This article was originally published in the 'Get to work' issue of MUSEUM magazine.
Imagine a train hurtling towards you. It’s a hundred metres away, close enough to hear the hum of the engine and the rattle of the tracks. You’ve been told what you need to do. You’ve been given four or five different ways you can avoid being struck by the mammoth locomotive—but you don’t do any of them. Instead, you conceptualise the millions of possible realities in which being run down by a fast-moving train might not be so bad. You consider how you might pick up the shattered pieces of your body after
the train obliterates you, rearranging it into something vaguely resembling what it looked like before. You know it will never look or function as it once did. But you don’t even consider stepping out of the way
This is where the world stands with the looming threat of automation. We are facing, within a decade or two, possibly sooner, the elimination of an entire stratum of low-income jobs. A joint study by Oxford University and the Oxford Martin School found, through analysis of labour markets in the United States, that up to 47% of jobs in America are susceptible to computerisation—and that it will be those with less skills and less formal education who will feel the brunt of this change earliest and hardest.
We can see this happening all around us. Real wages in western markets have more or less remained stagnant over the course of the past four decades. Through a system of broad deregulation and liberalisation of global economic administration, and the increasingly attractive benefits of automation, the share of income in the developed world has increasingly been captured by capital, leaving labour to pick up the scraps.
This is not a new or surprising development. Economist John Maynard Keynes frequently cited his belief that the rate of technological advancement would at some point outrun the pace at which we can find new uses for labour. Nonetheless, Keynes was optimistic. He believed, as many classical economists tend to argue now, that a swell in productivity would automatically translate into improved incomes, which would lead to increased demand for products and services. Behold: more jobs.
Keynes thought that this balancing act of automation and increased demand would produce a sort of utopia, wherein individuals worked only 15 hours per week, with technology taking care of the rest. There was reason to believe this would be the case—after all, society had weathered revolutions in labour-saving technology before, alongside a reduction in the working week. Everything from the invention of the loom to the industrial revolution itself had represented a replacement of human labour with vastly more efficient machinery.
To put it lightly, it didn’t quite eventuate that way. The logic of capitalism and the world’s baffling fixation on the inherent spiritual value of labour have used technology to force workers to work more. Anthropologist David Graeber, a leading figure and academic within the Occupy Wall Street argues that most of the world has been sequestered into “bullshit jobs”—administrative busywork largely conducted in front of a computer screen.
There is certainly a sense that there are a great many service jobs which have no real reason for existing. Liberals might argue that the very existence of these salaried roles justifies the labour applied to them; the market has demanded them. But there’s nevertheless a psychic malaise surrounding workers in nebulous fields from human resources to administration—the pervading sense that their roles produce nothing of real value, that the world might continue easily without them. There’s a good reason why unions have difficulty penetrating and organising these sectors: if upper management dispensed with these jobs entirely, what would be lost? What bargaining chips do workers possess?
Meanwhile, the app economy has created new spaces for traditionally structured (and protected) service jobs to be provided on a peer-to-peer basis. The sharing economy fractures traditional means of social organisation for radically individual modes of labour, through which workers can perform tasks at the whim of an algorithm. ‘Disruption’ has become a codeword for the development of efficient business models, based on the total obliteration of hard-won labour protections. Those low-income jobs that aren’t destroyed by automation risk will be performed by individuals stripped of their rights—compelled to work under a perverse system of digital feudalism, where they are granted access to corporate platforms and marketplaces on a contingent and unstable basis. This is access that can be denied at any time, for any reason, without the safeguards we have come to expect from employment relationships.
It’s important to recognise that market logic will take no prisoners when it comes to automation. If there are clear efficiencies in replacing workers with machines or with code, then employers will do so. Individual cases of charitable employers who invest in retraining programs or governments who do the same on a systemic level will not slow the hollowing out of the lower echelons of labour.
There are 3.5 million truck drivers in the United States. Approximately 7.8 million people are employed in roles ancillary to driving a truck. As I write, companies like Uber and Google are racing to automate the time-consuming, labour-intensive and dangerous profession of longhaul trucking; where staff cross borders bleary-eyed, strung out too far. Automation would vastly reduce costs for many companies that rely on the transportation of tangible products—why wouldn’t they automate? In the US alone, around 4,000 fatalities a year are caused by truck accidents; surely it is an imperative to reduce the harm.
