Crushing the Stack

If in The Code Economy Philip Auerswald managed to give us a succinct history of the algorithm, while leaving us with code that floats like a ghost in the ether lacking any anchor in our very much material, economic and political world. Benjamin Bratton tries to bring us back to earth. Bratton’s recent book, The Stack: On software and sovereignty provides us with a sort of schematic with which we can grasp the political economy of code and thus anchor it to the wider world.

The problem is that Bratton, unlike Auerswald, has given us this schematic in the almost impenetrable language of postmodern theory beyond the grasp of even educated readers. Surely this is important for as Ian Bogost pointed out in his review of The Stack: “The book risks becoming a tome to own and display, rather than a tool to use.” This is a shame because the public certainly is in need of maps through which they can understand and seek to control the computational infrastructure that is now embedded in every aspect of our lives, including, and perhaps especially, in our politics. And the failure to understand and democratically regulate such technology leaves society subject to the whims of the often egomaniacal and anti-democratic nerds who design and run such systems.

In that spirit, I’ll try my best below to simplify The Stack into a map we can actually understand and therefore might be inclined to use.

In The Stack Bratton observers that we have entered the era of what he calls “planetary scale computation.” Our whole global system of processing and exchanging information, from undersea fiber-optic cables, satellites, cell-phone towers, server farms, corporate and personal computers along with our ubiquitous smartphones he see sees as “an accidental megastructure” that we have cobbled together without really understanding what we are building. Bratton’s goal, in a sense, is to map this structure by treating it as a “stack”, dissecting it into what he hopes are clearly discernible “layers.” There are six of these: Earth, Cloud, City, Address, Interface and User.

It is the Earth layer that I find both the most important and the most often missed when it comes to discussions of the political economy of code. Far too often the Stack is represented as something that literally is virtual, disconnected from the biosphere in a way that the other complex artificial systems upon which we have come to depend, such as the food system or the energy system, could never be as a matter of simple common sense. And yet the Stack, just like everything else human beings do, is dependent upon and effects the earth. As Bratton puts it in his Lovecraftian prose:

The Stack terraforms the host planet by drinking and vomiting its elemental juices and spitting up mobile phones. After its short career as a little computing brick within a larger megamachine, its fate at the dying end of the electronics component life cycle is just as sad. What is called “electronic waste” inverts the process that pulls entropic reserves of metal and oil from the ground and given form, and instead partially disassembles them and reburies them, sometimes a continent away and sometimes right next door. (p.83)

The rare earth minerals upon which much of modern technology depends come at the cost of environmental degradation and even civil war, as seen in the Democratic Republic of Congo. Huge areas of the earth are now wastelands festooned with the obsolescent silicon of our discarded computers and cell phones picked over by the world’s poorest for whatever wealth might be salvaged.

The Stack consumes upwards of 10 percent of the world’s energy. It’s an amount that is growing despite the major tech players efforts to diminish its footprint by relocating servers in the arctic, and, perhaps soon, under the sea. Although gains in efficiency have, at least temporarily, slowed the rate of growth in energy use.

The threat to the earth from the Stack, as Bratton sees it, is that its ever growing energy and material requirements will end up destroying the carbon based life that created it. It’s an apocalyptic scenario that is less fanciful than it sounds for the Stack is something like the nervous system for the fossil fuel based civilization we have built. Absent our abandonment of that form of civilization we really will create a world that is only inhabitable by machines and machine-like life forms such as bacteria. Wall-e might have been a prophecy and not just a cartoon.

Yet Bratton also sees the Stack as our potential savior, or at least the only way possible without a massive die off of human beings, to get out of this jam. A company like Exxon Mobil with its dependence on satellites and super-computers is only possible with the leverage of the Stack, but then again so is the IPCC.

For the Stack allows us to see nature, to have the tools to monitor, respond to, and perhaps even interfere with the processes of nature many of which the Stack itself is throwing out of kilter. The Stack might even give us the possibility of finding an alternative source of power and construction for itself. One that is compatible with our own survival along with the rest of  life on earth.

