The Stack On Software and Sovereignty

Put differently, that situation is also characterized by an opposition between the global plasticity of the second planetary computer and the Anthropocenic crisis of ongoingness. While it may be that our position on the precipice of ecological collapse suggests an anti-cosmopolitanism based on sharing this G ö tterd ä mmerung, different economies have contributed to its inadvertent geoengineering in very different ways and it is certain that how some groups accommodate the coming crisis will be as equally pronounced and divided. It is sometimes said by the political right that the dangers posed by climate change are overblown by leftists keen on using centralized anti-market regulatory measures as a matter of principle and eager to mobilize them in the name of bad weather from the future, but the political opportunism of climate-change denialists represents the more ominous teleology. The longer that effective interventions and mitigations are forestalled, the more catastrophic the eventual outcomes will be, and the less likely that open and democratic societies will be able to manage the deluge of life or death consequences. Many on those the political right are well aware of this, but see it as an opportunity for the development of quasi-sovereign enclaves in which those who can will consolidate their wealth and deploy it as a bulwark against both the ecological consequences of industrialization and the populations that these effect most dramatically. For those who would prefer neo-Feudalism and/or tooth-and-nail libertarianism, inaction on climate change is not denialism, rather it is action on behalf of a different strategic conclusion.

Too often, notions of ecological cosmopolitics rely on rhetorical criteria of consensus, whether as a lifeboat ethics — that we are all in this together — or the supposed self-evidence of Earth ’ s archive seen as a single space that can be made more communitarian. However, while it draws on an ethics of rationality, the geodesign I would endorse doesn’t see dissensus as an exception to the norm but as a matter of fact (nor does it see it as only and necessarily “ political ” ). It doesn’t confuse the universality of a shared planetary ecology with how its political geography subdivides sovereignties, ones that are not only separate but may not even be interoperable. It is not that we will split the middle but that, quite specifically, we speak different languages, live in different worlds, and have different geophysical relations to capital as a design asset in relation to emergencies. That said, the current ecological emergency extends, rather than supersedes, the importance of totality as interpretive instrument, especially with regards to planetary-scale computation. When I peruse with fascination all the rightwing conspiracies theories about “ Green totalitarianism, ” Agenda 21, and carbon police confiscating McMansions, I can ’ t help but wonder if the right has a more active and forceful vision (however goofball) for the left than the left does. At the same time, we are sensitive that just as well-meaning initiatives too often sour into failure (and wonder if we should instead design evil things that will decay into something wonderful and successful) we know that the anti-strategy of neo-feudal secession may, however counterintuitively, usher in a geopolitics that provides a platform that — in time — may unwind the social hierarchies that inaugurated it. Who knows? They may get what they want, but they may not want what they get. ¹⁸ With regards to this, the next section considers possible futures of the next layer of The Stack, and how the delimitation of sovereignty from territory by global Cloud platforms, and the differing models of Cloud Polis that they represent, supports the momentum of generalized secession (for better and for worse) which may or may not evolve toward (away from) an institutionally normative Cloud feudalism .

306

The Stack to Come

1. Cloud Layer to Come: Cloud Feudalism and Its Discontents

Ground zero of robotic labor automation may be in California, but a few hundred miles south and east of Silicon Valley in the San Joaquin Valley, where 1 percent of the US farmland produces 8 percent of its agricultural output, generating at least \$36 billion for the state and over \$100 billion in related economic activity. The area is the engine that makes California the fifth largest supplier of food in the world, and its capital is the crestfallen metropolis of Fresno (population 500,000). Fresno often comes near last in urban quality-of-life polls, and even a quick visit to its dusty and mean avenues confirms the results. Roughly a third of all jobs in the city are tied directly to agriculture, which makes the economy particularly vulnerable to downward wage pressures, as well as to climate change – related drought. The pressures of increased efficiencies against crop losses, toward crop diversity, and speed of delivery make agriculture an important area for applied robotic automation (picking, sorting, transporting, as well as plantby-plant drone observation and diagnosis). These factors together nominate Fresno to enjoy first-mover disadvantage in the evolution of similar urban centers toward broadband-dependent manoralism and serfdom (and evacuation) and its economy toward Cloud feudalism. Detroit is the first case study we have of what automation can do to an insufficiently diversified urban system, dependent on intensive assembly labor (it is certainly not the only one, nor does it represent the inevitable outcome of physical computing as applied to infrastructure-scale manufacturing economies), but Fresno may soon take its place next to it. Future scenarios for Cloud feudal life in Fresno are mostly grim. Most remaining jobs might be related to servicing the automated logistics and warehousing of food packets, not so dissimilar to working in an Amazon warehouse or FedEx routing facility, while the surplus population that has not or cannot exit is largely unemployed and increasingly desperate. ¹⁹

