Notes on Automation
I can still remember the first time I saw a “self-checkout” machine at our local grocery store. As a kid, I would always look forward to the end of each grocery shopping trip, when my father and I would wait to be called up to a checkout aisle. The fun, of course, lay in trying to guess which of the eight aisle numbers would light up first to beckon us forward. As dozens of these trips came and went, I came to possess a fairly thorough understanding of the layout of the checkout area. Hence my alarm when, one day, I saw that aisles seven and eight had simply vanished. In their place stood unmanned kiosks, complete with plastic bags and flickering red scanners.
Rather than increasing my odds of guessing correctly, this development eliminated my little game altogether. For some time thereafter, my father usually opted to check out our items himself. It’s hard to begrudge his choice; the new kiosks were simply faster than the traditional checkout aisles. Yet there was a certain hollowness to the self-checkout process that no gains in efficiency could entirely remedy. My father and I came to that grocery store often enough to recognize the employees, or, in this case, to recognize their absence. One’s interaction with a human cashier — no matter how obligatory, how superficial — was a fundamental part of the consumptive ritual, a comforting trace of the idea that commercial activity represents, or even creates, an underlying system of social bonds. The self-checkout kiosk seemed to epitomize (almost comically, with its robotic voice and stark beige paint job) commercial society’s opposing tendency towards atomization and depersonalization.
Yet I suspect that this sense of alienation wasn’t the primary source of our unease. More basically, we must have felt guilty about benefiting from a machine that had eliminated someone’s job. Our guilt was likely compounded by a narrative that has gained a great deal of traction in recent years, particularly since the 2008 financial crisis. This narrative says that the self-checkout kiosk is not an isolated phenomenon but rather a particularly visible example of a more general and potentially catastrophic trend towards “automation.” As the story goes, technological advancements in fields like robotics and artificial intelligence are enabling machines to perform complex tasks more efficiently than their human counterparts. Such innovations reduce the number of employees necessary to produce certain goods and services and, in cases like the self-checkout kiosk, can even obviate entire lines of work. Some studies seem to support an apocalyptic interpretation of the automation thesis. A 2017 report by the McKinsey Global Institute, for instance, estimated that automation could eliminate 73 million jobs in the United States by 2030, rendering over a third of the US workforce unemployed. In the face of such numbers, the doomsayers claim, we may be forced to accept a reality in which there simply aren’t enough jobs to go around.
Other economists are more circumspect. Today’s fears about automation, they say, are really no different from the many technophobic ideologies and movements that have arisen since the Industrial Revolution. While the Luddites are surely the most infamous technophobes of modern European history (and, perhaps, the most ridiculed), we would be mistaken to think of technophobia as a fringe tendency; on the contrary, some of most esteemed political economists of the 18th and 19th centuries were similarly concerned about the ramifications of technological progress. David Ricardo, for one, worried that the widespread “substitution of machinery for human labor” could easily “render the population redundant,” while Karl Marx predicted that technologically-induced “overproduction” would lead to widespread unemployment and cyclical economic crises. The automation soothsayers claim, however, that such fears have been shown time and time again to be unfounded. Although technological innovation may lead to unemployment in the short term, such progress tends to create as many jobs as it destroys. Solving unemployment, then, is simply a matter of retraining workers to meet the needs of a changing economy.
My concern is not to adjudicate between the two sides of this debate, which ultimately hinges on empirical questions of economics that I am unqualified to discuss. I would like to suggest, rather, that the debate as it is now framed misses a more fundamental issue, which we might get at with reference to what John Maynard Keynes called the “economic problem.” Keynes defined the economic problem as the “struggle for subsistence” under conditions of scarcity. With this concept, Keynes is simply articulating the fact that human beings need certain things to survive and prosper, that acquiring these things requires substantial effort, and that there is a limited amount of such things to go around. Keynes thought of the economic problem as a fundamental condition of human existence — according to Keynes, the simple fact of scarcity, of needing to labor to survive, had profoundly shaped the consciousness, social arrangements, and modes of life of all human beings.