After the Earth layer comes the Cloud layer. It is here that Battron expands upon the ideas of Carl Schmitt. A jurist under the Nazi regime, Schmitt’s ideas about the international order have become popular among many on the left at least since the invasion of Iraq by the US in 2003 not as a prescription, but as a disturbingly prescient description of American politics and foreign policy in the wake of 9-11.

In his work The Nomos of the Earth Schmitt critiqued the American dominated international order that had begun with the US entry into WWI and reigned supreme during the Cold War  as a type of order that had, by slipping free the of the anchor of national sovereignty bound to clearly defined territories, set the world on the course of continuous interventions by states into each other’s domestic politics leading to the condition of permanent instability and the threat of total war.

Bratton updates Schmitt’s ideas for our era in which the control of infrastructure has superseded the occupation of territory as the route to power. Control over the nodes of  global networks, where assets are no longer measured in square miles, but in underwater cables, wireless towers, and satellites demands a distributed form of power, and hence helps explain the rise of multinational corporations to their current state of importance.

In terms of the Stack, these are the corporations that make up Bratton’s Cloud Layer, which include not only platforms such as Google and FaceBook, but the ISPs controlling much of the infrastructure upon which these companies (despite their best efforts to build such infrastructure themselves), continue to depend.

Bratton appears to see current geopolitics as a contest between two very different ideas regarding the future of the Cloud. There is the globalist vision found in Silicon Valley companies that aims to abandon the territorial limits of the nation-state and the Chinese model, which seeks to align the Cloud to the interests of the state. The first skirmish of this war Bratton notes was what he calls the Sino-Google War of 2009 in which Google under pressure from the Chinese government to censor its search results eventually withdrew from the country.

Unfortunately for Silicon Valley, along with those hoping we were witnessing the last gasp of the nation-state, not only did Google lose this war, it has recently moved to codify the terms of its surrender, while at the same time we have witnessed both a global resurgence of nationalism and the continuing role of the “deep-state” forcing the Cloud to conform to its interests.

Bratton also sees in the platform capitalism enabled by the Cloud the shape of a possible socialist future- a fulfillment of the dreams of rational, society-wide economic planning that was anticipated with the USSR’s Gosplan, and Project Cybersyn in pre-Pinochet Chile. The Stack isn’t the only book covering this increasingly important and interesting beat.

After the Cloud layer comes the City layer. It is in cities where the density of human population allows the technologies of the Stack to be most apparent. Cities, after all, are thick agglomerations of people and goods in motion all of which are looking for the most efficient path from point A to point B. Cities are composed privatized space made of innumerable walls that dictate entry and exit. They are the perfect laboratory for the logic and tools of the Stack. As Bratton puts it:

We recognize the city he describes as filled with suspicious responsive environments, from ATM PINs, to key cards and parking permits, e-tickets to branded entertainment, personalized recommendations from others who have purchased similar items, mobile social network transparencies, GPS-enabled monitoring of parolees, and customer phone tracking for retail layout optimization.  (p. 157)

Following the City layer we find the Address. In the Stack (or at least in the version of it dreamed up by salesmen for the Internet of Things), everything must have a location in the network, a link to which it can be connected to other persons and things. Something that lacks an address in some sense doesn’t exist for the Stack. An unconnected object or person fails to be a repository for information on which the Stack itself feeds.

We’ve only just entered the era in which our everyday objects speak to one another and in the process can reveal information we might have otherwise hidden about ourselves. What Bratton finds astounding is that in the Address layer we can see that the purpose of our communications infrastructure has become not for humans to communicate with other humans via machines, but for machines to communicate with other machines.