It’s a flawlessly simple equation: if it is profitable to automate, companies will do it. There will be no concern for what this profoundly dislocating change will augur for the communities it devastates. A firm replacing its factory workers with machines will not consider how their change will fracture the political landscape, nor will they imagine which consumers will be left with sufficient disposable income to buy products once every other organisation automates. It will be a perfect storm.
The shock, I suspect, will be unimaginable. Consider the situation we find ourselves in today, where the pressure cooker of globalisation has fomented a seething populist rage—one that finds expressions both on the left and the right. The economic displacement, spiritual fragmentation and pure, white-hot rage that gave us Donald Trump pales in comparison to what might follow. If a financial crisis and evaporating coal jobs were enough to shatter the blue wall and turn Michigan and Pennsylvania toward economic nationalism, what will happen when an entire class of workers has their livelihood pulled out from underneath them? How will we even begin to retrain a significant chunk of western workforces when so many welfare states are crumbling? What sort of unhinged political movements will emerge, feeding vampirically on fear? Automation is already leaving its mark on the world, but we can scarcely comprehend where it will lead. The question turns to how we ought to address this situation.
This is a problem that both governments and corporations seem unwilling or unable to consider in real, concrete ways. It isn’t necessarily because they don’t recognise the looming problem—the neoliberal system we have immersed ourselves in as a planet for the past three decades, predicated on growth and constant expansion, makes it near-impossible to change. Our liberal institutions, entrenched in notions of freedom, stop us from moving out of the way of the train.
The most widely proposed and most radical solution to the problems inaugurated by automation—short of actual socialist revolution—is the universal basic income. Proponents on both the left and right concede that there is little point in generating busywork for the majority of society out of some misplaced attachment to the protestant ethic, and suggest there ought to be a basic income provided to a country’s citizenry. The logic is that those who have no desire to work no longer need to do so, and can learn to survive on whatever subsistence is provided to them by the state.
There are good arguments against a universal basic income. It seems that within the current structure of the democratic nation-state it would be difficult to guarantee its continuation across generations and successive governments, growing at a rate commensurate to consumer price indexes and inflation. It’s near impossible to conceive of the world’s rapidly ossifying liberal democracies achieving any sort of lasting economic change in that form. Even the postwar social welfare consensus would be unimaginable in nearly any nation you could name.
But whether or not it would be achievable, it’s not something that has been seriously considered at a legislative level almost anywhere. Australia’s government barely acknowledges that the future of work is precarious, let alone considers any means by which it could be addressed. The United States is incapable of sustaining any kind of new federal project, let alone one that reimagines the very nature of work. Only in the smaller Scandinavian and central European economies, already accustomed to extensive welfare states and public ownership of production, are concepts like universal basic income contemplated on a trial basis.
The systems which are naturally emerging to fill the widening gaps left by collapsing social democratic projects and increasingly difficult job markets are not satisfactory. The ride-sharing economy popularised by Uber—which bases its pitch to potential drivers on supplementing a primary income—is facing a series of major crises within a decade of its existence. Much of this can be put down to Uber’s remarkably immoral conduct, which pushed the boundaries of acceptable corporate behaviour so aggressively that most reasonable consumers might eventually be tempted to reject it. But it’s also clear that though users love the convenience of the product, it cannot provide a model for how labour will work in the era of precarious employment.
And Uber has invested heavily in self-driving car technology. Even those systems which emerge to create new, practical jobs are themselves angled along a continuum towards automation. The company is the perfect example of a system eating itself—it exists nominally to connect riders to drivers to the benefit of both, but ultimately cannot sustainably exist unless the drivers (the labour) are robotised. Such was Uber’s desperation to move quickly to automated vehicles—less than a decade after their so-called ‘disruptive’ model emerged in the first place—that they repeatedly broke laws in numerous American jurisdictions by testing their driverless cars. (In late 2016, California were adamant the practice was illegal, though Texas governor Greg Abbott signed a bill in June this year, declaring companies are welcome to try self-driving cars on the state’s roads and highways with no driver present behind the wheel.)
As the rate of technological progress accelerates, it is easy to imagine this happening along a wide variety of ‘replacement’ sectors as quickly as they emerge. The economy will generate jobs and entire sectors of employment which are shambling zombies from the outset. Again, our liberal economic systems seem fundamentally incapable of redressing this problem. This is not a call-to-arms, nor is it a grave warning to policymakers of the consequences of inaction. It is a simple statement of fact: we are standing astride a railroad track, and a train is hurtling towards us. Whether or not we are capable of jumping out of the way increasingly seems like it isn’t up to us.