This is a lightly edited excerpt from Dread: Facing Futureless Futures (Wiley, 2021) by David Theo Goldberg.
Tracking-capitalism: The political economy of dread
In January 2014, rapper Jay-Z played a gig in Toronto. By geo-tracking their smartphones a marketing and analytics company was able to determine that roughly 13 000 fans attended the show at the Air Canada Arena, how long they spent there, and most notably where they had been both before and after. Across the 2010s, a significant shift was taking hold of social life.
Digital technology, as I have demonstrated in the preceding chapter, has dramatically transformed our worlds. Work, play, learning and education, politics and recreation are all conducted differently in the wake of its development and almost universal adoption. We communicate, interact, consume, exercise, mind our health, bank, invest and travel not just with greater speed but in significantly contrasting ways to the past. Even writing, reading and cultural production more generally have altered as a consequence. Personal computing and mobile devices are within arms’ reach through waking and sleeping hours. Life no longer turns off.
As the Jay-Z event signaled, all this was upending social, economic, political and cultural life in completely unanticipated ways. The large-scale changes of this past decade have had less visible but equally telling implications for work profiles and positions, for who gets hired and for what functionality, as well as in assessing job performance. The digital has taken its disruptability as a definitive operating condition of its widespread application. It disrupts longstanding and taken-for-granted modes of production, work structures and practices. Social relations and cultural expressions too. Conceptions like “the born digital” and “digital native” reveal a contrasting time and culture before and after.
A major factor spurring contemporary dread, especially in industrialised states, is intensifying concern over techno-replacement and irrelevance. Significant job loss in the foreseeable future has become a driving anxiety as robotification ramps up. This has certainly helped to generate sometimes seismic political shifts and outbursts in countries like Britain, the United States and France. Political forces have emerged here and elsewhere in part promising to reinvent or return to a nationalist political status quo ante. The “return” is often to an imagined state of privileging “purity”, the reinvention a tacit acknowledgment of the fabricated social tissue underpinning the projection. Foundational to this political promise has been a commitment to ensure that jobs remain in the seemingly abandoned workplaces while sustaining the welfare structure that together anchored especially white working- and middle-class life prospects and lifestyles for much of the last century.
The fog of unease descending recently, manifesting in dread as it spreads across life, is knotted with the creeping sense that this bubble of work and lifeline has burst. The rapid development of AI and machine learning that drives expanding robotification is projected to result in massive job disappearance over the next 25 years. The shop floor increasingly has become an awesome exhibit of computerised robots and machines. Predictions range anywhere from 30 to 60% job loss generally, most notably in manufacturing and repetitive rote work. Carl Frey and Michael Osborne (2013) estimate that approaching half of American jobs “are at risk”, a McKinsey Report that as many as a third will be lost. We are already witnessing massive technicisation in automobile, aircraft, computer and other high-tech manufacturing and assembly. Given the increasingly integrated nature of the global economy and its supply, distribution and consumption chains, robotification in one national site will impact work availability elsewhere, often in unpredictable ways. The disruptions caused by Covid-19 have accelerated these dynamics in real time.
One more sober analysis projects that in the next decade a little less than 10% of manufacturing jobs – 20 million worldwide – will be automated. Nearly three-quarters will be in China, rapidly leading the world in replacing manufacturing workers with robots. One in three of the world’s robots are being introduced into Chinese factories. In the past 15 years, every robot introduced has displaced more than a worker and a half, a process fast accelerating. The Oxford Economics study predicts that throughout the 2020s, China will lose more than 12 million jobs to automation, the United States 1.5 million, the European Union 2 million, and South Korea 800 000. Cities with more diversified economies will tend to be less impacted. In Britain, for example, manufacturing is located overwhelmingly in smaller cities, towns and rural areas while London offers a more varied range of work. Disruption will dramatically affect those areas less well prepared for the transition. The political shift is well under way, with some impacted working-class groups gravitating to more conservative – and, by implication, more conventionally conserving – political commitments.
