One January night last year, I disappeared into two protests of Donald Trump’s travel ban at once. My attention split between the jostling bodies around me — we chanted, “let them in!” — and the flux on Twitter. During the protest my tweets were like those of many other people who were there: quotations from the crowd, pictures of signs and of bemused travelers stepping into an airport terminal filled with shouting people. Some time into the protest, the physical jostling and chanting ceased to be a surrounding discomfort and felt more like a coming together. When my pictures or quotes from the protest were retweeted, I felt like it was only extending that feeling beyond the terminal. I pictured blue-lit faces elsewhere in dark space, solitary extensions of the crowd, even as the event itself continued to shake my bones and thunder in my ears.
But after I returned home, my fragments within the networked protest took on a life of their own. Overnight, my tweets accrued protesters and counter-protesters in numbers far beyond what I had seen in person, becoming a tireless cacophony as I slept, and by the morning I was looking at a protest almost unrecognizably transformed.
I had become the spectator to a thing I had made to reflect my autonomy as it became integrated into a different kind of affective machine
Elias Canetti, in Crowds and Power (1960), described the “discharge” of a crowd — before which the crowd doesn’t really exist — as the moment in which “distinctions are thrown off and all feel equal,” where “each man is as near the other as he is to himself; and an immense feeling of relief ensues.” This, of course, is an illusion; the members of a crowd are not, nor will they continue to feel, equal. “They return to their separate houses, they lie down on their own beds, they keep their possessions and their names,” Canetti writes. “ Caught in the amber of my Twitter feed was disquieting proof of a phenomenon I had appreciated the day before and now found threatening — that is, my part in the nationwide protest did not cease when I went home, but became crystallized and seemed to act on its own. In my absence, the slogans and images I had shared were inserted by others into new frames of reference, perforated by extraneous debates, borrowed, criticized, bundled alongside many voices offering nearly identical accounts. I had become a witness to my own words, just like the blue-lit faces of the audience I had imagined the night before, and from that perspective the words seemed like just another move in the ineffectual pistoning of American public discourse.
The same technology that had helped to call the protest into being made visible, by recording it, what it means to be swallowed by the crowd. At the protest, my feed had proved to me that I was both a participant and an observer, that my personality had not quite dissolved in the moment and the mass but remained outside it. Now I saw my “feed self” caught up in an even bigger crowd. Paradoxically, I had become the spectator to a thing I had made to reflect my autonomy as it became integrated into a different kind of affective machine.
I was frightened to see myself rendered no more than the periscope and megaphone of a machine. I was frightened to feel like a robot.
I’m not the first to be frightened by a convergence of the apparently robotic and the human. The human-passing robot is a theme for concern that shows up repeatedly in science fiction: Blade Runner, Terminator, Battlestar Galactica, Eve of Destruction. It’s also the subject of prolonged reflection in the academic world, where the Turing Test integrated the idea of human-passing robots into the effort to define intelligence. We test the power of our computing machines by seeing if they can pass for human, talk like us, beat us at chess. Now robot paranoia has emigrated from the realm of imagination and speculation to everyday life. Among the worries regularly encountered on op-ed pages are robots replacing us at work, robots passing as humans on social media to conduct political psyops, robot armies led by hackers of the insecure “internet of things,” and robots replacing us as objects of desire for each other.
But in a related development, the fear of robots has shifted to contempt, and increasingly, “robot” is being used as a human insult. At a memorable debate in the 2016 Republican Primary, Senator Marco Rubio awkwardly over-repeated a poorly phrased talking point and brought a mountain of derision down on his own head: He looked, people said, like a malfunctioning robot. Mark Zuckerberg receives similar comparisons for virtually everything he does. Repetition and predictability, and the appearance of acting out of premeditation rather than with human spontaneity, are scorned as robotic behavior.
Regardless of where one is on the political spectrum, one is vulnerable to some version of that insult. When the alt-right edgelords gleefully ask, “Are you triggered yet?” they are trying to project mastery of the presumed automatism of their target. On the left, Sanders supporters were accused of being “Bernie bots.” Trump’s army of social media warriors are sometimes treated to their own variant of the insult: “I get called a Russian bot 50 times a day,” one woman told Politico.
The all-purpose insult only works because of its all-purpose applicability. Every group accused of roboticism is perceived to be parroting party lines, choosing friends and enemies on the basis of blind team-allegiance. This means that having almost any identifiable political position makes you vulnerable to being called a robot. At the same time, the insult itself has become such a cliché that using it is robotic itself. The woman so offended to be called a robot in the Politico article earned the label by routinely participating in organized propaganda campaigns on Twitter, tweeting and retweeting a predetermined message hundreds of times a day — and ironically felt free to criticize what others said to her as rote and repetitive.
The robot is perhaps such a favored metaphor because real robots exist in the same space. From innocuous experiments in the generation of words — I recently made a bot that invents new story ideas, for example — to the deeply troubling astroturf botnets massed to manipulate public opinion, “bots” surround us online, resembling more and more how humans themselves communicate online, using the same small palette of pre-set responses that we do.