The next layer is that of the Interface. It is the world of programs and apps through which for most of us is the closest we get to code. Bratton says it better:

What are Apps? On the one hand, Apps are software applications and so operate within something like an application layer of a specific device-to-Cloud economy. However, because most of the real information processing is going on in the Cloud, and not in the device in your hand, the App is really more an interface to the real applications hidden away in data centers. As an interface, the App connects the remote device to oceans of data and brings those data to bear on the User’s immediate interests; as a data-gathering tool, the App sends data back to the central horde in response to how the User makes use of it. The App is also an interface between the User and his environment and the things within it, by aiding in looking, writing, subtitling, capturing, sorting, hearing, and linking things and events. (p.142)

The problem with apps is that they offer up an extremely narrow window on the world. Bratton is concerned about the political and social effects of such reality compression, a much darker version of Eli Pariser’s “filter bubble”, where the world itself is refracted into a shape that conforms to the individual’s particular fetishes, shattering a once shared social world.

The rise of filter bubbles are the first sign of a reality crisis Bratton thinks will only get worse with the perfection of augmented reality-there are already AR tours of the Grand Canyon that seek to prove creationism is true.

The Stack’s final layer is that of the User. Bratton here seems mainly concerned with expanding the definition of who or what constitutes one. There’s been a lot of hand-wringing about the use of bots since the 2016 election. California has even passed legislation to limit their use. Admittedly, these short, relatively easy to make programs that allow automated posts or calls are a major problem. Hell, over 90% of the phone calls I receive are now unsolicited robocalls, and given that I know I am not alone in this, such spam might just kill the phone call as a means of human communication. Ironically, the very reason we have cellphones in the first place.

Yet bots have also become the source of what many of us would consider not merely permissible, but desirable speech. It might upset me that countries like Russia and Saudi Arabia are vociferous users of bots to foster their interests among English speaking publics, or scammers using bots to pick people’s pockets, but I actually like the increasing use of bots by NGOs whose missions I support.

Bratton thus isn’t crazy for suggesting we give the bots some space in the form of “rights”. Things might move even further in this direction as bots become increasingly more sophisticated and personalized. Few would go so far as Jamie Susskind in his recent book Future Politics in suggesting we might replace representative government by a system of liquid democracy mediated by bots; one in which bots make political decisions for individuals based on the citizen’s preferences. But, here again, the proposal isn’t as ridiculous or reactionary as it might sound.

Given some issue to decide upon my bot could scan the position on the same by organizations and individuals I trust in regards to that issue. “My” votes on environmental policy could reflect some weighted measure between the views of the World Wildlife Fund, Bill Mckibben and the like, meaning I’d be more likely to make an informed vote than if I had pulled the lever on my own. This is not to say that I agree with this form of politics, or even believe it to be workable. Rather, I merely think that Bratton might be on to something here. That a key question in the User layer will be the place of bots- for good and ill.

The Stack, as Bratton has described it, is not without its problems and thus he ends his book with proposals for how we might build a better Stack. We could turn the Stack into a tool for the observation and management of the global environment. We could give Users design control over the interfaces that now dictate their lives, including the choice to enter and exit when we choose, a right that should be extended to the movement between states as well. We could use the power of platforms to revive something like centrally planned economies and their dream of eliminating waste and scarcity. We could harness the capacity of the Interface layer to build a world of plural utopias, extend and articulate the rights and responsibilities of users in a world full of bots.

Is Bratton right? Is this the world we are in, or at least headed towards. For my money, I think he gets some things spectacularly right, such as his explanation of the view of climate change within the political right:

“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.” (p.306)

Yet, elsewhere I think his views are not only wrong, but sometimes contradictory. I think he largely misses how the Stack is in large part a product of American empire. He, therefore, misinterprets the 2009 spat between Google and China as a battle between two models of future politics, rather than seeing the current splintering of the internet for what it is: the emergence of peer competitors in the arena of information over which the US has for so long been a hegemon.

