Читать книгу Duty Free Art - Hito Steyerl - Страница 9

A while ago I met an extremely interesting developer. He was working on smartphone camera technology. Photography is traditionally thought to represent what is out there by means of technology, ideally via an indexical link. But is this really true anymore? The developer explained to me that the technology for contemporary phone cameras is quite different from traditional cameras: the lenses are tiny and basically rubbish, which means that about half of the data being captured by the camera sensor is actually noise. The trick, then, is to write the algorithm to clean the noise, or rather to discern the picture from inside the noise.

But how can the camera know how to do this? Very simple: It scans all other pictures stored on the phone or on your social media networks and sifts through your contacts. It analyzes the pictures you already took, or those that are associated with you, and it tries to match faces and shapes to link them back to you. By comparing what you and your network already photographed, the algorithm guesses what you might have wanted to photograph now. It creates the present picture based on earlier pictures, on your/its memory. This new paradigm is being called computational photography.¹

The result might be a picture of something that never ever existed, but that the algorithm thinks you might like to see. This type of photography is speculative and relational. It is a gamble with probabilities that bets on inertia. It makes seeing unforeseen things more difficult. It will increase the amount of noise just as it will increase the amount of random interpretation.

And that’s not even to mention external interference into what your phone is recording. All sorts of systems are able to remotely turn your camera on or off: companies, governments, the military. It could be disabled in certain places—one could for instance block its recording function close to protests or conversely broadcast whatever it sees. Similarly, a device might be programmed to autopixelate, erase, or block secret, copyrighted, or sexual content. It might be fitted with a so-called dick algorithm to screen out NSFW (Not Suitable/Safe For Work) content, automodify pubic hair, stretch or omit bodies, exchange or collage context, or insert location-targeted advertising, pop-up windows, or live feeds. It might report you or someone from your network to police, PR agencies, or spammers. It might flag your debt, play your games, broadcast your heartbeat. Computational photography has expanded to cover all this.

It links control robotics, object recognition, and machine learning technologies. So if you take a picture on a smartphone, the results are not as premeditated as they are premediated. The picture might show something unexpected, because it might have cross-referenced many different databases: traffic control, medical databases, frenemy photo galleries on Facebook, credit card data, maps, and whatever else it wants.

Relational Photography

Computational photography is therefore inherently political —not in content but form. It is not only relational but also truly social, with countless systems and people potentially interfering with pictures before they even emerge as visible.² And of course this network is not neutral. It has rules and norms hardwired into its platforms, and they represent a mix of juridical, moral, aesthetic, technological, commercial, and bluntly hidden parameters and effects. You could end up airbrushed, wanted, redirected, taxed, deleted, remodeled, or replaced in your own picture. The camera turns into a social projector rather than a recorder. It shows a superposition of what it thinks you might want to look like plus what others think you should buy or be. But technology rarely does things on its own. Technology is programmed with conflicting goals and by many entities, and politics is a matter of defining how to separate its noise from its information.³

So what are the policies already in place that define the separation of noise from information, or that even define noise and information as such in the first place? Who or what decides what the camera will “see”? How is it being done? By whom or what? And why is this even important?

The Penis Problem

Let’s have a look at one example: drawing a line between face and butt, or between “acceptable” and “unacceptable” body parts. It is no coincidence that Facebook is called Facebook and not Buttbook, because you can’t have any butts on Facebook. But then how does it weed out the butts? A list leaked by an angry freelancer reveals the precise instructions given on how to build and maintain Facebook’s face, and it shows us what is well known, that nudity and sexual content are strictly off limits, except art nudity and male nipples, but also how its policies on violence are much more lax, with even decapitations and large amounts of blood acceptable.⁴ “Crushed heads, limbs etc are OK as long as no insides are showing,” reads one guideline. “Deep flesh wounds are ok to show; excessive blood is ok to show.” Those rules are still policed by humans, or more precisely by a global subcontracted workforce from Turkey, the Philippines, Morocco, Mexico, and India, working from home, earning around $4 per hour. These workers are hired to distinguish between acceptable body parts (faces) and unacceptable ones (butts). In principle, there is nothing wrong with having rules for publicly available imagery. Some sort of filtering process has to be implemented on online platforms: no one wants to be spammed with revenge porn or atrocities, regardless of there being markets for such imagery. The question concerns where and how to draw the line, as well as who draws it, and on whose behalf. Who decides on signal vs. noise?

