The Twittering Machine

by Richard Seymour

  At another level, that of meaning, one could say that addiction is a thwarted form of love. It is a passionate attachment to something that, slowly, occupies a larger and larger part of one’s mind. It exercises a veto over other loves, aspirations and dreams. It occupies attention, when attention is subject to economic scarcity. It usurps our ingenuity, when the goal in life becomes maintaining access to the object, staying close to it. For the Twittering Machine, this is good: it keeps us writing. In an attention economy, addiction is not so much a scourge as a mode of production.

  Anything that so captures our attention must be the object of intense fantasies. In the history of junkie literature, for example, drugs are magical, fairy-tale objects, summoning abundance from nothing, defying the laws of physics. Or so it seems at first. Thomas De Quincey’s famous Confessions of an English Opium-Eater, for example, stands out for its utopian air.22 With the first hit, he had discovered ‘the secret of happiness’, ‘a resurrection, from the lowest depth’, an ‘abyss of divine enjoyment’, ‘an apocalypse of the world within’, ‘portable ecstasies . . . corked up in a pint bottle’. He had discovered the magic beans, the goose that laid the golden egg, the flax spun into gold: emotional plenitude. A bounty comparable only to the oceanic bliss in pursuit of which mystics of all faiths have undergone extraordinary physical and mental rigours.

  As the high diminishes, however, the fantasies become darker. When the Catholic mystic and poet Francis Thompson sang of ‘The Poppy’, the source of his ‘withered dreams’, it was as though he had become a hapless husk for the magical substance:23

  The sleep-flower sways in the wheat its head,

  Heavy with dreams, as that with bread.

  [ . . . ]

  I hang ’mid men my needless head,

  And my fruit is dreams, as theirs is bread.

  [ . . . ]

  Love! I fall into the claws of Time:

  But lasts within a leavèd rhyme

  All that the world of me esteems –

  My wither’d dreams, my wither’d dreams.

  His head, the actual source of his dreams, was, just like the flower, a ‘needless’ drooping cocoon for the opium. He credited his remaining creative power to the drug. Addicts tend to fetishize the object of their addiction. They attribute to it their own agency and imagine that it holds great powers that it really doesn’t. At the same time, they suffer a profound subjective impoverishment: the addict is as poor as the object is rich.

  The Twittering Machine appears to have a similar magical quality. Technology has never been just technology. It is always a world of intense emotional attachments.24 The Twittering Machine promises to give us access to everything, limitlessly, allowing us to transcend the limitations of mere flesh. This is how the telecommunications firm, MCI, sold the internet two decades ago.25 People could communicate ‘mind to mind’. No race, no gender, no age, no infirmity. ‘There are only minds,’ the advertising breathlessly suggested. ‘Utopia? No . . . The internet. Where minds, doors and lives open up.’ This was digital Clintonism, a kind of thin liberal utopianism. Standing in a weak shadow of the opiate sublime, it promised an abundance of being, ageless immortality, protean plasticity beyond the bedrock of the body. The name of this abundance was connectivity, a truly magical substance.

  The social platforms give concentrated expression to this idea, turning it into a business model and raison d’être. Facebook’s first video advertisement reminded us that the universe ‘is vast and dark and makes us wonder if we are alone’.26 We build connections, it said, to ‘remind ourselves that we are not’. Connection was the basis for ‘a great nation’, ‘something people build so they can have a place where they belong’. By implication, the platforms would be nation-builders, through the power of connectivity. At the outset of the ‘Twitter revolutions’, this same magical substance was supposed to outflank the old regimes and engender democratic upheaval.

  But as the cyberpunk writer Bruce Sterling points out, connectivity is not necessarily a symbol of affluence and plenty.27 It is, in a sense, the poor who most prize connectivity. Not in the sense of the old classist stereotype that ‘the poor love their cellphones’: no powerful group would turn down the opportunities that smartphones and social media offer. The powerful simply engage differently with the machine. But any culture that values connectivity so highly must be as impoverished in its social life as a culture obsessed with happiness is bitterly depressed. What Bruce Alexander calls the state of permanent ‘psychosocial dislocation’ in late capitalism, with life overrun by the law of markets and competition, is the context for soaring addiction rates.28 It is as if the addictive relationship stands in for the social relationships that have been upended by the turbulence of capitalism.

