日，它是躁动于母腹中的快要成熟了的一个婴儿

I won’t include a translation because it doesn’t do the text justice, but suffice it to say that rays of heterotopian light are already visible through the cracks in the facade of modernity. As soon as you start looking, you can see it everywhere you look, from calls for a “Game B”, to elucidations of the social media dystopia likely without a cryptographic substrate, to expositions of how a category-theoretic treatment of economics would discourage the conceit of “ceteris paribus” decision-making, to random conversations about dysfunctional speculation-fueled property markets overheard in a Mexican restaurant in Kreuzberg. You’ve already seen it. Maybe you’re already living it. I didn’t invent it, you did. I just gave it a name. And I didn’t even do that.

Heterotopia is inevitable the moment we decide to make it so.

Now, for a little retroactive trust-funding of my own:

Thanks to Ding Yuanfang for the 103% inspiration without which even nothing wouldn’t have been possible. Thanks to Awa Sun Yin, Adrian Brink, Dev Ojha, and too many others to name for conversations and cooperation without which this post certainly wouldn’t have been. Thanks to Yuanfang Ding, Nicholas Campbell, and Dev Ojha for reviews of this post and helpful suggestions - all errors remaining are my own.