If I were to summarize Daniel Susskind’s new book “A World Without Work” in one sentence, it would be “The world without work is behind the corner but we shouldn’t worry too much about it.” The second part (about no worries) is certainly reassuring in this particular time as COVID-19 has accelerated transition towards a world with more automation and digitalization and less direct human involvement and interaction. This trend, while necessary under the current circumstances, has also heightened the fears and angst about our future where humans may be rendered useless by more capable machines.
Many universities around the world have transitioned completely towards online education, with neither professors nor students physically visiting the campus. Hundreds of colleges in the US have reversed or altered their reopening plans in the past several weeks after taking stock of COVID-19 testing availability, student and faculty safety concerns, state regulations and the worsening public health crisis. Whereas the education process is by definition human centric and neither professors (at least for the time being) nor students can be replaced with robots, the complementary jobs on the campus (catering, building maintenance, security, administration, etc.) have shrunk. Moreover, since the virtual lecture hall is virtually unlimited and can take as many people as you can only imagine, the demand for lecturers is likely to go down too. Similarly, the use of digital finance has soared throughout the world as did online shopping. In the US, which has traditionally lagged other countries in digital adoption, 35% of customers have increased their online banking usage during COVID-19. VISA saw more than 13 million customers in Latin America make their first-ever online transaction in the March quarter of this year. Online orders in retail industry in Europe peaked in April-May to 200% in the UK, 187% in Italy and 85% in Germany (in comparison to the previous year). Telemedicine has received an immense boost, with online consultations and diagnostics booming. Government services are increasingly delivered online (of course, some countries, such as Estonia, transitioned to comprehensive e-governance some time ago but this year, for instance, five utility providers in the Ukrainian city of Kharkiv have adopted e-document management systems for their customers). These days it is more common to confer online than in person (even at the highest level – think about the recent EU consultations over the situation in Belarus), with the consequential impact on all those jobs that physical operation of government offices maintains. The latest sensational clip on the social media allegedly shows the king of Bahrain accompanied by a huge robot bodyguard fully equipped with 360-degree cameras, anti-blast armor, inbuilt pistols, a machine gun, and a rocket launcher. Whatever else, one thing is sure: such a bodyguard cannot contract COVID-19.
So, the trend is there in front of our eyes, and Susskind is trying to make sense out of it building up on a vast existing literature on this topic (I would like to mention here Martin Ford’s The Lights in the Tunnel published over a decade ago and his more recent The Rise of the Robots (2016), which explored many of the same issues). The two issues that Susskind explores in the first part of his book concern the scope and direction of automation and its impact on labor substitution as a result of interaction of two opposing trends, complementing effect and substitution effect.
Susskind analyzes the common skepticism about the extent and depth of automation. Surely, there are occupations that require human qualities, such as ingenuity, originality, creativity. There are professions that require empathy, compassion, and care that only humans can display. So, the common reasoning goes, only routine and dull tasks can be automated freeing people for more challenging and interesting jobs. Here Susskind, quit rightly, introduces a distinction between the ability of machines to think and act in a human way (which is not there – as yet?) and their ability to perform certain activities. In a nutshell, to deliver a diagnosis or produce a legal document (apparently complex activities which machines have successfully mastered), the machines don’t have to be human like or think like us or think at all, in the sense we understand it. Hence, the encroachment of machines into areas traditionally considered reserved for humans will continue, at an accelerated speed, given their ever-increasing computing power and sophistication. Of course, there may be ethical and other limits to how much we want this encroachment to happen even if machines can carry out certain tasks more efficiently than humans. We may feel uneasy about fully automated decisions on parole or bank credit or about fully automated care activities for elderly and infirm, let alone fully automated combat activities, with machines responsible for identification and destruction of targets.