  The nature of this social poverty can be recognized in a situation typical of a social industry addict. We often use our smartphones to take us away from a social situation, without actually leaving that situation. It is as though we are both lonely and threatened by intimacy. We develop ways of simulating conversational awareness while attending to our phones, a technique known as ‘phubbing’. We experience this weirdly detached ‘uniform distancelessness’, as Christopher Bollas calls it.29 We become nodes in the network, equivalent to ‘smart’ devices, mere points of relay for fragments of information; as much extensions of the tablet or smartphone as they are of us. We prefer the machine when human relationships have become disappointing.30

  IV.

  Over the last twenty years, a number of apparently discrete social changes have taken place in the richer countries, above all in Europe and North America. First, a sharp decline in all forms of violence, including sexual violence, has been found in most of these societies.31 Almost simultaneously, these societies have seen a decline, almost a crash, in rates of alcohol and nicotine consumption, which historically have tended to be consumed socially.32 Finally, young people are having far less sex, something that has been subject to a great deal of mocking prurience. It seems odd, after all, that young people are more sexually liberal than their forebears, while at the same time more likely to avoid sex itself.

  One thing that these tendencies have in common, though, is that they all show a decline in sociality. Other data confirms this. Analysis of American post-Millennials by psychologist Jean Twenge finds that they are far less likely than their predecessors to go out, go on dates or have sex.33 This is one of the reasons for the plummeting teen pregnancy rate. The trend, she says, is strongly correlated with the ubiquity of smartphones prevalent since 2011–12. Cigarettes and alcohol, like the proverbial coffee, have been used as props for social interaction. It is no accident, says the psychoanalyst Darian Leader, that as soon as we abandoned cigarettes, the mobile phone appeared in our hands – as though we can’t face one another without some sort of medium.34 But the smartphone is not a prop for social interaction. It is an escape route, a way to connect with someone who isn’t there; or is only there as a written trace, a ghost in the machine.

  The fantasy of plenitude, the superabundance of online shit, may allow us to experience our social poverty as affluence, as in the fantasy that the internet and the social industry are ‘post-scarcity’.35 Like many fantasies, this has some basis in reality when not just ‘free stuff’, but even affection and romantic excitement can be accumulated in an objectified form as ‘likes’ and ‘matches’. But as with so many fairy tales, it is the fantasy, the wish fulfilment, of the poor. Social media are not the cause of this social impoverishment, any more than drugs are. They are just a more sophisticated remedy than booze and fags.

  But the Twittering Machine is a techno-political regime which in its own way absorbs any nascent desire to challenge these painful conditions. The literary critic Raymond Williams once wrote of certain technologies which promoted ‘mobile privatization’.36 While electrification and railway-building were public affairs, cars and personal stereos were simultaneously mobile and bound to the self-sufficient individual or family home. Silicon
Valley has taken this logic much further, extending privatization into the most public of spaces, soliciting our participation on a solitary basis. At the same time, it has taken the place of previous forms of self-medication. Just as the pharmaceutical giants are losing ground with their ‘one weird trick’ pill-shaped remedies for social distress, tech says ‘there’s an app for that’. The psychoanalyst Colette Soler has written of ‘the unprecedented development of techniques of listening targeting solitary voices in distress rather than really finding help for them’.37 The Twittering Machine is a technique for listening to solitary voices on a giant scale – shout at a politician, denounce a celebrity, rant at a CEO – the possibilities are endless.