Let’s go back to the elimination of sexual content. Is there an algorithm for this, like for face recognition? This question first arose publicly in the so-called Chatroulette conundrum. Chatroulette was a Russian online video service that allowed people to meet on the web. It quickly became famous for its “next” button, for which the term “unlike button” would be much too polite. The site’s audience at first exploded to 1.6 million users per month by 2010. But then a so-called “penis problem” emerged, referring to the many people who used the service to meet other people naked.⁵ The winner of a web contest called to “solve” the issue ingeniously suggested running a quick facial recognition or eye tracking scan on the video feeds—if no face was discernable, it would deduce that it must be a dick.⁶

This exact workflow was also used by the British Secret Service when it secretly bulk extracted user webcam stills in its spy program Optic Nerve. Video feeds of 1.8 million Yahoo users were intercepted in order to develop face and iris recognition technologies. But—maybe unsurprisingly—it turned out that around 7 percent of the content did not show faces at all. So—as suggested for Chatroulette—they ran face recognition scans on everything and tried to exclude the dicks for not being faces. It didn’t work so well. In a leaked document the GCHQ admits defeat: “There is no perfect ability to censor material which may be offensive.”⁷

Subsequent solutions became a bit more sophisticated. Probabilistic porn detection calculates the amount of skintoned pixels in certain regions of the picture, producing complicated taxonomic formulas, such as this one:

a. If the percentage of skin pixels relative to the image size is less than 15 percent, the image is not nude. Otherwise, go to the next step.

b. If the number of skin pixels in the largest skin region is less than 35% of the total skin count, the number of skin pixels in the second largest region is less than 30% of the total skin count and the number of skin pixels in the third largest region is less than 30% of the total skin count, the image is not nude.

c. If the number of skin pixels in the largest skin region is less than 45% of the total skin count, the image is not nude.

d. If the total skin count is less than 30% of the total number of pixels in the image and the number of skin pixels within the bounding polygon is less than 55 percent of the size of the polygon, the image is not nude.

e. If the number of skin regions is more than 60 and the average intensity within the polygon is less than 0.25, the image is not nude.

f. Otherwise, the image is nude.⁸

But this method got ridiculed pretty quickly because it produced so many false positives, including, as in some examples, wrapped meatballs, tanks, or machine guns. More recent porn-detection applications use self-learning technology based on neural networks, computational verb theory, and cognitive computation. They do not try to statistically guess at the image, but rather try to understand it by identifying objects through their relations.⁹

According to developer Tao Yang’s description, there is a whole new field of cognitive vision studies based on quantifying cognition as such, on making it measurable and computable.¹⁰ Even though there are still considerable technological difficulties, this effort represents a whole new level of formalization; a new order of images, a grammar of images, an algorithmic system of sexuality, surveillance, productivity, reputation, and computation that links with the grammatization of social relations by corporations and governments.

So how does this work? Yang’s porn-detection system must learn how to recognize objectionable parts by seeing a sizable mass of them in order to infer their relations. So basically you start by installing a lot of photos of the body parts you want eliminated on your computer. The database consists of folders full of body parts ready to enter formal relations. Not only pussy, nipple, asshole, and blowjob, but asshole, asshole/only and asshole/mixed_with_pussy. Based on this library, a whole range of detectors get ready to go to work: the breast detector, pussy detector, pubic hair detector, cunnilingus detector, blowjob detector, asshole detector, hand-touch-pussy detector. They identify fascinating sex-positions such as the Yawning and Octopus techniques, The Stopperage, Chambers Fuck, Fraser MacKenzie, Persuading of the Debtor, Playing of Cello, and Watching the Game (I am honestly terrified of even imagining Fraser MacKenzie).