In the past five years, robots have become less expensive, now undercutting the manufacturing costs of human labour. Microchips process more powerfully, batteries have longer life spans and networked operating systems have become significantly “smarter”. Production line breakdowns are easier to track and fix, output has become faster, costs have reduced. These trends are translating into automation developments beyond manufacturing, from warehouse distribution to service and some office tasks. The rate and range of job losses are bound to increase too. Amazon’s warehouses deploy small robots to pack its goods for shipping. These robots take a quarter or less of the time human packers require.
Robots have been designed to do the heavy lifting and repetitive functions in manufacturing and commerce more generally. They function faster and operate without incurring injuries. When they break down, they can be replaced by the next in line without worker compensation costs. As robots grow smarter, they will be able to take on more and more rote functions in offices too, such as filing, perhaps basic bookkeeping, composition, completing standardised regulatory forms, sending and tracking form correspondence, and the like.
In manufacturing, those most likely to be impacted by job loss, at least for the foreseeable future, are less well-off. Blacks pretty much everywhere and Latinos in the United States are significantly more likely than whites and Asians to occupy those positions affected by robotification. There is a long history of identifying technological innovation and development with whites, as well as their benefits. Digital technology is little different, especially if you include those historically designated “honorary whites” or “model minorities” (most notably Asians). The top lines of work to be impacted by automation include manufacturing production, office and administrative support, farming, fishing and forestry, transportation and moving. Those least likely range over domains requiring more agile reasoning and nuanced, intricate judgement, including business and financial operations, education and training (the likes of Udacity, Coursera and Khan Academy notwithstanding), as well as architecture and engineering. Workers with lower levels of education will be far more readily impacted than those with graduate education. Automation clearly will exacerbate rather than narrow existing economic and, by extension, political and social inequalities.
It turns out that robotification is having unexpected impacts on how work is done too. Robots operate so much faster on work activities involving rote manual but also increasingly cognitive functionality. This is being used to pressure humans in the chain of production and delivery. Workers are being driven to speed up in order to enable the machines to fulfil their quota. Studies of turnaround and delivery commitments at Amazon warehouses reveal that company management requires much quickened work rates for human workers in the production chain so that the robots are able to function efficiently. Robotic time is prompting management to drive the human work pace. Amazon is tracking warehouse workers on their capacity to meet impossibly quickened turnaround times, as a consequence causing much faster human breakdown due to fatigue and injury. Failure even marginally to meet quotas is a cause for firing and replacing workers. Work under these conditions is less a site of self-providing dignity than the fuel for days of dread.
Similarly, employers are using an app on wearables like Fitbit and Apple Watch to track employee performance. The app monitors non-work-related time away from an employee’s desk or office, time on the phone, quality of sleep at night, and so on, all of which demonstrably impact quality of work. The app’s developers, researchers at Dartmouth College, claim 80% accuracy – which is to say, of the tracked relationality between more standardly assessed performance and the app’s assessment. Automation, both cases are suggesting, is reinforcing the inclination among corporate leaders, pressured by shareholders and venture funders, to robotify human functionality.
The steady decline of manufacturing positions in industrialised economies as the global economy has become deeply integrated and workplaces increasingly automated has resulted in concerns about social disruption more broadly. Where once reliable work could be counted on for career-long employment, especially by the white male working class, employment opportunities today might just evaporate tomorrow. In the United States, it is estimated that a person changes jobs every four years on average, most notably in the earlier part of adult work life. They shift professions approximately every decade. This perhaps signals less that work is disappearing than that it is changing in nature. Service work is increasingly replacing manufacturing work. The shift away from mall shopping to online consumption means that entry-level positions in retail stores are giving way to warehouse packing and distribution work. Robotification requires not only design and manufacture but also ongoing systems service and operations management. Someone has to make sure the machines run well, to service or replace them, to match their capabilities to the job required. The gig or “on demand” economy is offering a wide range of work, more flexible if less secure in terms of benefits and healthcare provision. The ground is rumbling beneath our feet.