Bratton is also dismissive of privacy and enraptured by the Internet of Things in a way that can sometimes appear pollyannaish. After all, privacy isn’t just some antiquated right, but one of the few ways to keep hackable systems secure. That he views the IoT as something inevitable and almost metaphysical, rather than the mere marketing it so often is, leads me to believe he really hasn’t thought through what it means to surround ourselves with computers- that is to make everything in our environment hackable. Rather than being destined to plug everything into everything else, we may someday discover that this is not only unnecessary and dangerous, but denotes a serious misunderstanding of what computation is actually for.

Herein lies my main problem with the Stack: though radically different than Yuval Harari, Bratton too seems to have drank the Silicon Valley Kool Aid.  The Stack takes as its assumption that the apps flowing out of the likes of FaceBook and Google and the infrastructure behind them are not merely of world-historical, but of cosmic import. Matter is rearranging itself into a globe spanning intelligence with unlikely seeds like a Harvard nerd who wanted a website to rate hot-chicks. I just don’t buy it.

What I do buy is that the Stack as a concept, or something like it, will be a necessary tool for negotiating our era, where the borders between politics and technology have become completely blurred. One can imagine a much less loquacious and more reality-based version of Bratton’s book that used his layers to give us a better grasp of this situation. In the Earth layer we’d see the imperialism behind the rare-earth minerals underlying our technology, we’d see massive Chinese factories like those of FoxConn, the way in which earth destroying coal continues to be the primary energy source for the Stack.

In the Cloud layer we’d gain insight into server farms and monopolistic ISPs such as Comcast, and come to understand the fight over Net Neutrality. We’d be shown the contours of the global communications infrastructure and the way in which these are plugged into and policed by government actors such as the NSA.

In the City layer we’d interrogate idea of smart cities, along with the automation of inequality and digitization of citizenship along with exploring the role of computation in global finance. In the Address layer we’d uncover the scope of logistics and  find out how platforms such as Amazon work their magic, and ask whether it really was magic or just parasitism, and how we might use these insights for the public good, whether that meant nationalizing the platforms or breaking them into pieces.

In the User layer we’d take a hard look at the addictive psychology behind software, the owners and logic behind well-known companies such as FaceBook along with less well known such as MindGeek. Such an alternative version of the Stack, would not only better inform us as to what the Stack is, but suggest what we might actually do to build ourselves a better one.  

 

Advertisements

City As Superintelligence

Medieval Paris

A movement is afoot to cover some of the largest and most populated cities in the world with a sophisticated array of interconnected sensors, cameras, and recording devices, able to track and respond to every crime or traffic jam ,every crisis or pandemic, as if it were an artificial immune system spread out over hundreds of densely packed kilometers filled with millions of human beings. The movement goes by the name of smart-cities, or sometimes sentient cities, and the fate of the project is intimately tied to the fate of humanity in the 21st century and beyond because the question of how the city is organized will define the world we live in from here forwards -the beginning of era of urban mankind.

Here are just some of many possible examples of smart cities at work, there is the city of Sondgo in South Korea a kind of testing ground for companies such as Cisco which can experiment with integrated technologies, to quote a recent article on the subject, such as:

TelePresence system, an advanced videoconferencing technology that allows residents to access a wide range of services including remote health care, beauty consulting and remote learning, as well as touch screens that enable residents to control their unit’s energy use.

Another example would be IBM’s Smart City Initiative in Rio which has covered that city with a dense network of sensors and cameras that allow centralized monitoring and control of vital city functions, and was somewhat brazenly promoted by that city’s mayor during a TED Talk in 2012. New York has set up a similar system, but it is in the non-Western world where smart cities will live or die because it is there where almost all of the world’s increasingly rapid urbanization is taking place.

Thus India, which has yet to urbanize like its neighbor, and sometimes rival, China, has plans to build up to 100 smart cities with 4.5 billion of the funding towards such projects being provided by perhaps the most urbanized country on the planet- Japan.