This grammar as well as the library of partial objects are reminiscent of Roland Barthes’s notion of a “porn grammar,” where he describes the Marquis de Sade’s writings as a system of positions and body parts ready to permutate into every possible combination.¹¹ Yet this marginalized and openly persecuted system could be seen as a reflex of a more general grammar of knowledge deployed during the so-called Enlightenment. Michel Foucault as well as Theodor W. Adorno and Max Horkheimer compared de Sade’s sexual systems to mainstream systems of classification. Both were articulated by counting and sorting, by creating exhaustive, pedantic, and tedious taxonomies. And Mr. Yang’s enthusiasm for formalizing body parts and their relations to one another similarly reflects the huge endeavor of rendering cognition, imaging, and behavior as such increasingly quantifiable and commensurable to a system of exchange value based in data.

Undesirable body parts thus become elements of a new machine-readable image-based grammar that might usually operate in parallel to reputational and control networks, but that can also be linked to it at any time. Its structure might be a reflex of contemporary modes of harvesting, aggregating, and financializing data-based “knowledge” churned out by a cacophony of partly social algorithms embedded into technology.

Noise and Information

But let’s come back to the question we began with: What are the social and political algorithms that clear noise from information? The emphasis, again, is on politics not algorithms. Jacques Rancière has beautifully shown that this division corresponds to a much older social formula: to distinguish between noise and speech in order to divide a crowd between citizens and rabble.¹² If one wants not to take someone else seriously, or to limit their rights and status, one pretends that their speech is just noise, garbled groaning, or crying, and that they themselves must be devoid of reason and therefore exempt from being subjects, let alone holders of rights. In other words, this politics rests on an act of conscious decoding —separating “noise” from “information,” “speech” from “groan,” or “face” from “butt,” and from there neatly stacking the results into vertical class hierarchies.¹³ The algorithms now being fed into smartphone camera technology to define the image prior to its emergence are similar to this.

In light of Rancière’s proposition, we might still be dealing with a more traditional idea of politics as representation.¹⁴ If everyone is aurally (or visually) represented, and no one is discounted as noise, then equality might draw nearer. But the networks have changed so drastically that nearly every parameter of representative politics has shifted. By now, more people than ever are able to upload an almost unlimited number of self-representations. And the level of political participation by way of parliamentary democracy seems to have dwindled in the meantime. While pictures float in numbers, elites are shrinking and centralizing power.

On top of this, your face is getting disconnected—not only from your butt, but also from your voice and body. Your face is now an element—a face/mixed_with_phone, ready to be combined with any other item in the library. Captions are added, or textures, if needs be. Face prints are taken. An image becomes less a representation than a proxy, a mercenary of appearance, a floating texture-surface-commodity. Persons are montaged, dubbed, assembled, incorporated. Humans and things intermingle in ever-newer constellations to become bots or cyborgs.¹⁵ As humans feed affect, thought, and sociality into algorithms, algorithms feed back into what used to be called subjectivity. This shift is what has given way to a post-representational politics adrift within information space.¹⁶

Proxy Armies

Let’s look at one example of post-representational politics: political bot armies on Twitter. Twitter bots are bits of script that impersonate human activity on social media sites. In large, synchronized numbers they have become formidable political armies.¹⁷ A Twitter chat bot is an algorithm wearing a person’s face, a formula incorporated as animated spam. It is a scripted operation impersonating a human operation.

Bot armies distort discussions on Twitter hashtags by spamming them with advertisements, tourist pictures, or whatever. They basically add noise. Bot armies have been active in Mexico, Syria, Russia, and Turkey, where most political parties have been said to operate such bot armies. In Turkey, the ruling AKP alone was suspected of controlling 18,000 fake Twitter accounts using photos of Robbie Williams, Megan Fox, and other celebs: “In order to appear authentic, the accounts don’t just tweet out AKP hashtags; they also quote philosophers such as Thomas Hobbes and movies like PS: I Love You.”¹⁸

So who do bot armies represent, if anyone, and how do they do it? Let’s have a look at the AKP bots. Robbie Williams, Meg Fox, and Hakan43020638 are all advertising “Flappy Tayyip,” a cell-phone game starring the then AKP prime minister (now president) Tayyip Recep Erdoğan. The objective is to hijack or spam the hashtag #twitterturkey to protest PM Erdoğan’s banning of Twitter. Simultaneously, Erdoğan’s own Twitter bots set out to detourne the hashtag.

