Skip to content

The Automation Myth

Robots won’t build the classless society

Most important, computerization has reduced neither the workweek, the promise of all techno-utopias since the 1950s, nor the burden of physical work. We now work more than ever.

—Silvia Federici, “Reenchanting the World: Technology, the Body, and the Construction of the Commons”

In 1963, the Black working-class revolutionist James Boggs wrote of a coming cataclysm in American industrial production. As an autoworker at Chrysler in Detroit, Boggs had an intimate knowledge of the changes introduced on the shop floor and the impacts that reverberated from them across the city, seeing the ways in which these shifts in the technical aspects of the labor process came to affect the prospects for radical organizing. He saw one particular harbinger of this coming utter devastation of the working class, and especially the Black working class—automation. It makes sense that a Detroit autoworker would find himself especially attuned to this phenomenon. The contemporary usage of the word automation has its origins in the Automation Department at Ford Motor Company set up by vice president of manufacturing Delmar Harder in 1947—even though Harder’s actual proposals for the reorganization of work in Ford’s factories primarily relied on nineteenth-century technologies designed simply to speed up the production line. Throughout The American Revolution: Pages from a Negro Worker’s Notebook, Boggs speaks of automation in apocalyptic terms, issuing the grave pronouncement that “America is headed toward full unemployment, not full employment.” By the early 1960s when Boggs wrote of this fast-approaching wave of automation, the term had come to mean the replacement of jobs once done by human laborers, now performed by an integrated system of machines that themselves come to regulate the pace of production for the smaller number of workers on the line.

Central to Boggs’s analysis of automation is his insistence that the process will create a massive surplus population of “outsiders,” people whose labor has been made superfluous and obsolete who cannot get a foothold anywhere in the labor force. He argues that Black industrial workers will be hit first and hardest by this unemployment but assures his readers that it will come for all workers in time:

Automation replaces men. This of course is nothing new. What is new is that now, unlike most earlier periods, the displaced men have nowhere to go. The farmers displaced by mechanization of the farms in the 1920s could go to the cities and man the assembly lines. As for the work animals like the mule, they could just stop growing them. But automation displaces people, and you don’t just stop growing people even when they have been made expendable by the system.

Boggs’s vision of the world to come is a harrowing one indeed: even as work becomes increasingly scarce, the capitalist world-system will inevitably still condition the right to live upon the production of surplus value. This warning of automation’s capacity to create a crisis of joblessness and an ever-expanding surplus population still echoes in the contemporary discourse which always suggests a world of labor on the precipice of coming undone. Though this description of full unemployment did not unfold as Boggs predicted, the processes of deindustrialization and class decomposition certainly did, not just in the United States but all around the globe, even in nations whose industrial booms came much later in the twentieth century. To what degree can we blame automation for this shift? Has the coming onslaught of automation merely been delayed or are we living through a longue durée of disappearing labor? What is the future of work? And above all else, how can we collectively ensure the possibilities for organizing ourselves as the ground shifts underneath our feet?

Three more recent works of economic analysis provide a much-needed critical lens on the future of work, each arguing that not only is an impending future of laborlessness unlikely to arrive but also that the promise of automation (always doubling as a thinly veiled threat to workers) has been used to paper over the larger structural fault lines of a festering global regime of capital accumulation. Taken collectively, they portray a world economic system in long-term crisis and a global labor force increasingly stuck in low-wage service work while living under austerity regimes that have stripped both labor protections and social services to shreds. Automation’s acolytes often suggest that the hour of drudgery’s end is nigh, that a techno-utopian abundance, a land of milk and honey, is just around the corner if only we can have faith in technology’s promises just a little bit longer. Contrary to this assertion, these authors demonstrate that the work we do is in fact changing, just not in the ways we have so often been palliatively foretold by the carnival barkers of automation.

In 2020’s Automation and the Future of Work, Aaron Benanav contends that the resurgence of automation-related anxiety stems from the current harsh economic reality: “There are simply too few jobs for too many people.” The cause of this labor crisis, however, is not simply that automation has increased labor productivity and inversely decreased the demand for labor; instead, Benanav argues that the current under-demand for labor stems from “overcrowded global markets for manufactures, declining rates of investment in fixed capital, and a corresponding economic slowdown.” Certain segments of the industrial economy saw some processes automated, but this hardly explains why industrial output growth has stagnated for over a half century. As global manufacturing capacity grew, so, too, did competition in the global marketplace, which led to falling profits and, eventually, to manufacturers limiting output. Rather than deindustrialization happening in one country due to offshoring production to another, this widespread industrial overcapacity explains why it occurred worldwide, gradually moving outward from the capitalist core in the global North to the periphery in the Global South. As a result, companies are no longer willing to expend money in long-term fixed capital assets, opting instead for stock buybacks. With no alternative to the growth engine of manufacturing, capital chases its own tail, opting for the short-term windfalls of financialization. This mass scale disinvestment explains why so many analysts put their faith in automation. If the massive post-1945 manufacturing growth was but a brief blip on a long-term chart of capitalism’s global trajectory, the managers of capital can no longer jumpstart the engine. Instead they proclaim that we must reimagine and reorganize capital to save it, and that automation has or will force us to make these changes. But their new system relies upon a rate of growth it cannot possibly achieve, revealing the problem of overcapacity to be far more dire.