  Rather than reducing addiction to a chemical experience, then, we have to look at what problems addiction might be solving. In an arresting image, Marcus Gilroy-Ware compares social media to a fridge which has something new in it every time we look.38 It might only be a half-empty tube of tomato paste, an out-of-date yoghurt or last night’s scraps. And we might not really be hungry. But at least we understand hunger, in a way that we don’t necessarily understand the obscure feelings of dissatisfaction that sent us to the fridge in the first place. We have the option of treating this opaque wanting as if it were hunger, to be satisfied with a feed. But what is it that we’re eating?

  V.

  Why is addiction such a useful economic model for social industry giants – indeed, for so many businesses? How does it work with the informational politics of the machine? And what does it say about the relationship of this machine to its users? Part of the answer lies in the mid-twentieth-century behaviourist revolt against free will. A revolt with a strange utopian dimension.

  This is paradoxical, since the idea of free will is central to the liberal market-based system we inhabit. We’re supposed to be able to decide what we prefer within the rules – rules which English philosopher Thomas Hobbes, in a pregnant metaphor, compared to the ‘laws of gaming’.39 We may not decide the rules, but we decide where to place our bets and when to ante up. And on the face of it, that surely is what we do on the social industry. No one forces us to be there, and no one tells us what to post, ‘like’ or click. And yet our interactions with the machine are conditioned. Critics of social media like Jaron Lanier argue that the user experience is designed much like the famous ‘Skinner Box’ or ‘operant conditioning chamber’ invented by the pioneering behaviourist B. F. Skinner. In this chamber, the behaviour of laboratory rats was conditioned by stimuli – lights, noises and food. Each of these stimuli constituted a ‘reinforcement’, either positive or negative, which would reward some forms of behaviour and discourage others. In the Skinner Box, test subjects are taught how to behave through conditioning. And if this model has found its way into the mobile apps, gaming and social industries, it might reflect the way that behaviourist ideas have achieved a surprising renaissance among businessmen and policymakers in recent decades.

  B. F. Skinner was not just a behavioural scientist, alongside peers, such as Pavlov, Thorndike and Watson.40 He was also a radical social reformer. For him, abandoning the myth of free will, and reorganizing society as an elaborate laboratory in which behaviour was carefully moulded by stimuli, was a utopian pursuit. This made him slightly different from the policymakers and academics of his era, for whom behavioural science was supposed to secure the social order and help the US win the Cold War against Russia. The behavioural scientists at Harvard were closely linked to the US Military, and Skinner himself had cooperated with the military during the Second World War.41 One of his major experiments was Project Pelican, where he deployed his theory of ‘operant conditioning’ to train pigeons to fly planes and drop lethal missiles while keeping pilots out of harm’s way. The programme was surprisingly successful, but was never implemented.42 In the Cold War years, however, Skinner was sceptical about the widespread anti-communism of the time, and was suspected by the authorities because he opposed nuclear testing. He was far more interested in reforming American society than Russian society.

  To reform American society, Skinner had to destroy what he thought were its ruinous myths of ‘freedom’ and ‘will’. These concepts, he claimed, were literally nonsense: they described no observable reality. The same was true for other terms used to define mental states. In Science and Human Behavior, Skinner insisted that emotions were ‘fictional causes’ of, and an unscientific way of describing, behaviour.43 All of these states could be redescribed as behaviour produced by a good stimulus or a bad stimulus: a ‘positive’ or ‘negative’ reinforcement. Frustration, for example, was the behaviour emitted by a test subject not receiving an accustomed reinforcement. Loneliness was just a special form of frustration. It was not that Skinner didn’t believe in mental states. He was, like most behaviourists, agnostic about them. As long as he had the means to observe behaviour up close, he didn’t need to infer anything about mental states.44

  The utopian undercurrent of this approach was the belief that human behaviour could be regulated to avoid unnecessary harm. This was first fully outlined in Skinner’s bestselling science-fiction utopian novel, Walden Two.45 The title evoked the libertine philosophy of Henry David Thoreau, and Skinner even expressed some interest in nineteenth-century anarchism. But the utopian community of the book is closer to the ‘Bensalem’ of Francis Bacon’s New Atlantis, a New World colony ruled by a scientific caste dedicated to enlightenment. Rather than being run by scientists directly, however, Walden Two is ruled by behavioural engineering: a sort of algorithm, manipulating the environment to produce good citizens. The algorithm could go on being updated to account for the latest scientific research, and it would be free of the moralism and bullying associated with doctrines of ‘free will’. Since choices were determined by reinforcements, bad behaviour reflected a failure in the system. Punishment was abandoned, restrictions on sexual love dropped and the workload radically reduced to give workers more time for creativity.