China continues to urbanize at a historically unprecedented pace with 250 million of its people- the equivalent of the entire population of the United States a generation ago- to move to its cities in the next 12 years. (I didn’t forget a zero.) There you have a city that few of us have even heard of – Chongqing, – which The Guardian several years back characterized as “the fastest growing urban center on the planet”  with more people in it than the entire countries of Peru and Iraq. No doubt in response to urbanization pressure, and at least back in 2011, Cisco was helping that city with its so-called Peaceful Chongqing Project an attempt to blanket the city in 500,000 video surveillance cameras- a collaboration that was possibly derailed by allegations by Edward Snowden that the NSA had infiltrated or co-opted U.S. companies.

Yet there are other smart-city initiatives that go beyond monitoring technologies. Under this rubric should fall the renewed interest in arcologies- massive buildings that contain within them an entire city, and thus in principle allow a city to be managed in terms of its climate, flows, etc. in the same way the internal environment of a skyscraper can be managed. China had an arcology in the works in Dongtan, which appears to have been scrapped over corruption and cost overrun concerns. Dubai has its green arcology in Masdar City, but it’s in Russia in the grip of a 21st century version of czarism, of all places, where the mother of all arcologies is planned, architect Norman Foster’s Crystal Island which, if actually built, would be the largest structure on the planet.

On the surface, there is actually much to like about smart-cities and their related arcologies. Smart-cities hold out the promise of greater efficiency for an energy starved and warming world. They should allow city management to be more responsive to citizens. All things being equal, smart-cities should be better than “dumb” ones at responding to everything from common fires and traffic accidents to major man- made and natural disasters. If Wendy Orent is correct as she wrote in a recent issue of AEON that we have less to fear from pandemics emerging from the wilderness such as Ebola than those that evolve in areas of extreme human density, smart-city applications should make the response to pandemics both quicker and more effective.

Especially in terms of arcologies, smart-cities represent something relatively new. We’ve had our two major models of the city since the early to mid-20th century, whether the skyscraper cities pioneered by New York and Chicago or the cul-de-sac suburban sprawl of cities dominated by the automobile like Phoenix. Cities going up now in the developing world certainly look more modern than American cities many of whose infrastructure is in a state of decay, but the model is the same, with the marked exception of all those super-trains.

All that said there are problems with smart-cities and the thinker who has written most extensively on the subject Anthony M. Townsend lays them out excellently in his book Smart Cities: Big-Data, Civic Hackers and the Quest for a New Utopia. Townsend sees three potential problems with smart-cities- they might prove, in his terms, “buggy. brittle, and bugged”.  

Like all software, the kinds that will be used to run smart-cities might exhibit unwanted surprises. We’ve seen this in some of the most sophisticated software we have running, financial trading algorithms whose “flash crashes” have felled billion dollar companies.

The loss of money, even a great deal of money, is something any reasonably healthy society should be able to absorb, but what if buggy software made essential services go off line over an extended period? Cascades from services now coupled by smart-city software could take out electricity and communication in a heat wave or re-run of last winter’s “polar vortex” and lead to loss of life. Perhaps having functions separated in silos and with a good deal of redundancy, even at the cost of inefficiency, is  “smarter” than having them tightly coupled and under one system. That is, after all, how the human brain works.  

Smart-cities might also be brittle. We might not be able to see that we had built a precarious architecture that could collapse in the face of some stressor or effort to intentionally harm- ahem– Windows. Computers crash and sometimes do so for reasons we are completely at a loss to identify. Or, imagine someone blackmailing a city by threatening to shut it down after having hacked its management system. Old school dumb-cities don’t really crash, even if they can sicken and die, and its hard to say they can be hacked.

Would we be in danger of decreasing a city’s resilience by compressing its complexity into an algorithm? If something like Stephen Wolfram’s principles of computational equivalence  and computational irreducibility is correct then the city is already a kind of computation and no model we can create of it will ever be more effective than this natural computation itself.