Automation and the Future of Work suggests that what we see is not Boggs’s eschatology of an economy moving towards full unemployment but rather one of incessantly growing underemployment. With overcapacity in industry, reduced unemployment benefits, and low union density, workers all over the world have been forced into deskilled, low-paying jobs with little to no substantive labor protections:

The postindustrial economy we have inherited, finally on a world scale, is, however, rather unlike the one whose emergence the American sociologist Daniel Bell first predicted in 1973: instead of an economy of researchers, tennis instructors, and Michelin-rated chefs, ours is predominately one of side-street barbers, domestic servants, fruit-cart vendors, and Walmart shelf stackers.

In deindustrialized economies, workers have been absorbed by the stagnant service industries, an area which sees output growth predominately by expanding how many people it employs. Faced with a ceaselessly low demand for labor, millions end up stuck accepting low wages lest someone just as desperate step in to fill their position. Stuck in these dead-end jobs (many in the so-called gig economy), countless folks end up caught in a cycle of debt just to get by, with no way to get out. With workers thus unorganized and unprotected, employers can simply pay the bare minimum, assured that they can always secure replacements for alienated employees.

But what if automation comes for jobs in the service sector? Wouldn’t this surely trigger Boggs’s predicted cataclysm of mass permanent unemployment? Jason E. Smith’s Smart Machines and Service Work: Automation in an Age of Stagnation argues that not only do companies lack any incentive to invest in this technology (why spend money on fixed technological capital when you can rely on paying workers far less to do the same job) but also that many jobs in the service sector, like care work, simply cannot be automated as they rely on processes no machine can perform. Unlike Benanav’s meticulous exposition of quantitative data, Smart Machines and Service Work is a work of political economic critique, which deftly disassembles the primary assumptions the field of automation (and capitalist economics more generally) rests upon. “The notion of a service sector, like the category of ‘services’ more generally, conceals as much as it clarifies,” Smith writes. “The more critical pressure is placed on this concept, the less useful it becomes as a tool of analysis.” If four out of five jobs in the United States qualify as existing within the service sector, what exactly ties together the labor of a day trader, a security guard, a middle manager, a line cook, and a teacher?

Smith, placing himself in a Marxist tradition of critique, suggests that we instead understand our current economic stagnation as a result of the proliferation of labor that does not produce value. He points to the expansion of managerial supervision and circulatory labor as prime examples of unproductive labor. Whereas the fast-food worker performs work that looks and feels like Fordist production, the manager, like the cop, largely serves to enforce discipline and surveillance. Similarly, those vast circuits of transit which take a commodity from its point of production, to a shipping container, to a port, to a warehouse distribution center, and eventually to your doorstep in a single click, do not actually create value, they merely realize it. Most of the gains in productivity appear less as a miracle of automation and more as employers sweating increasingly more out of a workforce they see as disposable.

Key to Smith’s argument is his rigorous skepticism towards much technology as such, rightly analyzing how most of the so-called developments of our era are little more than reworkings of prior technics, just packaged a bit sleeker, made a little smaller, and running a bit faster, but never able to fulfill even their modest promises of making life and work more tolerable. He writes that “the age of the computer has turned out to be a dud,” unable to turn around a sluggish world economy, while also observing that most much touted innovations amount to little more than “a tsunami of infantilizing gadgets that double as tracking collars for adults and children alike.” Most of the workplace innovations these companies have engineered amount to renting out digital platforms, charging tolls for their use, mining consumer data, and dissolving labor contracts by reclassifying their workers as freelancers. None of these advance us even an inch toward the lives of abundant leisure so many automation theorists promise. Instead, so long as companies continue to benefit from more technically efficient production processes, the service industry will continue to have a vast pool of labor to draw from, ensuring that these workers will always be cheaper than any permanent investments in automating even that segment of work which could be done by machines. Under such a capitalist regime, full automation of work simply cannot happen.