As the “rise of the robots” continues unabated, the complementing effect of automation (its ability to create new jobs instead of the ones retired due to automation) will recede, and an increasing number of people will be driven out of work, resulting in an end of the Age of Labor. Sure, it will not happen at once and will not happen at the same time everywhere but it will, eventually – this is what Susskind argues. He also looks – helpfully but rather inconclusively – into the relationship between technology and inequality, a topic that has been featuring much more prominently in the past decade, once we discovered that the “end of history” is nowhere in sight. He notes that technology is often the main driver of inequality, both directly by increasing the pay of high-skilled workers, or encouraging firms to use more traditional capital relative to labor, and indirectly, by promoting globalization and other economic shifts. Susskind subscribes to the view that inequality is essentially a social phenomenon that can be regulated (at least to some extent) institutionally – a view that has been excellently substantiated with a wealth of historic factual data by Thomas Piketty in Capital and Ideology.
That is, this takeover of jobs by machines is bound to happen unless humans also become more effective. What I mean here is the human capacity enhancing brain-machine interfaces. As I write this, Elon Mask has just unveiled a major update of his biotech startup Neuralink and demonstrated a “working device” for the company’s hyped brain-machine interface. Neuralink’s technology, the founder claims, will “merge biological intelligence with machine intelligence” and perhaps help us survive the “AI apocalypse”. The efforts to enhance human capacities via a machine link remains controversial and the full implication of this trend is yet to be understood. There is probably no stronger critic of these attempts than the philosopher Slavoj Žižek in whose opinion this could not only radically change, if not destroy, our human nature but will also give politicians an opportunity to turn artificial intelligence into favor and enslave us. Such a policy would control people and force them into acts that they would not even know they really did not. This is something that contains totalitarian implications: “The main question for me is – and I have no answer – how such technology will affect our own perception of ourselves. We will still be perceived as free living beings or we will be regulated by the digital machinery. Now comes the key point of this thinking – we may not even know that they are regulating us.”
But apart from the rather fancy possibility that humans may become omnipotent in a symbiosis with machines, what else may act is mitigants against the growing loss of jobs under the relentless pressure of automation? Susskind’s (careful) optimism is built on one (big) premise: the Big State. The Big State is supposed to ensure just redistributive policies through taxing capital and big businesses to introduce a kind of Universal Basic Income (yeah, again!) with some strings attached (Susskind calls it “a Conditional Basic Income”). He makes a step further suggesting that the Big State should also be instrumental in sharing out the valuable capital itself, which is the source of that income in the first place (a similar proposal is advanced by Piketty who suggests a grant amount to be paid to every citizen to equalize the starting conditions). Further, the Big State is supposed to actively defend the world of work from the changes that are unfolding, not just be passively carried along on the current of technological progress. It should be the labor-supporting state: stepping in to support workers during the transition, to make sure that whatever jobs remain are well-paid and high-quality. The Big State is also supposed to keep in reign the Big Tech by establishing a Political Power Oversight Authority which would regulate the economic power of these big companies to allow regulators to identify, in a clear and systematic way, when power is being misused. Lastly, the Big State will define the leisure policies and the (shrank) world of work to maximize their utility for the society as a whole.
This is a conclusion, particularly symptomatic in light of the current developments after the COVID-19 pandemic has demonstrated, in the words of Marianna Mazzucato, “the big failure of small government”. After years of advocacy for small government tinkering on the margins of capitalist economies as some sort of a janitor correcting market failures, the issue of strong government directing and guiding development and ensuring fair distributive policies in the interest of society as a whole, not just its most privileged groups, is taking more and more hold. This is the major concern of many economists and political scientists, such as Joseph Stiglitz, Fred Block, Branko Milanović and Marianna Mazzucato (whose works you can find reviewed in this blog), to name just a few. Isn’t it the best demonstration of the Marxist central thesis that at some point the productive forces come into conflict with the existing relations of production, resulting in a new type of relations of production that adequately accommodate the potential of the productive forces?
David Harvey noted once that a conscientious researcher of the relationship between technologies and labor inevitably stumbles upon the underlying contradiction between labor and capital that is driving the trend of increasing automation. To what extent different researchers acknowledge this is a matter of their choice (and political inclination). Martin Ford in The Lights in the Tunnel recognized that there may be some sort of broad similarity between his argument and that of Marx but was at pains to distance himself from the damaging consequences of any such association. This is not an issue for Susskind because he does not see that his analysis (and the solutions he proposes) are latently driven by the need for continued capital circulation. Capital exists as a continuous flow of value. Capital must circulate continuously or die. The speed of circulation is also important.