The Cloud is able to treat “ food ” — defined here as culturally and economically desirable modules of proteins, vitamins, and sugars — as parcels or data packets in a deep supply chain that incorporates local climate, soil, nutrients, seed systems, individual plant care, harvesting, sorting, warehousing, packaging, refrigeration, and global product destination optimization, as well as the metagovernance of demand modeling, crop diversification, qualitative and quantitative research, and, one hopes as well, the geogovernance of nutrition and food health (this is to say nothing of multistory hydroponic megastructures growing onions, orchids, and okra in deep midnight). The intelligent industrialization of food is potentially an extremely positive (even crucial) Anthropocenic strategy. If Heidegger self-servingly compared industrial agriculture to “ Auschwitz, ” and today ’ s biofundamentalists sometimes use similar imagery to demonize genetically modified organisms, we may prefer instead to see the design of food platforms as less about preserving the experiential simulation of preindustrial farming and eating ( “ A is for Apple ” ) and more like molecular gastronomy on a landscape scale, inventing amazing new forms from the calculative slurry. In their seams of leisure, craft farms can be left to those with the time to indulge pornographic fantasies of prelapsarian originalism. That is to say, Cloud feudalism is neither desirable nor inevitable, even according to the processes that are trying hard to birth it today. ²⁰

The Stack to Come

307

None of this gleaming promise screens out a reality determined by vastly inequitable relations to the wealth of the Fresno that we have and the Fresno toward which we slouch. In time, cities such as these may not need so many people but will have them nevertheless. They may not really be cities so much as city-shaped refugee camps, and like all other camps, they are the inverse image of the enclaves that spawn them. Those employed “ in ” Fresno may not even be inside city limits. If we follow the thread of Alex Rivera ’ s Sleep Dealer (2008), a film in which California ’ s agriculture is served by drone pilot/robot fruit pickers working remotely from behind the sovereign wall separating the United States and Mexico, it is not unreasonable to imagine a further logistic dehumanization of Fresno ’ s on-site population. ²¹ Perhaps the costs of piloting agricultural labor will be held down by global wage arbitrage, pickers in Tijuana competing with pickers in Jakarta and Juneau to provide fast and cheap results. That is, formal national jurisdiction may have far less to do with the economics of Cloud feudalism than with whichever Cloud Polis, enclave platform, or urban camp happens to counts a given worker as one of its Users. The elevation of labor systems like Amazon ’ s Mechanical Turk, TaskRabbit, and Uber to infrastructural scale suggests several paradoxical and even contradictory outcomes, both positive and negative. One of these is well summarized as: “ I ’ m really looking forward to a future in which service employees are leased Google Glass so they can complete courses in for-profit trade schools while simultaneously earning health care vouchers instead of actual currency and Soylent instead of actual food. ” ²² We should add, however, that the lease terms on that Glass set are conditional on whether the User actually won the bid to pilot-pick avocados.