Let’s look at Hakan43020638 more closely: a bot consisting of a copy-pasted face plus product placement. It takes only a matter of minutes to connect his face to a body by way of a Google image search. On his business Twitter account it turns out he sells his underwear: he works online as an affective web service provider.¹⁹ Let’s call this version Murat, to throw yet another alias into the fray. But who is the bot wearing Murat’s face and who is a bot army representing? Why would Hakan43020638 be quoting Thomas Hobbes of all philosophers? And which book? Let’s guess he’s quoting from Hobbes’s most important work Leviathan. Leviathan is the name of a social contract enforced by an absolute sovereign in order to fend off the dangers presented by a “state of nature” in which humans prey upon one another. With Leviathan there are no more militias and there is no more molecular warfare of everyone against everyone.

But now we seem to be in a situation where state systems grounded in such social contracts seem to fall apart in many places and nothing is left but a set of policed relational metadata, emoji, and hijacked hashtags. A bot army is a contemporary vox populi, the voice of the people according to social networks. It can be a Facebook militia, your low-cost personalized mob, your digital mercenaries, or some sort of proxy porn. Imagine your photo being used for one of these bots. It is the moment when your picture becomes quite autonomous, active, even militant. Bot armies are celebrity militias, wildly jump-cutting between glamour, sectarianism, porn, corruption, and conservative religious ideologies. Post-representative politics are a war of bot armies against one another, of Hakan against Murat, of face against butt.

This may be why the AKP pornstar bots desperately quote Hobbes: they are already sick of the war of Robbie Williams (IDF) against Robbie Williams (Electronic Syrian Army) against Robbie Williams (PRI/AAP), they are sick of retweeting spam for autocrats—and are hoping for just any entity organizing day care, gun control, and affordable dentistry, whether it’s called Leviathan or Moby Dick or even Flappy Tayyip. They seem to say: we’d go for just about any social contract you’ve got!²⁰

Now let us go even one step further. Because a model for this might already be on the horizon. And unsurprisingly, it also involves algorithms.

Blockchain

Blockchain governance seems to fulfill the hopes for a new social contract.²¹ “Decentralized Autonomous Organizations” would record and store transactions in blockchains akin to the one used to run and validate bitcoin. But those public digital ledgers could equally encode votes or laws. Take for instance bitcongress, which is in the process of developing a decentralized voting and legislation system (www.bitcongress.org). While this could be a model to restore accountability and circumvent power monopolies, it means above all that social rules hardwired with technology emerge as Leviathan 2.0:

When disassociated from the programmers who design them, trustless blockchains floating above human affairs contain the specter of rule by algorithms … This is essentially the vision of the internet techno-leviathan, a deified crypto-sovereign whose rules we can contract to.²²

Even though this is a decentralized process with no single entity at the controls, it doesn’t necessarily mean no one controls it. Just like smartphone photography, it needs to be told how to work: by a multitude of conflicting interests. More importantly, this would replace bots as proxy “people” with bots as governance. But then again, which bots are we talking about? Who programs them? Are they cyborgs? Do they have faces or butts? And who is drawing the line? Are they cheerleaders of social and informational entropy? Killing machines? Or a new crowd, of which we are already a part?²³

Let’s come back to the beginning again: How to separate signal from noise? And how does the old political technology of using this distinction in order to rule change with algorithmic technology? In all examples, the definition of noise rested increasingly on scripted operations, on automating representation and/or decision-making. On the other hand, this process potentially introduces so much feedback that representation becomes a rather unpredictable operation that looks more like the weather than a Xerox machine. Likeliness becomes subject to likelihood—reality is just another factor in an extended calculation of probability. In this situation, proxies become crucial semi-autonomous actors.

Proxy Politics

To better understand proxy politics, we could start by drawing up a checklist:

Does your camera decide what appears in your photographs?

Does it go off when you smile?

And will it fire in a next step if you don’t?

Do underpaid outsourced IT workers in BRIC countries manage your pictures of breastfeeds and decapitations on your social media feeds?

Is Elizabeth Taylor tweeting your work?

Have some of your other fans’ bots decided to classify your work as urinary mature porn?