And here’s the contradiction nested in the heart of capitalist technological development: new technologies and automation accelerate the speed of capital circulation by reducing production, delivery and payment processes but at the same time, it erodes the purchasing power and the aggregate demand of exactly the same groups that consume most (in percentage terms). The Warren Buffets and Bill Gates of this world do not consume nearly as much in percentage terms of their income, let alone their wealth, as a Bangladeshi factory worker or a Ugandan farmer. There is an apocryphal story about Ford (the other Ford, Henry, the founder of the Ford Motor Company) and a trade union boss looking at a fully automated production line in from of them. Ford says: “See these machines? They don’t ask for a pay rise and never go on strike.” The trade union boss dryly returns: “And will they also buy your cars?”
Susskind’s discussion of technology and its impact on work is strangely divorced from the real work of capitalist production with its purposive drive of capital to maximize profits, facilitate endless capital accumulation and reproduce capitalist class power. Where he occasionally does mention Marx, it is rather superficial and not really central to the argument. For Susskind, technological progress appears as a standalone phenomenon, a thing-in-itself. This is exactly what Marx called “fetishism”: when the certain relationships (especially production and exchange) are perceived not as relationships among people, but as social relationships among things (for example, the money and commodities exchanged in market trade). In a capitalist society where capital is striving to establish control and dominance over labor (a trend Susskind mentions in his book), the only solution to prevent the death of capitalism under continuous reduction of social labor as a source of value is a Keynesian-type solution to boost aggregate demand. (One immediately recalls Keynes advice to the government to pay people to dig wholes in the ground and then fill them up.)
This is why all proposed solutions (not only by Susskind but by many others as well) boil down to some kind of a state-mandated tax system to recuperate the productivity gains created by the new technologies and then redistribute these funds as purchasing power to the dispossessed masses on an incentivized basis. In return for the funds people are expected to commit to creative or worthy social activities and contribute to the common good (a version of the Conditional Basic Income suggested by Susskind).
As David Harvey stresses in Seventeen Contradictions and the End of Capitalism, Marx considered the issue of automation from the production side and saw a number of alternatives to the tendency of profit rates to fall as a result of labor-saving innovations, such as the opening up new product lines that were labor-intensive; a pattern of innovation that was devoted as much to capital saving as to labor saving and so on. As a minimum, it indicates a state that is not only correcting market failures but proactively directs and guides the process of innovation focusing on those with maximum social value. Do we really need further advances in e-trading where the speed of trading is measured in milliseconds? How much social value does it create? Is an ever-growing level of consumption (which is just a way to maximize profit) a good thing (let alone, sustainable)? Short of a social revolution, changing these patterns requires a much more balanced relationship between labor and capital, a trend that Paul Mason and Branko Milanović, among others, see as an emerging possibility of reversing this relationship so that labor hires capital and not the other way around.
Of course, the looming question is why and how these relations will change and whether it is possible in an evolutionary way. Marx argued in Grundrisse that technological innovation was a crucial weapon in class struggle and that the purpose of innovation under capitalism was to replace labor power where this is lacking, but rather in order to reduce massively available labor power to its necessary measure. For him, however, automation was a source of hope, rather than of concern, on two accounts. Firstly, he believed that increasing automation hastens the demise of capitalism by increasing pauperization of the labor force and thus sharpens the contradictions immanent in the capitalist mode of production. Secondly, he believed that automation was the foundation of communism that would allow transition from the realm of necessity to the realm of freedom for humanity. In a situation when the means of production are socialized (and social labor remains the source of value), the fundamental capitalist contradiction between labor and capital is lifted, eliminating the problem of large numbers of redundant workers.
I nostalgically recall discussions during my high school and student years in the Soviet Union about the communist future. The future seemed bright. Of course, it was meant to be a highly automated future with minimum hard physical work and an abundance of goods created due to this high level of automation. Such a future carried no danger of unemployment, let alone poverty, because communist society will create the best conditions for full self-realization of its citizens who will conscientiously strive for a maximum contribution. This is no doubts an idyllic picture. My point however is that the future full of automation and technological innovation in a communist society was a source of inspiration and hope, not of anxiety and desperation as Susskind’s book demonstrates.