Every group accused of roboticism is perceived to be choosing friends and enemies on the basis of blind team-allegiance. Having almost any political position makes you vulnerable
But the chants of protesters massed to express themselves with one voice may be both as robotic as the retweets that help a viral marketing campaign land as well as a form of massively enhanced collective agency. Conceiving of a protest as robotic because protesters form a kind of team, share a message, and repeat a phrase smuggles a false implication into an apparently innocent descriptive metaphor. The robot acts on programming rather than thought: It is not free. This hidden premise of the insult explains why it is so tempting to project roboticism outward onto our enemies. After the protests against Trump’s travel ban, one of the right’s favorite ideas kept coming up: We protesters had all been paid by George Soros to be there. What other explanation could there be for such a massive, spontaneous, apparently well-coordinated demonstration? How else did we all know to chant “let them in?” The enemies of collective agency can’t afford to acknowledge the possibility of self-organization: It is so much more comforting to imagine your enemy as an an army of pre-programmed robots than a real human crowd.
When I looked at my Twitter feed the day after the protest, as if by osmosis I saw myself with those same cynical eyes. I knew I hadn’t been paid by Soros, of course, but I felt discomfort at an apparent roboticism all the same. Did my discomfort indicate that I had been infected by an atmosphere that cynically misconstrues all collective action as unfree, misapplying the robot insult with reckless abandon?
The jurist and philosopher Carl Schmitt bluntly articulated why human political interactions tend toward a kind of automatism. His ideas are dark and properly suspect, for reasons his ominous nickname — “the crown jurist of the Third Reich” — makes clear. But they also possess great explanatory power: In The Concept of the Political, Schmitt argued that political disagreements are not political because of where they take place, whether the halls of government or the most private dinner table; nor because of who disagrees, whether the most prominent politicians or the most tuned-out high school students; nor because of what the disagreements are intended to accomplish, whether legislating or winning an argument over the turkey and stuffing. Political disagreements, Schmitt wrote, are political because they divide us into friends and enemies: Friend/enemy is the political distinction. In politics, it comes before everything and contains everything. Arguing about principles is a rationalization of or distraction from the fundamental reality of struggle. The principles are almost arbitrary.
When some ordinary disagreement becomes political, in Schmitt’s view, the content of the disagreement becomes secondary to the fact of antagonism. Thus we often see alliances that appear hypocritical but make perfect sense from inside the political distinction. Religious groups embrace monstrously immoral figures; doves invite hawks into their resistance; rebels and the establishment lock arms — motivated to be friends because they share enemies. Likewise, from within the political distinction, enemies crop up in the strangest places. In The Concept of the Political, Schmitt writes:
The political enemy need not be morally evil or aesthetically ugly; he need not appear as an economic competitor, and it may even be advantageous to engage with him in business transactions. But he is, nevertheless, the other, the stranger; and it is sufficient for his nature that he is, in a specially intense way, existentially something different and alien, so that in the extreme case conflicts with him are possible.
Politicization, in Schmitt’s sense, is a kind of roboticization. The reasoning, dithering, discriminating processes of ordinary life are swept away, replaced by a brutal algorithm of antagonism. Conceiving of roboticism in this way highlights reaction rather than repetition. What makes the political distinction take on a life of its own is the way enemies bond us to friends, reactively. Taken as a prescription — as Schmitt took it — this insight leads down a very dark road to the idea that a nation should always keep a public enemy on hand to unite the citizenry. Taken as an explanation, however, the same idea can be spun in a hopeful direction: The travel ban protests formed a more or less spontaneous collectivity because a massively oppressive act of enmity had created a corresponding group of friends.
In the days after the January protests, the political eruption was discussed ad nauseam, footage aired on every channel. Each time I’d hear the familiar chant “let them in!” booming from a wall-mounted TV in a public building, I’d feel a peculiar mixture of the energy I remembered and the discomfort that had followed. My anger had vaulted me into a specific politicization and I had already expressed it, submerging my individuality in a crowd. Thanks to the technologies that allow us to broadcast, preserve, and review such moments, I had seen what that meant, from inside and outside.
Politicization, in Schmitt’s sense, is a kind of roboticization. The dithering processes of ordinary life are swept away, replaced by a brutal algorithm of antagonism
Gradually the self-consciousness faded, until, over a year later, I heard a surprising rumor that brought all my doubts raging back. In David Wolff’s Fire and Fury, he reports that the timing of Trump’s travel ban, just before a weekend when the maximum number of protesters could be expected to turn out, was urged by his cartoonishly evil then-advisor, Steve Bannon. He wanted, Wolff reports, the “snowflakes to show up at the airports and riot.” In other words, the executive order was an attempt to “trigger” his enemies from the largest possible platform. It’s no surprise that a former editor like Bannon, familiar with the world of media and social media, where provocation and revenue pursue a rigorous alliance, would come up with an idea like that. Ironically, while Bannon attempted to create a Schmitt-style robot-enemy, his people, and in fact his own magazine, trumpeted the idea that the protesters had all been paid by Soros and were not spontaneously responding to the ban. This might seem like a contradiction, but both the provocation and the rumor rely upon the belittling assumption that those who would attend such a protest must be in some way unthinking.
The idea that I had been part of an intentionally provoked protest resurrected my misgivings. But then I realized: The assumption that the normal politics of democracy are expressions of robotic mindlessness is more dangerous than either Bannon’s provocation or Breitbart’s rumor. Bannon could no more control the crowd he had provoked than readers of his magazine could cynically explain it away. But to fear the moral physics according to which an outrage raises an outcry, or to perceive collective action as a loss of agency, is an insidious and genuine danger. The renewal of politics requires us to disentangle the image of a team as a bunch of bots with a party line from the political unity of collective agency.
This essay is part of a collection on the theme of TEAMS. Also from this week, Vicky Osterweil on how YouTube has weaponized the pleasures of hanging out, and Tom Thor Buchanan on making an enemy of your body.