Cloud feudalism can be understood as a particular distribution of power between central and commanding platform servers and quasi-autonomous, if relatively powerless, network clients as applied to human economic geography. Others have articulated the problems associated with these sorts of arrangements, their deflationary impact on demand-side growth, and their ultimate macroeconomic instability. Under such regimes, platform economics works to monopolize power and wealth into centripetal consolidations of extracted value, such that the ratio of value realized by those Users who collaborate with the platform commons ( User platform value) to those who own claims on infrastructural profits (platform surplus value) is grotesquely misaligned. For political dispositions across the ideological spectrum, the supposed solution is making freedom of personal autonomy from platforms more absolute, up to and including generalized secession. From one side, “ the main achievement of the nation-state in the last century has been the establishment of a uniform grid of heavily policed barriers across the world. … Given a free movement of people, the whole neoliberal project would collapse. ” ²³ From another side, “ If you value freedom, then I think that ‘ exit ’ comes out way ahead of ‘ voice ’ as a mechanism by which people can express their preferences. ” ²⁴ Under some real-world situations, the hard partition keeping people in or keeping people out is itself given sovereign status, with its surface imbued with maximal gateway intelligence, and in others the “ elective ” coordination of free market actors and agents is seen in a proliferation of modular desiring machines, like a North Korean stadium pageant without an actual country behind it, all decisions linked by an ontological proletariat writing the rules of proprietary semantic webs. If everyone (in principle) has the right of exit and to opt out of their citizenship end user agreement for another offered elsewhere, but all the good spots have already be taken by high-end Cloud polities that feature strong exclusionary membership bylaws keeping the plurality of humans at bay, then the differences between state violence on the border and posted rules of the gated community, between positive and negative freedom essentially, are dark and bitter comedy.

308

The Stack to Come

The previous discussions of nomos emphasize the arbitrary but meaningful geometry of political geographic subdivisions, including secession. It begins with an image of the Earth ’ s surface differentiated by chemical arrangements of water, land, and air. Nation-states can be read in relation to these as petroglyphs written by law. They are keen to concretize the integrity of virtual boundaries when ingress and egress are seen to compromise asymmetries of power between zones of a synthetic landscape of loops and bubbles. Enclaves, exclaves, and especially colonies drew another jigsaw on top of the first, nonlinear and noncontiguous outlines of sovereign control rule. ²⁵ At various times, this becomes a momentum for jurisdictional integration (as for the American colonies in the 1770s, the European Union in the 1990s, Italy and Germany in the nineteenth century), and at other times it is an equal momentum for disintegration (as for the United States in the 1860s, India/Pakistan in the 1940s, the USSR in the 1990s, and seemingly everywhere today). It takes different forms and seeks different ends: schemes to break up California would mean political autonomy for Silicon Valley and several more Senate seats; Special Economic Zones freeing up markets for commodity assemblage by keeping hands and fingers on call in special factory camps; sovereign wealth funds turning states into corporate actors; Supreme Court rulings turning corporate actors into holders of religious and political speech rights; neo-Confederates once again taking control of one of the major US political parties; Saudi Arabia buying sovereign farmland in Indonesia to secure its food future; the hard geopolitics of ongoing state-sponsored spying, hacking, and mutual recriminations; and so on. The great migrations, psychodemographic segmentations, and biopolitical wall building are not always about claiming more and more zero-sum territory at the expense of rivals. Sometime it means the opposite, even deliberately shrinking domains of sovereign interest (Australia recently relinquished its sovereign claim on a nearby island so that it would not have to take responsibility for refugees who landed there hoping to gain admittance to the jurisdiction and its protections). However, these examples of generalized secession are not interoperable with one another, at least not in the same way that the Westphalian platform for political consolidation made the form and formats of states compatible. Instead, according to pressures of combination and separation, integration and disintegration, bull markets in both nonpolarity and hyperpolarity collaborate to assassinate unipolar geopolitical architectures.

