The Poverty of Left Accelerationism: A review of Srnicek and Williams, “Inventing the Future”

by Jehu

Nick Srnicek and Alex Williams have written a book, Inventing the Future, proposing the complete automation of production and a reduction of hours of labor. The proposal is fascinating and stands head and shoulders above the gruel typically on offer on the Left.  Nevertheless it is poorly argued and in serious need of additional theoretical development.

The meat of the book can be found in chapter 6, where the authors discuss the Holy Grail of Left Politics, non-reformist reforms — reforms that, of themselves, have revolutionary implications, that force society to go beyond existing capitalist relations. To this end they propose four demands they believe are necessary, “to start building a platform for a post-work society.”

These demands are:

1. Full automation of production
2. The reduction of the working week
3. The provision of a basic income
4. The diminishment of the work ethic

I will spend some time reviewing it here.


What is preventing a fully automated future? Why isn’t wage labor going away, despite all predictions to the contrary?

The first question is raised by this passage from the book:

“Capitalism has long been synonymous with rapid changes in technology: driven by the imperative to accumulate, the means of production are continually transformed.”

The second question is raised by this passage:

“From the beginning of capitalism, workers have struggled against the imposition of fixed working hours, and the demand for shorter hours was a key component of the early labour movement.”

The authors argue both classes in capitalist society are trying to get rid of wage labor, each for their own reasons. The workers want more leisure, the capitalists want more profit. Between the two of them, wage labor appears doomed.

However, if these two assertions are true, why then do the authors still find it necessary to include demands for full automation and less work in their program? Why is there a need to demand things that appear to already have been unfolding for the whole of the 19th, 20th and 21st centuries?

In proposing a third demand for universal basic income, the authors seem to offer an answer of sorts. Complete automation of production and a general shortening of labor hours may be incompatible with the continuation of wage labor:

“These first two proposals equate to the reduction of labour demand through full automation, and the reduction of labour supply through the shortening of the working week. The combined outcome of these measures would be the liberation of a significant amount of free time without a reduction in economic output or a significant increase in unemployment. Yet this free time will be of little value if people continue struggling to make ends meet. As Paul Mattick puts it, ‘the leisure of the starving, or the needy, is no leisure at all but a relentless activity aimed at staying alive or improving their situation’. The underemployed, for instance, have plenty of free time but lack the means to enjoy it.”

The demands proposed by the authors thus rests on a contradiction. Srnicek and Williams insist automation and labor hours reduction would free up society from labor while increasing the production of material wealth. But then they insist the resultant free time created may have little value because people would be struggling to make ends meet. Automation does not necessarily lead to the emancipation of society from labor; it can just as easily lead to plain old-fashion unemployment.

The authors appear to conflate free time with unemployment, when they suggest free time might lead to a struggle to make ends meet. At the same time, they appear to argue against such a conflation, because they argue reducing labor hours and automation increases output without a significant increase in unemployment. If complete automation of production and a reduction of hours of labor increases output without a significant increase in unemployment, why will there be a struggle to make ends meet?

Somewhere in making their argument for complete automation of production Srnicek and Williams seem to have dropped a stitch.

I will return to this point later, but first…

Who wants complete automation of production?

Srnicek and Williams state their case for complete automation of production this way:

“Our first demand is for a fully automated economy. Using the latest technological developments, such an economy would aim to liberate humanity from the drudgery of work while simultaneously producing increasing amounts of wealth.”

To whom is this demand for a fully automated economy being made? And, on whose behalf?

If the replacement of human labor by automation is going to become a political project of the left, this immediately presents them with a huge political problem. The Left fancies itself the political representative of the very people who will end up being automated out of a job. The Left has to sell complete automation of production to the very people who will be displaced in production by machines. Mind you, this is a serious political obstacle: wage workers are not folks for whom labor is an idle fancy. The people to whom complete automation of production has to be sold live only so long as they can sell their labor power to capital. Automation directly threatens the sale of labor power. The Left will have to convince wage workers that elimination of their jobs is a benefit for them, despite the obvious fact they will be summarily replaced by robots on the production line, at the checkout counter, in the office cubicle, classroom and eventually even in the doctor’s office and care work.