Keynes invokes the economic problem in his 1930 essay “Economic Possibilities for our Grandchildren,” written at the outset of the Great Depression. He begins the essay by refuting the argument that declining productivity is to blame for widespread unemployment. The real problem, he writes, is that technology is advancing too quickly, that “our discovery of means of economising the use of labour [is] outrunning the pace at which we can find new uses for labour.” This argument sounds markedly similar to today’s accounts of the dangers of automation. Yet this comparison is misleading, since Keynes parts company with the automation doomsayers in one crucial respect: for Keynes, the possibility of widespread and persistent “technological unemployment” heralded liberation rather than catastrophe. Such a development would indicate that some portion of society’s labor is superfluous — that is, that society could work less, potentially much less, and still satisfy all of its needs and wants. This hypothetical society could overcome the economic problem and lead humanity back to a kind of Eden — albeit an Eden in which man’s own mechanical creations (rather than his creator) sustain his world of leisure and plenty. In keeping with this vision, Keynes famously predicted that technological progress would enable people of our generation to work a mere 15 hours each week. For the remainder of the week, people would be free to pursue activities that are not directed towards maintaining the basic conditions of life. (Politics, sport, arts and crafts, learning, and invention all come to mind.)
Obviously, Keynes’s utopian vision of a machine-driven leisure society has not come to pass. Contemporary citizens of the United States work about as much as their grandparents did. Why did Keynes get this question so wrong? Many would follow the automation soothsayers and argue that Keynes’s pronouncements about technological unemployment were too bold. “Have no fear,” they would say. “Each time technology satisfies some human desire and antiquates some corresponding line of work, the very same technology creates new wants and new job opportunities. After all, no one needed a social media coordinator before the invention of Facebook!” This line of thinking is at once pernicious and revealing. For what is presented as an empirical explanation for the persistence of good jobs is really a moral justification of labor. If it is true that technological innovations create new wants as quickly as they obviate old ones, and if work is justified as a means of satisfying human wants, then it is impossible to argue that technological progress should result in a shorter work week.
The problem with this reasoning, from a moral perspective, lies in its initial premise. It may be true that technological innovations create needs and wants of quantitatively equal worth to those they satisfy — people seem to purchase the new goods and services produced or made possible by new technologies just as enthusiastically as they did older goods and services. But when one compares the needs served by labor two or three generations ago with the so-called “needs” served by labor today in qualitative terms, a stark contrast emerges. In our grandparents’ day, the social necessity of most work was beyond doubt. If a large portion of the labor force had suddenly taken an extended vacation in 1920, when the majority of people worked in agriculture and manufacturing, the most basic conditions of human life would have been threatened. Today, by contrast, “it is not entirely clear how humanity would suffer were all private equity CEOs, lobbyists, PR researchers, actuaries, telemarketers, bailiffs or legal consultants to similarly vanish.” So writes anthropologist David Graeber in his withering essay “On the Phenomenon of Bullshit Jobs.” According to Graeber, jobs in agriculture and manufacturing were not replaced by jobs in the “service sector” (the term “service” suggests the satisfaction of some legitimate need) so much as by jobs in the “ballooning … administrative sector,” which seem to serve little purpose at all — at least, that is, outside of the internal logics of finance capitalism and political horse-trading.
The 40-hour work week seems justifiable when labor is providing people with food and simple household necessities. But when technology has allowed us to meet these fundamental needs with relative ease, and when the “needs” served by labor come to include the “need” for a perfect Facebook page, the “need” to efficiently design and market disposable consumer goods, or the “need” to accurately bet on the futures of national currencies, the endurance of the 40-hour work week begins to take on the character of a social pathology.
Indeed, as technological progress makes labor less and less necessary, as evinced by the growth of Graeber’s “administrative sector,” we seem to cling to labor ever more tightly. As Graeber writes, “It’s as if someone were out there making up pointless jobs just for the sake of keeping us all working.” In other words, from a societal perspective, work looks increasingly like a choice rather than a necessity.* The interesting question, then, is not whether technological progress will give us the capability to overcome the economic problem and drastically shorten the work week. Maybe we have already reached this stage of technological development; if not, we certainly seem to be headed in the right direction. The interesting question, rather, is how to account for our desperate will to work, for our stubborn refusal to transcend scarcity and toil, regardless of whether or not such transcendence is within the realm of technological possibility.
