We've Already Lost the Battle Against the Machines
Picture by Anthony Harvey PA Archive/PA Images


This story is over 5 years old.

Job Stealing Robots

We've Already Lost the Battle Against the Machines

Robots are taking our jobs because our lives are already robotic.

It's finally happening; the robots are finally taking over. All the vast techno-nightmares that hacked into your imagination are about to come true. The self-driving cars will carry out their massacres on the streets, ramming into crowds of innocent pedestrians according to their obscure and misanthropic moral code. Your fridge and your kettle—isn't it cool that they're connected to the internet, I can tell the kettle to boil using the app!—have already risen up in revolt, smashing their data-chains, taking away the only thing that can still give you pleasure.


Panic in the cities, toasters burning their owners, washing machines eating you alive with their flappy Perspex mouths. It can't be long now: the war of the machines is clanking horribly on the horizon, and soon every person will have their weak organic flesh churned up in the spinning teeth of the robots, to be turned into lubricant or biofuel; the globe-straddling technical apparatus has no need for us any more, and with a computer's cold precision it'll decide to kill us all off. It's already started.

Automation is no longer just a problem for people who do actually useful work—building things or knocking them down, driving stuff around, tilling the fields, putting products together in factories or digging through the earth. It's come for your job. It's come for the boring respectable office work you're trying to avoid by reading this right now. In Japan—of course it's Japan—the Fukoku Mutual Life Insurance company has fired 34 of its workers and replaced them with an artificial intelligence system. After all, computers don't take endless tea breaks, or get pregnant, or read vice dot com instead of doing what they're paid to do. The IBM Watson Explorer system is better than you in every conceivable respect. Not just at work: one day you'll come home to find a touchscreen in your house, sleeping with your partner, raising your children, watering your allotment, algorithmically generating that novel you've always wanted to write. You're done. Get over it.


At least, that's one way of looking at it—one that's shared by all the people quite reasonably afraid of losing their jobs, the plutocrats giggling to themselves over the prospect of a fully-automated money-generator, and the new brand of techno-leftists seeing in our crisis the prospect of a future without work. But their mistake is imagining that the robots are only taking over now. In fact, it happened a very long time ago.

What are the jobs that are starting to be taken over by machines? The 34 white-collar workers about to lose their livelihoods to a sophisticated tin-can opener work at an insurance firm; their job is to look at legal data and calculate the lowest possible payout their company can give to sick and injured people while still not breaking the law or their contract. What this job really consists of is the bloodless, inhuman, mathematical administration of other people's lives. And to be clear, this isn't their fault; it's a job, everyone needs one, and in a world whose main output is small anonymous tragedy none of us can really escape being complicit. There are very few people who grow up with soaring dreams of being a claims adjustor—it tends to just kind of happen, as if automatically; all the machinery is already in place. Our form of social organisation is one that's already robotic—it functions all by itself, entirely alienated from the wants and needs of the people who use and operate it. It's just that its moving parts are made of human bodies, sat behind millions of desks, doing all the meaningless and unrelated bits of paperwork that together compromise the code for an enormous, stupid engine. If so much intellectual labour can be done by machines, what kind of work is it that we're doing?

The American sociologist Lewis Mumford called these forms "megamachines", the machine seen not as a distinct object but as a way of looking at hierarchical organisations. The pyramids of Egypt were built through the power of "murderous coercion", one capable of "turning men into mechanical objects and assembling these objects into a machine", something vast and intricate, even if the most sophisticated objects involved were ropes and pulleys. Later machines, he noted, weren't necessarily new inventions; they only replicated the work already being done by humans, but cut down on the amount of human labour-power required; all our sophisticated robotics is just a faint echo of the whip and the chains of ancient slave societies. (And, as Marx pointed out, with every new generation of technology the whole apparatus becomes less able to sustain itself and more expropriative.) The megamachine isn't for anything; it has no object other than its own reproduction. And since the Industrial Revolution they've grown more totalising every year. In commodity-production, all people are reduced to their labour-power. We don't need to worry about people being minced up by indifferent mechanical wheels in the near future; it's already happening now. It's been happening for centuries.

Replacing human bodies with software is just the final step; it can only happen once what we do is already robotised, once people have already been turned into objects. Another detail: elsewhere in Japan, the economy, trade and industry ministry is introducing another AI to take over the work of bureaucrats and researchers, spitting out stats and information for ministers before they face their colleagues or the press. If the rational and vicious micromanagement of human existence is becoming fully automated, then of course it's politics that's next.

In the United States, it was announced after the election—in what was hoped to be a triumphant flourish, but turned out to be more of a measly, shamefaced excuse—that the Hillary Clinton campaign's strategy was devised not by human tacticians who actually knew about the people who would be going to the polls, but by a sophisticated AI nicknamed Ada. The computer was fed vast amounts of polling and demographic data, and attempted to reasonably predict what Donald Trump would do, and the best way to counter it. As we all found out, reasonably predicting what Donald Trump will do is a pretty stupid game. But again, the automation was just the last step. Throughout the campaign, Hillary Clinton gave every impression of being a faulty machine: stiff and inelegant, prone to suddenly crashing, spitting out the same information over and over again, running through what was clearly a set of optimal programming rules. She was something imposed on the world, like so much new technology, not anything it ever really wanted. Or look at that old video of George Osborne repeating the same drab soundbite over and over again to a news team, stricken and sweaty, utterly mechanical. The machine war already happened. We lost.