The Stack to Come

309

Most are interested not only in drawing new legal lines for self-interested benefit, but also in building the armature of occupation that will enforce those lines and give their physical boundaries the force of law. These initiatives thereby spawn their own exceptions. China has several important urban jurisdictional anomalies, such the special status of Hong Kong, Macau, and the Shanghai foreign concessions, and more recently, its urbanization has been steered somewhat by a strict distinction between urban and rural systems and populations, most clearly symbolized by the houkou license system, which also generated a huge population of internal “ illegal aliens. ” The growth of new cities, almost from scratch, but mostly on top of where old ones stood a few years ago, drew on the fungibility of enforcing this credentialization of infrastructural access. Elsewhere, a new kind of tabula rasa urbanism is seen in the charter city movement, as evangelized by New York University economist Paul Romer. The city remains a crucial site for challenges to traditional spatial models of sovereignty and innovation in what might augment or displace them, and Romer has advised plans for a “ charter city ” in Trujillo, Honduras, one administered by the courts of Mauritius and, in principal, open to qualified persons who may choose to reside there. Here individual sovereignty is derived from the access to and use of a common urban infrastructure network more than autochthonous genealogy. Such schemes are certainly not without the dystopian potential of the absolute capitalization of habitat, but at the same time, in the delinking of transient sovereignty from fixed geography, they also contain an important kernel of utopian potential for geodesign invention, one that instead of eradicating place would reestablish it anew for a network society. ²⁶ One hopes so, but the more immediate context of contemporary sovereignty markets would seem by appearances to tilt toward privatopias.

310

The Stack to Come

Neil Blomkomp ’ s film Elysium (2013) is based on the parable of two Earths, one wallowing in the Anthropocenic decay of a burned-out planet and one orbiting above in bucolic (if run-of-the-mill) privilege. ²⁷ Like the Cloud megastructures discussed in the City layer chapter, feudal megastructures such as these always come in pairs. Bloomfield Hills comes with South Detroit; Silicon Valley comes with its San Joaquin Valley. Today new enclave developments, and soon charter cities looping around them, are marketed as branded service platforms. In time, they will require more than this. In order to fully urbanize secession, they will have to take on the status of “ homeland ” and mobilize patriotism against the temptations of “ exit. ” Disney ’ s Celebration, Florida, is a landmark project here, from its branded mythology to its status as a self-governing city and county. Elsewhere developers recognize that a fetish for arbitrary distinctions of hierarchy isn ’ t a bug but a feature, and so at The Oaks, north of Los Angeles, residents who pass through one gate from the outside world still are excluded from the gated community inside the gated community, known as The Estates of The Oaks. In dense cities, enclaves are more vertical than horizontal and branded according to discreet (and discrete) address coordinates. In New York, One57 (Christian de Portzamparc, architect) and 432 Park Avenue (Rafael Vi ñ oly) towers near Central Park, and 56 Leonard in Tribeca (Herzog & de Meuron) are just a few options. The demos of the modern city presents certain difficulties not suffered by planners of the orbiting Elysium, however, as evidenced by concern in New York over so-called “ poor doors ” that would filter high-income from low-income residents of the same Westside tower, One Riverside Park. Readers of J. G. Ballard ’ s novel High Rise (1975) will also note the tendency for communities that live within the same building envelope, but strongly differentiated by status, to lurch inexorably toward tribal violence (and hence have a negative impact on unit resale value across the board). ²⁸

In light of this, it is not surprising to see enclave/camp conditions coexisting side by side and even sharing walls and borders, flipping interiority quickly or very slowly or not at all. My home university, University of California, San Diego in La Jolla, sits in the northwestern-most region of the San Diego/Tijuana metroplex, which is for many purposes a single city that happens to be bifurcated by an international border. A strategy of militarized luxury urbanism also has strong appeal to the affluent of Latin American cities in which violence is an everyday worry. For example, in San Salvador, “ wealthy Salvadorans can retreat to residential compounds that resemble a militarized version of a Palm Beach retirement community, complete with golf carts. Behind high walls and even higher voltage wires, one economist gushed to me: ‘ This place has everything — we never have to go outside! ’ For the rest, those who stay and those who get sent back, gangland drama is a fact of life. ” ²⁹ But it is not as if the poor do not have their own megastructure gated communities, with their own special dispositions given for exceptional enforcement of rules. Consider the Dr. Manuel de la Pila housing block in Ponce, Puerto Rico, one of hundreds built after World War II:

The Stack to Come

311

When it was first built Dr. Pila was an open community. But early one November morning in 1994, two years after a private firm had taken over its management, three helicopters carrying national guards and police descended upon the project, officially occupying it. Operation Centurion, popularly known as Mano Dura Contra el Crimen (Strong Arm Against Crime), had dictated that the largest, presumably most dangerous public housing projects should be gated in order to reduce crime. Over the course of four years, nearly a quarter of Puerto Rico ’ s 337 public housing developments were “ rescued ” or “ occupied, ” leading to arrests of residents, the establishment of police outposts, and the erection of fences to control movement. Dr. Pila became a gated public housing development.