I would like to discuss two plausible explanations for this drive towards labor from the philosophical tradition. The first account, Marxian in origin, argues that the persistence of the 40-hour work week has nothing to do with the “will” of laborers at all. Rather, structural economic forces are to blame, forces that operate irrespective of the state of technological progress or people’s attitudes towards work. This account, which found its canonical expression in G.A. Cohen’s Karl Marx’s Theory of History: A Defense, is essentially an interpretation of Marx’s historical materialism. It will, therefore, be useful to sketch out Marx’s philosophy of history in the abstract in order to better understand the philosophical underpinnings of Cohen’s more specific claims about capitalism and work.
If Marx has anything like a “first principle,” it is that human beings are fundamentally producers, and that human history is characterized by the progressive development of humanity’s “productive forces,” the technologies and knowledges that people use to make things. For Marx, historical change is the outcome of a dialectical interaction between these productive forces and society’s “relations of production,” the economic and legal structures within which production occurs. A given set of relations of production can either promote or hinder the development and effective use of humanity’s productive forces. In cases of the latter variety, Marx holds that the existing relations of production must eventually give way to new relations that will allow the productive forces to develop further or to be utilized more effectively.
Marx cited the transition from feudalism to capitalism as a particularly consequential example of this process. Feudal relations of production — that is, guild regulations, apprenticeship statutes, and other forms of economic protectionism — shielded old techniques of production from competition, thereby hindering the development of humanity’s productive forces. The rise of capitalist relations of production liberated these forces for further growth, as strong private property rights promoted investment and market competition drove innovation. As is well known, Marx holds that a further set of relations of production, which he calls “communism,” lies beyond capitalism as the endpoint of humanity’s historical development. Marx’s historical materialism predicts that the transition from capitalism to communism will follow the pattern of the transition from feudalism to capitalism: late capitalism will begin to fetter humanity’s productive forces, necessitating a shift to new, communistic relations of production.
Cohen elegantly explicates the nature of this contradiction. The problem, Cohen claims, is not that capitalism fails to promote technological innovation in the manner of feudalism. Capitalism fetters humanity’s productive forces in a subtly different manner, by preventing people from using the formidable technological powers already at their disposal in a way that promotes human well-being. Cohen frames the problem as follows: whenever technological progress leads to gains in productivity, people must choose between two options: to keep labor hours constant and increase output, or to reduce labor hours and keep output constant (or some combination of the two). It stands to reason that in certain situations, when the work week is long and output is already quite substantial (as is the case in the present day), people would find a shorter work week more desirable than ever greater quantities of consumer goods. Yet capitalism is systematically biased in favor of increasing output rather than shortening the work week. In fact, two essential characteristics of capitalist relations of production — the profit motive and market competition — all but ensure that firms will adopt a strategy of output maximization. If a firm fails to capitalize on technological innovations by expanding production in a given market (or, when production in a given market can expand no further, by reallocating resources to more receptive markets), it will be outcompeted by other firms and cease to be a lucrative enterprise. Technologically-induced gains in productivity will fail to result in more leisure time for the average worker, then, so long as the profit motive and market competition continue to dictate the nature of production.
Cohen’s Marxist account of the relationship between capitalism and the length of the work week is a useful corrective to Keynes’s essay, which tends to elide the fact that significant economic reforms would have to precede any radical attenuation of the work week. One need not accept all the dogmas of historical materialism to recognize that the economic deck, so to speak, is stacked in favor of work. Keynes is, however, attentive to a powerful subjective dimension of the present-day persistence of work that is often trivialized in orthodox Marxist accounts. Keynes writes:
There is no country and no people, I think, who can look forward to the age of leisure and of abundance without a dread. For we have been trained too long to strive and not to enjoy. It is a fearful problem for the ordinary person, with no special talents, to occupy himself, especially if he no longer has roots in the soil or in custom or in the beloved conventions of a traditional society.
Keynes’s pessimism in this passage stems from his intuition that work is one of the primary practices in the modern social world that imbues people’s lives with a sense of purpose. Keynes implies that work acquired this significance following the dissolution of “traditional” European forms of social life; in these past societies, the social whole commanded the devotions of the individual and provided her with manifold opportunities to engage in meaningful activity, often by making sacrifices for and executing duties to the community at large. As this “traditional” society gave way to a more atomized “mass” society in the 18th and 19th centuries, Keynes suggests Europeans found a new source of meaning (or, at least, a new “occupation”) in work. Precisely how work came to possess this new status — improbably, it should be noted — is a fascinating story, but one which lies outside the scope of our present concerns.† Instead, I would like to reconstruct one particularly sophisticated account of the subjective significance that work holds for modern individuals, which, I think, offers profound insights into our contemporaries’ refusal to give up on the idea of the career, the 40-hour work week, and all the other trappings of “work” in its modern guise.
