Is automation really such a threat?

Why anxiety over technology is simply misplaced anxiety over capitalism

In the Western economies, public interest in technology has in recent years been coloured by a sort of techno-pessimism, or “techlash”. Distrust of the technology giants has gotten to the point where even The Economist is advising them on how to redeem themselves in the face of public opinion. After attacks on hackerspaces in France, The Guardian asked if 2018 would be the year of the neo-luddite. In September 2017, The Atlantic wondered if “smartphones destroyed a generation?” And it seems like hardly a day goes by without another panicked headline on the dangers of looming automation.

Automation—the replacement of human by machine labour—has particularly captured our imagination lately. Images of mass technological unemployment are now common in the press and among policy makers. The cover of the much ridiculed pamphlet by Conservative MP Alan Mak illustrated the point perfectly, depicting what it described as “LABOUR’S DYSTOPIAN FUTURE” in which humanoid robots are queuing for work in a factory while being attacked by (presumably) unemployed humans.

On its own, however, automation is not necessarily a threat. It is the political and economic system in which it is implemented that should worry us. Ambitious left-wing policies could cushion any potential downsides of automation while could providing the groundwork for a fully automated, post-work society. On the other hand, under the current neoliberal model, increasing automation would most likely be used to strengthen the power of the ruling class, at the cost of increased precarity and worsened living standards for the rest.

Bullshit jobs

Automation has, however, been on our minds for quite a while now, despite recent coverage treating it as a new problem.Take, for example, this prediction by Keynes from 1930:

For many ages to come the old Adam will be so strong in us that everybody will need to do some work if he is to be contented. We shall do more things for ourselves than is usual with the rich to-day, only too glad to have small duties and tasks and routines. But beyond this, we shall endeavour to spread the bread thin on the butter – to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!

Keynes’ prediction of fifteen-hour weeks has not aged well, but his quote shows how the fear of automation—and its obverse, the promise of a post-work society—has been around for quite some time. Previous waves of automation have hit a range of occupations, including textile workers, bank tellers, car-builders and dockworkers, and each wave has been accompanied by fears of permanent technological unemployment. On the other hand, these waves also show the promise of a society liberated from work.

So far, even while capitalism has destroyed certain types of work, it has done an excellent job at creating others. Out of the ashes of dying sectors and occupations, entirely new categories of work are created, some of them utterly unimaginable to previous generations. In a way, capitalism might even be too good at this—one often wonders how many telemarketers or social media managers are really necessary from the standpoint of societal need. As David Graeber explains in On the Phenomenon of Bullshit Jobs,

But rather than allowing a massive reduction of working hours to free the world’s population to pursue their own projects, pleasures, visions, and ideas, we have seen the ballooning not even so much of the “service” sector as of the administrative sector, up to and including the creation of whole new industries like financial services or telemarketing, or the unprecedented expansion of sectors like corporate law, academic and health administration, human resources, and public relations.

The current wave of automation might be different due to advances in computational technology—specifically, the prospect of artificial intelligence—which could allow it to destroy jobs faster than capitalism can create them. Still, it is far from certain that there will be a massive aggregate decline in the total amount of jobs to be found in society.

A human-driven process?

One major obstacle to developing the political will to successfully dealing with automation is the way it’s often presented as if it’s a natural process. This in turn inculcates a sort of helplessness among the public—there’s nothing we can do about it, so why even bother?—while simultaneously acting as a smokescreen to distract us from its role in maintaining the power of capital. After all, as Grace Blakeley argues in Robots Aren’t Coming for Our Jobs – Capitalists Are,

Talking about ‘globalisation’ allows us to construct the idea of ‘globalisation’ as an inevitable, impersonal trend, driven by the agentless forces of history. The nominalisation completely obscures the fact that ‘to globalise’ is a verb – the same goes for ‘to automate’. ‘Automation’ doesn’t just happen – tasks are automated by people.

Ultimately, automation is a human-driven process, and it is human beings who decide whether or not to implement it in any particular environment, weighing the choice between investing in machines versus employing (lower-cost) humans to do the work. The motive here, of course, is profit. In places where a unionised workforce has managed extract enough concessions to threaten profits, automation will seem like a particularly enticing prospect for capital—what Beverly J. Silver calls a “technological fix” to crises of profitability.