The opportunity to live behind walls and under armed guard is available anywhere, it seems. It may even be true that in some circumstances, Users of a city-shaped camp (its autochthonous refugees), may have their clean water supply turned off if service fees are overdue. ³⁰ What remains persistent, however, is that the mutualizing interiority/exteriority of every enclave/camp condition is reversible, even if movement between one zone and another and the right of exit is not. ³¹ The geodesign problems of generalizable secession are now dominated by the artificially narrowed interest of capital transactions (and transactors) to protect themselves from direct contact with their own externalities, not by the general interest to seek out viable access to that capital. This is neither a necessary or conclusive situation, and increasing the delamination of sovereignty from territory and the emergence of computational platform infrastructures may be both the means and the ends to better alternatives. The impetus toward secession and an activist stance on platform sovereignty (which are not the same things) is obviously not limited to building borders and, as this book argues, The Stack ’ s delinking of infrastructural systems from place also relinks them again in new ways. Two operations interweave legacy megacities with the Cloud and allow Interface layers to spin out new media at multiple scales. A new medium brings new noise and new noise brings new music. Unlike the zero-sum subdivision of a land behind walls, requiring Users to vote with their personal presence, the Cloud Polis (in principle) comes to you wherever you are. Many of those who engineer these platforms are not only well aware of these architectures of proto-sovereignty; they evangelize them, sometimes ingeniously, sometimes anxiously, and sometimes idiotically.

Demonstrating all three of these in various measure, venture capitalist Balaji Srinivasan summarized the rationale of secession in his Internet-famous speech “ Silicon Valley ’ s Ultimate Exit. ” ³² His address is worth examining for what it clarifies about the current state of the popular discourse on Cloud Polis, how that discourse is so strikingly inadequate, and how it is received and misunderstood. Srinivasan ’ s key point was that as the loci of twentieth-century American power (in governance, media, finance, entertainment, education, military) are giving way to a new economic system based on Silicon Valley software, that the old order will inevitably defend its flagging legitimacy by blaming software macroeconomics for the world ’ s problems. He argues that the best remedy is to allow for “ free zones ” in which the “ world run by Silicon Valley ” could be tried, tested, and demonstrated without interference for all to observe. “ We need to run the experiment, to show what a society run by Silicon Valley looks like without affecting anyone who wants to live under the Paper Belt, ” he implores. The inevitable success of this new polity will make it obvious that it is the sovereign platform of choice for anyone clear-minded enough to care. For this, he argues, the essential criterion of liberty is the right to withdraw from unworkable regimes. Without naming the source, he draws on the thesis of Albert Hirschman ’ s Exit, Voice and Loyalty: Responses to Decline in Firms, Organizations, and States (1970), which addresses the remedies available to consumers and citizens in the face of deteriorating quality of goods and services (including, for example, a failed or repressive state). ³³ “ Voice ” is the right to petition or reform institutions to which one belongs, and “ exit ” is the right to get up and go, to emigrate, to change teams, to start anew. Addressing the diverse audience, Srinivasan says, “ I would bet that exit is a reason why half of this audience is alive. Many of us have our ancestors who came from China, Vietnam, Korea, Iran, places where there ’ s war or famine, economic basket cases. Exit is something that I believe we need to preserve, and exit is what this talk is about. ” He compares states and Cloud platforms directly and quickly asks us to evaluate one like the other. “ Is the United States the Microsoft of nations? ” he asks an audience for whom Microsoft represents everything old, inflexible, arbitrary, poorly designed, coercive, monolithic, and unfortunate about large platforms. Just as progress in the Cloud depends on Users’ ability to opt out of one platform and opt in to another that provides more robust, trustworthy, and well-designed services, he asks, doesn ’ t the same apply to governing platforms called states? Leaving iOS for Android does not involve navigating the armed guards at the Berlin Wall, or walking across the Sonoran desert (as of this writing), so why can ’ t movement between platforms of political sovereignty work the same way? We note that he does not specify the convergence of Cloud platforms absorbing traditional functions of the state with states rotating and evolving into Cloud platforms. That convergence may confirm or undermine the argument about the necessary superiority of one form over the other.