This account is to be found in Hegel’s Elements of the Philosophy of Right, the definitive statement of Hegel’s ethical, social, and political thought. Hegel’s goal in the famous third section of the text is to enumerate the primary institutions of the modern social world and to justify these institutions as potentially rational or good (even if they fail to live up to this potential in their actually existing manifestations). In Hegel’s view, an institution qualifies as rational or good if it promotes what Frederick Neuhouser has termed “social freedom.”‡ Social freedom is an extremely rich concept, more complex than either the “negative” freedom of the liberal theorists or the moral autonomy proposed by Kant and his followers. The sole aspect of social freedom that concerns us here, however, is Hegel’s claim that modern individuals both acquire and actualize their conceptions of self, and thereby come to perceive their lives as meaningful and themselves as valuable, by performing particular roles in rational social institutions. This formulation is, admittedly, quite abstract — to make Hegel’s point more concrete, I’ll detail the way in which this aspect of social freedom obtains in the specific context of “civil society,” one of the three institutions that constitute Hegel’s idea of the rational social order. Usefully for our broader purpose, the roles that individuals inhabit in civil society are above all professional roles; thus, any explanation of the way social freedom is realized in civil society will necessarily contain Hegel’s account of the significance of work.
Hegel uses the term “civil society” to denote the realm of employment, production, and market exchange, something similar to our contemporary idea of the “economic sphere” of life. The participant in civil society conceives of herself as a private individual, largely without obligations to others, whose primary goal is to advance her personal (usually pecuniary) interest by working in a profession and playing the market. According to Hegel, however, participation in civil society is not merely of instrumental value to social members as a means of acquiring money or goods. Rather, members of civil society also view their participation as intrinsically valuable, for it is in this participation that members — to repeat the above formulation — both acquire and actualize their conceptions of self.
By an individual’s “self-conception,” I am referring to the individual’s sense of her own identity, to those aspects of her being that she views as integral to her selfhood. Hegel’s basic thought is that modern individuals derive their self-conceptions from the roles they adopt in rational social institutions. The socially free individual comes to view her social role not as a set of oppressive obligations (that is, as something alien) but rather as an identity that she can affirm as her own. Crucially, this socially-derived identity in turn furnishes the individual with the projects and goals that she takes to be essential to her life’s purpose and that endow her life with meaning. In the context of civil society, the social identities Hegel has in mind are professions: in conceiving of oneself as a doctor, or a plumber, or a teacher of Hegelian philosophy, one acquires definite ideas about the capacities one should cultivate and the type of conduct one should uphold as the bearer of this identity, which functions as a roadmap of sorts for the course of one’s life.
For Hegel, however, work does not simply shape the self-conceptions of social members at the abstract level of consciousness; rather, work also offers social members the opportunity to actualize their self-conceptions in the objective world and to achieve recognition from fellow members of civil society. In the act of plumbing — or, rather, in the act of plumbing well, following a period of serious study — the plumber’s lived experience is unified with her self-conception. Further, since the plumber’s professional identity is socially derived, the excellence with which the particular individual embodies the abstract category of “plumber” is evident to society at large. Through the competent performance of her chosen occupation, then, the plumber achieves social recognition as an individual of standing or value. Crucially, it is this social recognition that makes the plumber conscious of the fact that she is successfully enacting her self-conception in the sense outlined above. Social members’ consciousness of the unity between their self-conceptions and their actions is, ultimately, the source of their self-esteem. Insofar as members of civil society define their identities in terms of their professional roles, each member’s very sense of self consists in the feeling that she is effectively performing the duties of her chosen occupation. Members of civil society who lack this self-consciousness, conversely, perceive themselves — that is, perceive their selves — as fundamentally incomplete.
If Hegel’s characterization of civil society is to be believed, then we must accept that work plays a tremendously important role in the modern social world as a practice through which individuals constitute their subjectivities and acquire a sense of self-esteem. But do people today really derive meaningful self-conceptions from their professional lives in the sense that Hegel describes? While Hegel’s account of work’s significance still appears plausible in certain cases (many teachers, doctors, and skilled craftsmen, for instance, seem to take their professional identities quite seriously and derive satisfaction from doing their jobs well), it would be hard to argue that the Walmart employee, the call center operator, or the investment banker could (or should) take an attitude towards her job of the sort detailed above.