The other side of this problem are the owners of capital, who are solely motivated by profit. At present, automation is introduced not as a means to liberate humanity from the drudgery of work or increase the production of material wealth, but to accumulate wealth in a specific form: capitalist profits. Srnicek and Williams want to push automation beyond this limited end to actually emancipate society from work and poverty; which is to say, they want to go beyond the limits imposed on automation by the owners of capital.

The demand for complete automation of production means the Left will find itself opposing both capital and wage labor. They will be forced to confront wage labor because they want to automate the wage worker’s labor no matter its impact on her job. And they will be forced to confront capital, because they want complete automation of the production of material wealth no matter its impact on profits.

Moreover, we still have not figured out to whom this demand is to be made.

Do we make our demand to the worker, who is only concerned to sell her labor power? Do we make it to the capitalist, who is only concerned about his profits? Do we make it to the existing state that only functions as the ideal representative of the capital-wage labor relation? There does not appear to be any class within existing society to either champion this demand as its own aim or realize it.

The missing middle

Nevertheless, I am not against this aim; in fact I completely agree with it. My sole problem with this extreme program is that Srnicek and Williams offer us no credible path for its realization within the limits of present society.

This is essentially the criticism the folks over at De-Arrest Editorial Services (DES) made against Srnicek and Williams earlier this year. DES argue the book lacks punch because the text fails “to provide any indicator of the middle-term conflict that would be necessitated by a movement for the abolition of wage-labour on a global scale.”

However, the folks at DES are wrong to blame Srnicek and Williams for this. If Srnicek and Williams fail to offer a realistic path to realize their demands, it may just be because no social class actually wants to fully automate production of material wealth.

This is a second contradiction in Srnicek and Williams argument. Society seems to be moving inexorably toward complete automation, despite the fact no class in society wants complete automation. The absence of a middle-term conflict noted by DES simply expresses the fact that the abolition of wage-labour on a global scale is at nowhere the conscious starting point of any class in society. Instead, it is the inevitable product of a process over which neither class has any control.

The difficulty Srnicek and Williams ran into is thus the problem critical theory spectacularly failed to solve:

If there is no revolutionary subject, who the fuck is abolishing wage labor even as we speak?

Although the complete automation of production has no protagonist in society, the evidence provided by Srnicek and Williams suggests it is increasingly difficult for the wage worker to sell her labor power:

“As we have seen, there is a growing population of people that are situated outside formal, waged work, making do with minimal welfare benefits, informal subsistence work, or by illegal means. In all cases, the lives of these people are characterised by poverty, precarity and insecurity. Increasingly, there are simply not enough jobs to employ everyone. As the hegemonic order predicated upon decent and stable jobs breaks down, social control is likely to revert to increasingly coercive measures: harsher workfare, heightened antagonisms over immigration, stricter controls on the movement of peoples, and mass incarceration for those who resist being cast aside. This is the crisis of work facing neoliberalism and the surplus populations who make up most of the world’s labour force.”

Although we can identify no class in society demanding the abolition of wage labor, wage labor is facing a crisis as an increasing number of people are unable to sell their labor power.

The great unsolved crime of the epoch

Here is the ultimate crime of the epoch with no clues, no fingerprints, no weapon. Someone is killing off wage labor, but all the usual suspects seem to have airtight alibis. The wage worker is absolutely dependent on selling her labor power to survive. The capitalist is equally dependent on the sale of labor power for his profits. And the state has worked tirelessly for eight decades now to expand ‘economic growth’, i.e., wage labor.

Essentially, the whole of critical theory since World War II has been trying to solve this Whodunit without success. As both DES and the authors suggest, capital has been trying to abolish wage labor since its inception and seems well on its way to accomplishing this, but never once in this entire epoch has the abolition of wage labor been the conscious aim of either class in the capitalist relation. As a practical matter the opposite has been true, at least since the 1930s: both classes have aimed to expand wage labor, not abolish it.

Since neither class in the capitalist relation wants to abolish wage labor, it is no surprise that the state, the ideal representative of the capitalist relationship, has aimed for the constant expansion of wage labor, i.e., economic growth for the past 80 years.

And here we come to the third, and by far the most popular demand advanced by Srnicek and Williams in their book, universal basic income.

Like most supporters of universal basic income, Srnicek and Williams appear to believe that poverty, precarity and insecurity can be overcome by as simple a measure as handing out a pile of worthless currency to everyone in society. However, they note this approach to complete automation of production is fraught with serious complications — any program for a universal basic income requires specific legislation from the state and is subject to competing political forces.