Such a fix will, of course, interact with labour market dynamics. In the longer term, if enough jobs are automated away, that could increase the supply of workers for lower-paid jobs. In the absence of either a legal framework or strong organised labour movement to protect those workers, the result could be downward pressure on wages in those jobs. This would make it cheaper to hire human beings for those jobs than to invest in technology for automating them away—at least in the short run—which could lessen management’s desire for automation.

As a result, we should not think of automation as a purely technological process. A more realistic understanding of automation would see it as an interaction between technology and economics, typically for the purpose of augmenting class power and increasing profitability.

Policy alternatives

Effective policy alternatives for dealing with automation are plenty, but they generally err on the left side of the political spectrum. Traditional left-wing policies such as a good welfare state, powerful trade unions and a national retraining system would be able to cushion most of its negative effects. The New York Times recently explained that fears of automation are lower in Sweden than in the rest of the world, largely because of their strong welfare state. In other words, what we need to fight the Davos man’s biggest headache—mass technological unemployment—is what he spent the last few decades cutting.

If this new wave of automation does prove to cause a secular decline in employment, and capitalism thus does not manage to invent new work at the same pace as automation, a stronger renewal of the welfare state would most likely be needed. This could mean strengthening existing provisions while also implementing a variety of newer proposals.

This could be things like a shorter work week, a historical demand of the labour movement which also shows up in Keynes’ dream of a 15-hour working week. The idea behind an universal basic income would also be useful, particularly because it questions the current transformation of the welfare state into a disciplinary mechanism for the poor. Nevertheless, it has significant problems regarding efficiency and associated costs.

Another proposal that particularly struck my attention was the “Universal Basic Services” idea proposed in a UCL report:

The UK should provide citizens with free housing, food, transport and IT to counter the threat of worsening inequality and job insecurity posed by technological advances, a report launched by the Institute for Global Prosperity recommends.

Building on the ethos that saw the establishment of the NHS and public education – that essential services should be free at the point of need – the plan would “raise the floor” of basic services all citizens can expect, providing better protection for workers in the face of rapid advances in technology and automation.

Of course, the idea is limited in scope, and its monetary impact might be low for the entire population. Nevertheless, it could serve to de-commodify certain basic services for the population and could lead to public control over key infrastructural goods like Internet provision. It would also provide a stronger safety net for the unemployed, which would in turn bolster the bargaining position for workers as a whole.


What scares me the most about automation is not the lack of policy alternatives to deal with this new wave of automation—there are plenty of options, after all. What’s truly frightening is the conservatism of our elites.

We are in the twilight of neoliberalism, and yet elites have been actively resisting the possibility of change, even in the wake of the 2008 crisis. This process has historical parallels: when the Roosevelt administration wanted to implement the New Deal during the 1930s, they were resisted tooth and nail by business elites, even though the New Deal was a relatively limited implementation of a welfare state. It eventually took a war that killed 60 million and the threat of the Soviet Union to convince North American business elites that some state intervention might be appropriate.

A modern-day equivalent of this might be Macron in France. He does set out a strategy for dealing with artificial intelligence, coupled with a planned €1.5 billion investment in the field. He also talks the standard establishment policy talk on AI: attention to ethics, retraining for replaced workers and a disrupt-or-be-disrupted mentality. Yet at the same time, he attacks trade unions and social movements, privatises education and practices standard austerity politics, thereby destroying potential sources of resilience against automation.

Consequently, the problem is not technology, as the answers to its negative spillovers are relatively straightforward. The problem is that our elites refuse to step away from their failed policies, or even try to take advantage of those negative spillovers. It will probably require a powerful labour movement to drag them, kicking and screaming, into a solution that works.

Beyond capitalism

In order to deal with the negative effects of automation in a humane way, we will need to go beyond neoliberalism, and beyond the establishment that has spearheaded it for the last few decades. As socialists, however, we should also look beyond that. Under capitalism, automation occupies a strange position: capitalism automates human labour, yet also forces humans to sell their labour for an income. Something that requires us to work less to create the same collective wealth thereby becomes a threat as people lose their jobs and, as a result, their ability to survive. Perhaps it’s time to resolve that contradiction by transcending it, through embracing a new model of organising the economy that isn’t centred around the wage-labour of the working class.

Under such a model, our approach to automation could be much more than attempting to mitigate the damage it causes. With the levers of the economy under collective control, automation could push us into a post-work, or at least partly post-work, society. Every advance in automation would then mean more time for other, socially useful activities, and even a collective lowering of working hours. Maybe we would even see Keynes’ vision of a three-hour day during our lifetimes.