Yet before we dismiss Hegel’s discussion of work as irrelevant to our present social reality, we should remind ourselves of the purpose of Hegel’s project. Hegel did not mean to argue that the civil society of his day actually provided people with social freedom; rather, he claimed that the institution of civil society in the abstract had the potential to provide everyone with a meaningful, self-defining job. In simpler terms, Hegel believed that such a society could be realized within the framework of existing social institutions, and that reform, rather than revolution, was the order of the day. Our contemporaries are under no delusions that labor in the present day uniformly resembles Hegel’s idea of what labor would look like in the rational social order (although, I would submit, Hegel’s account is not entirely off the mark even as a description of our current situation). I think, rather, that something closer to Hegel’s true philosophy is motivating our generation’s refusal to give up on labor: the sense that even if “work” in its current form often consists of empty drudgery or shameless egoism, a society in which work really did possess the significance that Hegel imagined might just be achievable, waiting in the wings of historical possibility.
A string of recent articles in publications like The Baffler and The Atlantic seem to have taken up the project that Keynes began in “Economic Possibilities for our Grandchildren.” These pieces either assume that human beings have the ability to overcome the economic problem and insist that we abolish work as quickly as possible (see James Livingston’s “Why Work?”), or they assume that the advent of a post-work society is likelier than not and try to envision what such a society might look like (see Derek Thompson’s “A World Without Work”). I’m broadly sympathetic to the motivations of these authors, Keynes included; if technological progress has truly placed human beings on the brink of overcoming the economic problem, then debating the possible implications of such a transition is clearly a matter of great importance. Furthermore, I share the optimism that a “post-work” society could be a world of greater equality and human flourishing — a world, that is, in which each individual has the opportunity to freely develop her capacities and engage in truly self-determined and self-affirming projects.
However, as the bulk of this essay has surely made clear, I also think that anyone who views a “world without work” as the inevitable result of technological progress, or who casually dismisses the power of the forces that stand opposed to the abolition of work, is essentially misguided. The amount of time that people devote to work appears to be unrelated to the state of technological development. Our economies have been endlessly creative in providing people with new (and increasingly superfluous) opportunities to labor, thereby replacing the jobs lost to technological innovation. Merely possessing the technological capability to transcend scarcity and toil does not, therefore, ensure that human beings will actually succeed in doing so. Before Keynes’s vision of an automated leisure society can become a reality — and, perhaps more importantly, a reality worth living in — we must both reform an economic structure that severely hinders efforts to shorten the working day and find new institutions or practices through which human beings can constitute their subjective identities and acquire a sense of self-esteem. These concerns, rather than musings about productivity rates and the pace of technological advancement, should occupy all future programs for the realization of a post-work age.
* Again, I have no quarrel with economists who claim that technological progress will not necessarily result in widespread unemployment. It is entirely possible that our economy will continue to successfully conjure up new jobs to replace those lost to automation. My concern, rather, is with the quality of these new jobs, the seriousness of the “needs” served by these jobs, and what the widespread existence of such “bullshit” jobs says about the necessity of the 40-hour work week.
† The idea that work was something dignified, rather than a necessary evil to be avoided whenever possible, was a novel development in Western intellectual history. The Greeks generally viewed manual labor as a degrading activity appropriate only for slaves, while early Christendom held a decidedly ambivalent attitude towards manual labor (which, after all, was humanity’s punishment for Adam and Eve’s disobedience in the Garden). An influential explanation for the shift in attitudes towards labor was formulated by Max Weber in his The Protestant Ethic and the Spirit of Capitalism. Broadly, Weber argues that aspects of Protestant theology, such as the Lutheran notion of the calling and the Calvinist doctrine of predestination, had the unintended effect of endowing labor in the service of capital accumulation with religious significance and eventually a more general, secular ethical value.
‡ My interpretation of Hegel is deeply indebted to Frederick Neuhouser’s Actualizing Freedom: Foundations of Hegel’s Social Theory (see especially ch. 3), as well as his lectures on Hegel’s social thought.