Are some of these bots busily enumerating geographic locations alongside bodily orifices?

Is your total result something like this?

Welcome to the age of proxy politics!

A proxy is “an agent or substitute authorized to act for another person or a document which authorizes the agent so to act” (Wikipedia). But a proxy could now also be a device with a bad hair day. A scrap of script caught up in a dress-code double bind. Or a “Persuading the Debtor” detector throwing a tantrum over genital pixel probability. Or a delegation of chat bots casually pasting pro-Putin hair lotion ads to your Instagram. It could also be something much more serious, wrecking your life in a similar way—sry life!

Proxies are devices or scripts tasked with getting rid of noise as well as bot armies hell-bent on producing it. They are masks, persons, avatars, routers, nodes, templates, or generic placeholders. They share an element of unpredictability—which is all the more paradoxical considering that they arise as result of maxed-out probabilities. But proxies are not only bots and avatars, nor is proxy politics restricted to datascapes. Proxy warfare is quite a standard model of warfare—one of the most important examples being the Spanish Civil War. Proxies add echo, subterfuge, distortion, and confusion to geopolitics. Armies posing as militias (or the other way around) reconfigure or explode territories and redistribute sovereignties. Companies pose as guerillas and legionnaires as suburban Tupperware clubs. A proxy army is made of guns for hire, with more or less ideological decoration. The border between private security, PMCs, freelance insurgents, armed stand-ins, state hackers, and people who just got in the way has become blurry. Remember that corporate armies were crucial in establishing colonial empires (the East India Company among others) and that the word “company” itself is derived from the name for a military unit. Proxy warfare is a prime example of a post-Leviathan reality.

Now that this whole range of activities has gone online, it turns out that proxy warfare is partly the continuation of PR by different means.²⁴ Besides marketing tools repurposed for counterinsurgency ops there is a whole range of government hacking (and counter-hacking) campaigns that require slightly more advanced skills. But not always. As the leftist Turkish hacker group Redhack reported, the password of the Ankara police servers was 12345.²⁵

To state that online proxy politics is reorganizing geopolitics would be similar to stating that burgers tend to reorganize cows. Indeed, just as meatloaf arranges parts of cows with plastic, fossil remnants, and elements formerly known as paper, proxy politics positions companies, nation-states, hacker detachments, FIFA, and the Duchess of Cambridge as equally relevant entities. Those proxies tear up territories by creating netscapes that are partly unlinked from geography and national jurisdiction.

But proxy politics also works the other way. A simple default example of proxy politics is the use of proxy servers to try to bypass local web censorship or communications restrictions. Whenever people use VPNs and other internet proxies to escape online restrictions or conceal their IP address, proxy politics is given a different twist. In countries like Iran and China, VPNs are very much in use.²⁶ In practice though, in many countries, companies close to censor-happy governments also run the VPNs in an exemplary display of efficient inconsistency. In Turkey, people used even more rudimentary methods—changing their DNS settings to tunnel out of Turkish dataspace, virtually tweeting from Hong Kong and Venezuela during Erdoğan’s short-lived Twitter ban.

In proxy politics the question is literally how to act or represent by using stand-ins (or being used by them)—and also how to use intermediaries to detourne the signals or noise of others. And proxy politics itself can also be turned around and redeployed. Proxy politics stacks surfaces, nodes, terrains, and textures—or disconnects them from one another. It disconnects body parts and switches them on and off to create often astonishing and unforeseen combinations—even faces with butts, so to speak. They can undermine the seemingly mandatory decision between face or butt or even the idea that both have got to belong to the same body. In the space of proxy politics, bodies could be Leviathans, hashtags, juridical persons, nation-states, hair-transplant devices, or freelance SWAT teams. Body is added to bodies by proxy and by stand-in. But these combinations also subtract bodies (and their parts) and erase them from the realm of never-ending surface to face enduring invisibility. In the end, however, a face without a butt cannot sit. It has to take a stand. And a butt without a face needs a stand-in for most kinds of communication. Proxy politics happens between taking a stand and using a stand-in. It is in the territory of displacement, stacking, subterfuge, and montage that both the worst and the best things happen.