Or, to make my meaning clearer, imagine that you had a person that had suffered some horrible accident where to save them you had to replace all of his body’s natural information processing with a computer program. Such a program who have to regulate everything from breathing to metabolism to muscle movement ,along with the immune system, and exchange between neurons, not to mention a dozen other things. My guess is that you’d have to go out many generations of such programs before they are anywhere near workable. That the first generations would miss important elements, be based on wrong assumptions on how things worked, and would be loaded with perhaps catastrophic design errors that you couldn’t identify until the program was fully run in multiple iterations.

We are blissfully unaware that we are the product of billions of years of “engineering” where “design” failures were weeded out by evolution. Cities have only a few thousand years of a similar type of evolution behind them, but trying to control a great number of their functions via algorithms run by “command centers” might pose similar risks to my body example.  Reducing city functions to something we can compute in silicon might oversimplify the city in such a way as to reduce its resilience to stressors cities have naturally evolved to absorb. That is, there is, in all use of “Big-Data”, a temptation to interpret reality only in light of the model that scaffolds this data or reframe problems in ways that can mathematically be modeled. We set ourselves up for crises when we confuse the map with the territory or as Jaron Lanier said:

 What makes something real is that it is Impossible to represent it to completion.

Lastly, and as my initial examples of smart-cities should have indicated, smart-cities are by design bugged. They are bugged so as to surveil their citizens in an effort to prevent crime or terrorism or even just respond to accidents or disasters. Yet the promise of safety comes at the cost of one of the virtues of city living – the freedom granted from anonymity. But even if we care nothing for such things I’ve got news- trading privacy for security doesn’t even work.

Chongqing may spend tens of millions of dollars installing CCTV cameras, but would be hooligans or criminals or just people who don’t like being watched such as those in London, have a twenty dollar answer to all these gizmos- it’s called a hoodie. Likewise, a simple pattern of dollar store facepaint, strategically applied, can short-circuit the most sophisticated facial recognition software. I will never cease to be amazed at human ingenuity.    

We need to acknowledge that it is largely companies or individuals with extremely deep pockets and even deeper political connections that are promoting this model of the city. Townsend estimates it is potentially a 100 billion dollar business. We need to exercise our historical memory and recall how it was automobile companies that lobbied for and ended up creating our world of sprawl. Before investing millions or even billions cities need to have an idea of what kind of future they want to have and not be swayed by the latest technological trends.

This is especially the case when it comes to cities in the developing world where the conditions often resemble something more out of Dicken’s 19th century than even the 20th. When I enthusiastically asked a Chinese student about the arcology at Dongtan he responded with something like “Fools! Who would want to live in such a thing! It’s a waste of money. We need clean air and water, not such craziness!” And he’s no doubt largely right. And perhaps we might be happy that the project ultimately unraveled and say with Thoreau:

As for your high towers and monuments, there was a crazy fellow once in this town who undertook to dig through to China, and he got so far that, as he said, he heard the Chinese pots and kettles rattle; but I think that I shall not go out of my way to admire the hole which he made. Many are concerned about the monuments of the West and the East — to know who built them. For my part, I should like to know who in those days did not build them — who were above such trifling.

The age of connectivity, if it’s done thoughtfully, could bring us cities that are cleaner, greener, more able to deal with the shocks of disaster or respond to the spread of disease. Truly smart-cities should support a more active citizenry, a less tone-deaf bureaucracy, a more socially and culturally rich, entertaining and more civil life- the very reasons human beings have chosen to live in cities in the first place .

If the age of connectivity is done wrong cities will have poured scarce resources down a hole of corruption as deep as the one dug by Thoreau’s townsman, will have turned vibrant cultural and historical environments into corporate “flat-pack” versions of tomorrowland, and most frighteningly of all turned the potential democratic agora of the city into a massive panopticon of Orwellian monitoring and control. Cities, in the Wolfram not Bostrom sense, are already a sort of super-intelligence or better, hive mind of its interconnected yet free individuals more vibrant and important than any human built structure imaginable.  Will we let them stay that way?