312

The Stack to Come

The more valuable point he makes (almost offhandedly) is one that does not require the geographic region of Silicon Valley to redraw lines on the map as such. “ That ’ s the thing about exit is: you can take as much or as little of it as you want. You don ’ t have to actually go and get your own island; you can do the equivalent of dual-booting or telecommuting. You can opt out, exit at whatever level [emphasis mine] you prefer. ” Especially if we take “ level ” to correspond strongly or weakly with the layers of The Stack, he is correct but perhaps not entirely aware of the cascading implications of this observation. As I discussed at length above, different Users may initiate columns up and down The Stack from many different positions and will be interpolated by the geoscopic interests of different Interfaces, will make use of what those interests are able to Address, will be situated within the interfacial envelopes of a given City, variously interiorizing and exteriorizing his location, will be enrolled by one or many competing Cloud platforms and supporting energy systems of varying interoperability. Different Users will have different columns available and unavailable to them, by definition, according to the determinant strength of one of more governing layers activated. The City layers of Pyongyang and Palo Alto will curtail or enable such very different possibilities for the other layers in The Stack that the City layer may have final leverage over the others. Or, in a perfectly walled garden, the platform coherence of the Cloud layer may be so closed and perfected that the Interface, Address, and User layers, are decided within a very small margin of variance. That is, for any real column up and down The Stack, there is a range between the strong incorporation of the sovereign operations of all layers together into complete vertical systems and the strong modularity of the layers into separate and counterprogrammable components. That range is not equally distributed to all Users everywhere, and the most important variations may be possible for some Users but not for others, and so exit is never available equally to and from all platform Users.

The Stack to Come

313

For our geodesign initiative, the important pressure point is that the sovereign operations of any one layer in The Stack are not necessarily reducible to or even interoperable with the others, and so redesigning (reprogramming, relegislating, rearticulating) the sovereign techniques of those layers can directly affect the others and shift the relative leverage between them. There is nothing inevitable about any one angle, and the recombinant logics of Stack platforms allow in principle for the future replacement and displacement of the machinery of one layer. Further, as the activation of a column occurs down and up all layers of The Stack from wherever a User may be physically located, the reformation of the sovereign qualities of any one layer may reverberate up to Users across the world from and to one another, but because they are located differently, the ultimate effects of that reverberation will not be uniform. As both activate that reprogrammed layer, one User may be opting out while another may be opting in, even as their gestures appear identical. What is up for one may be down for the other; what is exit for one may be entrance for another. As a User initiates multiple columns over time, even simultaneously, she may be inside the sovereign terms of one version of one layer (say, a Cloud platform or the apparent geolocation of her IP address) while simultaneously inside those same layers as the User of another column in another stack. That is, the political interpolation of the User never finally resolves into the biography of one single person in the same way that the identity of the citizen did and does. The management of multiple User identities and political positions is less a psychological disorder than the politics of everyday life. The design of the terms and norms of that irresolvable multiplicity is inseparable from the design of the sovereign characteristics of the Stack layers that mediate it, and in this way the simplistic state-versus-market metaphysics of mainstream Silicon Valley technolibertarianism rather undersells the disruptive potential of its own product.