If the threat of automation remains far smaller than the existential problems of a capitalist world-system spluttering along, unable to grow, and beginning to show the inevitable signs of rot, then surely we must begin to look to other escape hatches. Benanav argues that neither neo-Keynesianism nor UBI programs are adequate solutions. Without a social movement capable of embarking on a quest of production that could actually place the means of our labor into our own hands, piecemeal reforms remain inevitably susceptible to the capital strike. Given that most workers “are paid low wages, forced out of cities, denied public services, are not unionized, and do not perform activities that are located at strategic points in the economy,” Smith doubts that any residual forms of twentieth-century organizing will be able to adequately address the fragmented, isolated, and alienated conditions so many now work within. To echo that old Lenin query, what then is to be done?

While Benanav and Smith both provide crucial insight into the causes of global stagnation and its effects on the kinds of work we do now, Gavin Mueller’s Breaking Things at Work views the story from an entirely different vantage, focusing on the workers who have struggled against mechanization and automation from its inception, telling a cultural history of machine smashing as revolutionary praxis against the onslaught of capital’s attempts to take more control of the workplace. Mueller’s work is a counter-history of automation, attending to all those who have fought back at every turn, acting out of a desire to maintain as much collective autonomy over what it means to work as possible. Mueller argues that technology “robs people of the feeling that they can control their own lives, that they can set the terms of their world.” He centers his story around the Luddites, a group of “discontented weavers, croppers, and other textile workers,” who in the 1810s embarked on covert machine-smashing riots in Britain to protest new technologies being introduced that would turn their professions into deskilled piecework. He quickly dispels the common misconception of Luddism as somehow synonymous with a vague and regressive primitivism and instead reclaims the legacy of the Luddites as one in which everyday workers engaged in collective action against unwanted changes in their labor conditions. In Mueller’s formulation, “to be a good Marxist is also to be a Luddite.” The Luddites serve as inspiration to our present. They mark a tradition to be learned from, one which can remind us that despite all those managers and guards placed in our workplaces to ensure speedups and intensifications run smoothly, workers themselves can always throw a hitch into the operation, can always resist the attempts to deskill and gigify our jobs.

Mueller argues that automation does not so much replace humans with machines as it remolds work altogether, “isolating and rearranging tasks, altering job descriptions, and hollowing out middle-tier occupations.” In reaction to this, workers have historically participated in a variety of resistances which echo these Luddite methods, including wildcat strikes, slowdowns, sabotage, theft, and outright refusal to work. Union musicians in St. Louis “planted time bombs that damaged the Vitaphone sound film equipment that had replaced them”; American soldiers in Vietnam refused orders, destroyed equipment, and fragged their commanders; more recently, ILWU dockworkers in Seattle “damaged freight cars, dumped grain, and attacked windows with baseball bats during a contract dispute,” and Amazon warehouse workers continue to find ways to defy the surveillance regime meant to regiment their every movement in the workplace. Breaking Things at Work draws these legacies into a cumulative strategy for how we might come together to combat the daily indignities and miseries of contemporary work, advocating on behalf of a politics of deceleration, “a politics of slowing down change, undermining technological progress, and limiting capital’s rapacity, while developing organization and cultivating militancy.”

The decelerationist tendency of Luddism then is primarily about recapturing control of the terms not just of our work but of all of our social relations. It signals the possibility of exerting collective force over our lives, which have been increasingly dictated by the terms of the wealthy. Mueller suggests that this can be coupled with a politics of degrowth, of envisioning a world not built on the mass extraction of resources required for the continual expansion of profits “as the advanced technology of Silicon Valley is built with rare earth metals mined ‘artisinally’ by children using primitive tools.” Along with deceleration and degrowth, we can add decarbonization to this mix, since the current organization of work is directly responsible for the rapidly accelerating destruction of the planet. The political praxis of Luddism offers up a way of linking concurrent struggles together and of acting now, for coming together within and outside of the workplace to reassert our collective autonomy, to begin to actually build a new world by taking apart our present one.

Ultimately, so much of the discourse around automation reveals what we currently lack. We do not possess the leisure hours we deserve. Our workplaces already treat us like the robots they threaten to replace us with. We have ridden out a pandemic ever more reliant upon our smartphones and computers while being reminded of how isolated so many of us already had been made by these very technologies. Capitalists themselves have no solution to offer, not even an appealing fantasy to sell. Even the ludicrous dreams of Bezos and Musk to use space as an escape hatch from a climate ravaged world expressly rely upon transporting the division of labor off-world fully intact. As Boggs told us in 1963, we desperately desire to lead lives with meaning that decouple our work and our worth. We need a life not beholden to the value form. We need a labor no longer doggedly paced by Frederick Taylor’s stopwatch. We need a planet that remains inhabitable and biodiverse. Machines are not coming to make a better world for us. Robots won’t build the classless society. That historical task, as always, remains solely our own.