The idea that machines might one day replace people in the productive process has haunted economic analysis since its foundation. After all, there are obvious physical limitations on what a human being can accomplish (in a day, a year or a lifetime), but increasingly sophisticated machinery could be infinitely improved in its productivity. The only questions have been: will people one day be completely redundant, and what will be the social and political consequences as this situation becomes more imminent?
Neither technology nor economics can, on their own, provide reliable predictions of the outcomes of these struggles. Once the potential for the automation of a series of service activities has been created, the questions of how these are implemented, and what provision is made for the people displaced, are for the interest groups mobilised in societies to settle, through political contest or, in the last resort, violent conflict.
David Ricardo was the first political economist to set out the circumstances in which automation would lead to unemployment and falling wages. He recognised situations in which capitalists might see the mechanisation of production as a means of increasing production without increasing output. He considered that this was something that could happen for short periods, so it was rational at such times for workers to oppose the introduction of new technology (as they often did in the 1820s and early 1830s in the North of England). But he believed that, in the longer term, economic growth could be sustained by increased proportions of national income (GDP) flowing into capital investment, improved methods of production, a mobile and adaptable workforce, population growth, and elastic demand for manufactured goods, at home and abroad (Ricardo , Principles of Political Economy and Taxation (1817), especially chap. 31).
Up to 1870, the Industrial Revolution in the UK seemed to justify Ricardoâs optimism, with only short periods of recession and extensive unemployment after 1835 (Ireland was the exception, with periodic famines and mass emigration). But then Germany and the USA became faster-growing industrial economies; massive supplies of rural labourers in the former, and new immigrants to the latter, contrasted with a British society in which farming had already become a source of less than 15 per cent of all employment and self-employment (Walker 1979, p. 271). It was not until the Great Depression of the 1930s, with the rise of Nazism in Germany and Fascism in Italy, that redundancy and mass poverty spread throughout the developed world economy.
War and the active labour-market policies of Social Democratic governments followed, until many of these issues again afflicted industrialised countries in the 1970s and early 1980s. By then, industrial production was being relocated in the Far East, Central and South America; services were replacing manufacturing as the main employers, while remaining manufacturing processes were increasingly automated. Now that Artificial Intelligence (AI) can take the place of many service workers and their managers, there appears to be nowhere else for them to go.
Some cities seem to be prospering through their success in combining university research and development in the field of robotics with the manufacture of products (from artificial limbs and autonomous wheelchairs to childrenâs toys) using AI. One of these is Bristol, which has large-scale aerospace industries, but also small new engineering companies (innovating in fields such as aids for social care) many of which take interns from the cityâs two universitiesâsomething to be celebrated in the right-leaning press (Sunday Telegraph, 25 August 2019).
The recent literature on this topic seems to be polarised between starry-eyed optimists, who see AI and robotics as the keys to future personal liberation, and gloomy pessimists, who predict that the potential gains from new technologies will simply be captured by dominant financial interests, while the most vulnerable members of societies are impoverished and coerced.
In this introductory chapter, I shall review these rival analyses and consider what might determine the outcome of the struggles, both economic and political, over automationâs impact. The issues are extremely complex, becauseâas already occurred over 60 years agoâtechnological change gives rise to global as well as regional and national effects. Already a phenomenon that was not foreseen by commentators on the demise of Soviet-style state socialism, the rise of authoritarian regimes in the USA, Brazil, Poland and Hungary, as well as their consolidation in Russia and China, has affected the likely outcome of the current wave of automation (Jordan 2019).
This is because of a close link between the Western version of the new authoritarianism and the social policy response to automation that can be traced back to the early 1970s. First in the UK and USA, and subsequently in Europe, the introduction of means-tested supplements to low wages (eventually called âtax creditsâ in the Anglophone countries) created a class of workers and their householdsâwith echoes of the Poor Law in Britain during the Napoleonic Wars (Jordan 1973), an insecure âprecariatâ (Standing 2017)âwhich was coerced and sanctioned by the benefits authorities, under schemes for âworkfareâ (Mead 1986) and âwelfare-to-workâ (Jordan 1986, 1996; Standing 2017; Haagh 2019).
This book will trace the relevance of these policies for the likely outcome of the political and economic struggle over automation of service activities and employments. I shall argue that none of this is yet determined, but that the optimists are making a big mistake if they imagine that the technology itselfâwhatever its potential for liberationâwill actually lead to freedom and leisure for workers, let alone the collapse of capitalism, without a mighty political struggle.
Science Fiction: Threat or Utopia
Automation has always seemed to carry a threat to humanityâs social systems. The earliest robots, designed by Leonardo da Vinci, were intended to inspire awe and to intimidate human armies. The first, a mechanical lion, was meant to impress the king of France on a state visit to Milan; the second imitated the movements of a knight in armourâboth were designed to move automatically through spring and clockwork mechanisms (BBC Radio 4, âLeonardoâs Lost Robotsâ, 2 May 2019). The Italian literature of the period showed that these were not isolated experiments, but parts of a Renaissance obsession with the potentially destructive relationships between machines and human societies.
Such fears were revived towards the end of the nineteenth century. In The War of the Worlds, H.G. Wells imagined an attack on suburban Surrey by an invading army of robots. Over the intervening years, robotics have transformed manufacturing, first in the advanced industrial economies of the West, and more recently in the Asian ones to which much of the mass production of goods had been transferred since the 1960s. But new concerns about the ethical dilemmas around increasingly humanoid automata are reflected in the novel Machines Like Me by Ian McEwan (2019).
In this, âAdamâ, a robot advertised as a âcompanionâ, is not only able to do practical tasks, but also to âthinkâ and discuss. Those purchasing this model can select among options for âpersonalityâ, such as âextraversionâ. The complexities of relationships between such automata and their âownersâ are rich material for the novelist.
These possibilities had also been explored in the series âHumansâ (Channel 4 TV, 2015â18), in which âSallyâ, a âSynthâ, âthe help you always wantedâ, was a life-like humanoid in a new generation of âPersona Syntheticsâ, closer to human in her characteristics than any previous automata.
Of course, as remote-control machines have become more sophisticated, the potential for their malevolent use by human beings increases exponentially. Drones have been used to devastating effect in recent wars; they have also been flown to disrupt passenger and freight aircraft by unidentified miscreants. So travel and trade are vulnerable to the abuse of these technologies, hitherto in isolated acts of opportunism rather than systematic attacks.
However, the much larger issue for the future of economies and societies is how AI and robotics are deployed in the services which now dominate employment. Optimists such as Rutger Bregman (2017) insist that a great many of these are âbullshit jobsâ, created as make-work roles in management and the professions (Utopia for Realists and How We Can Get There, pp. 162â5). If the potential for translating new technological innovations were to be fully realised, he claims, and the power of banks could be restrained (pp. 105â8), we could afford to pay unconditional Universal Basic Incomes (UBIs ) to everyone, without any requirement for them to work. All this, he argues, implies that the automation of every kind of routine work allows more labour power to be used in services...