314

The Stack to Come

The trial balloon of an information technology free zone echoes remarks made in 2013 by Google CEO Larry Page during his Google I/O keynote speech, as he wondered aloud about a “ Google Island ” where we “ set aside a part of the world. … I think as tech nologists we should have some safe places where we can try out new things and figure out the effect on society. What ’ s the effect on people, without having to deploy it to the whole world. ” ³⁴ Page compared his speculative temporary autonomous zone to Burning Man, a city that comes and goes on an annual cycle unlike an actual island, and so of only limited experimental significance and value. Page ’ s thought bubble presents more than few problems, even taken at face value (especially taken at face value). “ Without having to deploy it to the whole world ” implies that the island is a society-sized laboratory with all the standard measures in place to isolate and contain contaminants from entering or exiting. ³⁵ Accordingly the test-bed can prototype the ideal software society within its formal boundary only if that border is enforced by draconian walls and firewalls. This purified closure would either validate or violate the inspiration for the island as a project, depending on your motivation. If it is not breached and remains a vacuum, then the findings will have diminished significance outside the artificially hermetic environment. If data or some other aspect of the Island microsociety were to leak out, or some of the outside world to link in, on a regular basis, as happened with Biosphere 2 in the Arizona desert, then the experimental noninterference would be breached, to whatever significant or insignificant degree. If breached, then the island is really just one locus among many others in “ the whole world, ” one that may layer experiments one on another in dense concentration, but one that does what it does in direct (if filtered) relationship with other enclaves and camps. By the measure of Srinivasan ’ s admonitions, this is a mixed blessing, as the rarified testing grounds must still make sense of unprogrammable input from the normal outside, a concession that may disappoint some purists. On the other hand, his strenuous defense of exit must apply even to the utopias to which emigrants relocate, not just the ones from which they originated. People, data, carbon, air, and capital have to be able to leave Google Island for the principle and narrative to hold true. The only way to have it both ways is to solve for the closed totality of island and then “ deploy it to the whole world, ” where planetary boundaries drawn against outer space may provide the same closure as the island drawn against the ocean: to the make whole world into “ the island. ” The inclination toward that information universalism is more familiar to Google ’ s information cosmopolitanism project than it is to that of the strong secessionists.

There are other future scenarios to spin out from the leverage of this hypothetical fissure, and they all may really be a more polite way of describing the present from the comfort of a slightly off-stage futuristic perspective. If formal Westphalian political geography is to be further delinked from last-instance sovereignty, we could as easily imagine (if not as easily implement) a situation in which a state ’ s services are avail able to Users anywhere, according to their interests and choices, some determined by economic privileges, others by cultural and brand allegiances, others by whether they are simply a human or not. As Cloud platforms absorb state functions and states realize Cloud platform topologies and methods, The Stack does not necessarily privilege either in favor of the other. State and Cloud platforms may consolidate their competing and complementary services based on different strategies for protocol lock-in — some geographic, some constitutional, some by historical legacy, some contractual, some extractive, some by taxing premiums, some mere convention. For example, we may draw a scenario in which the European Union (or some quasi-state actor, perhaps a religious polity or a Cloud platform) offers far more secure and useful digital identity services than California, so those of us who are now California residents choose to be part of that platform instead of or as well as being Californians at other layers and other times, and effectively pay “ taxes ” to a state other than the one on our passports according to terms of service. In time, people from different locales might become more intertwined with the services and political conflicts of the EU than they are of their home nation, because it is the EU platform that is actually providing them a preferred secure economic and cultural identity, access to transactions, even education and other social services. Perhaps someone can choose to live under the EU platform ’ s data laws even if they mostly reside in Los Angeles, and perhaps, by the terms of a datatraffic encryption and securitization mutual nonaggression pact among platforms, no one can (or does) stop them. Perhaps they don ’ t even give up their US passports, because doing so wouldn ’ t really matter. In this scenario, some platform strategies would rely on interoperability among federated universal networks (maybe Google ’ s or maybe not), while others would entertain the advantages and suffer the disadvantages of closed loops (such as Iran ’ s continuously pending “ Halal Internet ” ). However, for our User from California, movement is still restricted. Perhaps the greater state services apparatus in China is part of a transnational bloc that for whatever reasons has been programmed to prevent interoperation with the EU platform. The walled garden topography becomes, once again, one of antagonistic fiefdoms and empires, but ones that are more agnostic as to the latitude and longitude coordinates of Users, citizens, clients, and members.

The Stack to Come

315