  Skinner tried, repeatedly, to develop technologies that would implement his ideas. For example, in the post-war era he developed and marketed a teaching machine to eliminate failure in the classroom. The machine would pose quick questions or supply sentences with blanks to be filled in. The students would mark the answer on a strip of paper that the machine would read and assess. This was a perfect behaviourist technology, because it treated its users as learning machines. It varied the pace and pattern of stimuli to keep users attentive, much as Facebook algorithms engage users, effectively ‘teaching’ them how to behave on the machine by varying the content of their feed. For Skinner, the machine would remove the arbitrariness and inefficiencies of human teachers. And it would change student behaviour by teaching them to be right.

  One obvious problem with this is that quite a lot of what there is to teach can’t be quickly tested. You can test knowledge of historical dates, mathematical equations and capital cities. Anything more complex, like critical analysis, is beyond the ken of a machine. When there is no right answer, students have to learn how to be wrong. They have to give up and mourn their mistaken belief that they know everything.46 Another problem is that we are not learning machines. So what can a teaching machine do with the part of us that never learns anything? How to educate the part of us that stubbornly entertains unrealistic fantasies and unreasonable passions, regardless of reality, cleaving to self-destruction in the face of all warning? Behaviourism blithely overlooks this everyday reality, or treats it as an inconvenience to work around. Yet arguably it is this irrational kernel, this human oddity, that gives us the desire to learn anything in the first place.

  The most important problem with teaching machines, though, is political. In Walden Two, the community is overseen by a benign tyrant, Frazier. In defending his techniques, Frazier argues that the alternative is to leave them in the hands of wicked movements like the Nazis. This comparison only serves to illustrate his authoritarianism. The fantasy is that it is possible to know, through scientific research, what is g
ood and how people ought to live. It is a fantasy in which meaning is replaced by technique, and all that is contrary, disputatious and unpleasant in social life is replaced by a smooth surface and flow. (Perhaps it is no coincidence that the aesthetic of late capitalism, and particularly of smartphones and apps, is so obsessed with smoothness and flow.)47 This requires relentless intrusive surveillance and laboratory-like manipulation of the entire population. But the secret of the good life is not something that can be known, it being different for everyone. So, behind the rule of science and technology, there has to be a tyranny somewhere making these decisions. A small number of real-world communities attempted to emulate Walden Two, with varying degrees of success, one of the main drawbacks being that leaders often identified with the benevolent authoritarianism of Frazier.48

  Radical behaviourism produced bad utopias and bad theory. Beginning in the 1970s, it was overtaken in the field of psychology by cognitive approaches which were more interested in analysing mental states. Nonetheless, bad theory sometimes produces useful techniques. For example, a teaching machine might not know anything about human desires, but a highly sophisticated machine with enough data could learn to manipulate them. By picking up on regular behavioural patterns, it could learn how to ‘teach’ minds, to train the brain’s attention in particular ways. Sure enough, behaviourist ideas have gained traction. Having lost ground in psychology, they filtered into neuroscience, which was taking an aggressively reductive turn. By the early 1990s, brain scientists had come to believe that mental states could be explained by the physical structure of the brain, which in turn could be explained by genetics and environment. Rather than wrestle with the complexities of mind, meaning and motivation, it was sufficient to study the brain as an organism. This belief was not only congruent with behaviourist ideas about conditioning, but was strongly influenced by behaviourism.49 And it was extremely useful for the pharmaceutical giants. For example, if mental states like depression or anxiety could be understood as chemical states, they could be treated with ‘happy’ pills.