Thus they observe:

“The demand for a UBI, however, is subject to competing hegemonic forces. It is just as open to being mobilised for a libertarian dystopia as for a post-work society – an ambiguity that has led many to mistakenly conflate the two poles.”

The argument in favor of a UBI is that it appears to directly attack poverty, precarity and insecurity by handing out a wad of cash to all members of society. Therefore, it appears to overcome the problem identified by Mattick that free time can just as easily appear in the form of starvation and unemployment.

The obvious argument against UBI is that the legislation itself is not given; it must be the product of a political struggle between the two classes. Yet the history of the 20th century class struggle has demonstrated beyond any possibility of refutation that the working class has had its ass handed to it every time it has directly confronted its antagonist, the bourgeoisie. If the 20th century is remarkable for anything, it is the conclusion the proletariat cannot afford any half-measures that do not directly attack wage labor frontally.

Despite this, Srnicek and Williams seem to think UBI can work if we can just observe three caveats:

1. it must provide a sufficient amount of income to live on;
2. it must be provided to everyone unconditionally; and,
3. it cannot be allowed to replace the present welfare state.

In the face of 20th century history and the countless defeats of the proletariat at the hands of the bourgeoisie, who can credibly maintain the working class can define and defend a universal basic income program over the strenuous opposition of the capitalists? Moreover who credibly maintains the state will suddenly change its 40 year neoliberal pattern of siding with the capitalists against the working class?

If complete automation is so good, why do you need UBI?

My difficulty in accepting Srnicek and Williams argument in favor of UBI, however, is not primarily focused on the obvious problems of the working class defining and defending a UBI scheme on their own terms; rather, it is instead concerned with why they feel it necessary to offer it in the first place.

I suspect Srnicek and Williams offer a demand for UBI because they secretly believe the complete automation of production will leave the working class in worse shape than it is now. My bet is that the authors secretly accept poverty, precarity and insecurity will worsen if the production of material wealth is completely automated. They may not say this openly and honestly to their readers, but the hints can be easily inferred from their own words.

Let’s look again at that revealing intro to their demand for UBI first cited above:

“These first two proposals equate to the reduction of labour demand through full automation, and the reduction of labour supply through the shortening of the working week. The combined outcome of these measures would be the liberation of a significant amount of free time without a reduction in economic output or a significant increase in unemployment. Yet this free time will be of little value if people continue struggling to make ends meet. As Paul Mattick puts it, ‘the leisure of the starving, or the needy, is no leisure at all but a relentless activity aimed at staying alive or improving their situation’.

Basically, in this passage Srnicek and Williams equate the complete automation of production with unemployment in this passage. And this is for good reason: Under the existing mode of production automation always advances at the expense of the worker. The aim of automation is to replace living labor with machines for the sole purpose of increasing profits. Srnicek and Williams have gotten into their heads the automation of material production necessarily has this characteristic.

In the capitalist mode of production, automation leads to unemployment because automation itself, as the authors note, is driven by accumulation, by profit. By contrast, Srnicek and Williams have stated that the aim of their demand is to push automation of material production “beyond the acceptable parameters of capitalist social relations”, to push automation beyond the limits imposed on it by the aim of capitalist production for profit.

Thus we have a third contradiction buried in the book: Srnicek and Williams propose we advance to complete automation of material production even if this abolishes production for profit. This, they assure us, will create increasing amounts of material wealth; yet, at the same time, they suggest it must result in poverty, precarity and insecurity, “the leisure of the starving”, unless we also have universal basic income.

How do we reconcile this disparity?

The argument that automation produces more wealth flatly contradicts the argument that people will continue struggling to make ends meet. If automation creates more wealth, but people continue to struggle to make ends meet, where will the increased wealth be going?

To paraphrase the criticism raised by the folks at DES, Srnicek and Williams have neglected to show us a middle-term conflict that explains how increased production of material wealth must result in a struggle to make ends meet. This explanation is absolutely necessary, because it is the sole reason they raise the demand for a universal basic income. Unless they can explain why complete automation of production necessarily leads to poverty, the demand for a UBI has no purpose at all.

And if it has no purpose at all, why would we fight for it, since a UBI program necessarily requires a state that we all know is, and has been for at least 40 years, completely subordinated to capital.