From a refreshing trip to Scotland this summer I brought back, among other things, a visit to two great bookshops: the indescribable Leakey in the northern city of Inverness and the Topping & Company bookshop in Edinburgh. Consequence: a good handful of books in the suitcase as a pastime and intellectual fuel with which to pass the autumn.
One such book is Daniel Susskind’s ‘A World Without Work’. The book, well structured, is articulated in three parts of similar argumentative weight: the Context, the Threat, and the Response.
The Context
Susskind establishes early on the playing field of his work: the economic implications of automation or, in the updated version of the concept, of Artificial Intelligence (AI). And, with that, he designates the teams: on the one hand, those who argue that AI will bring new and more jobs. That the gains will compensate for the losses, that the phenomenon will not escape the phenomenon that economists are so relieved about, that of Schumpeter’s ‘creative destruction’.
And in the opposing team, those who say no, that with AI everything will be different, because its exponential development will not stop at tasks that are complementary to human ones, but that, sooner rather than later, we will see how it is capable of performing more and more substitute tasks.
The fact is that, to date, technology has favoured work. Today, despite technological development, more people are employed than ever before. Moreover, the number of hours worked per person is at a record low and productivity is at a record high.
This is because, of the two forces at work in relation to technology and labour, the substitutive and the complementary, it is the latter that has acted more strongly. In agriculture, for example, machinery has replaced labour, but today more food is produced than ever before (the ‘pie’ has grown bigger), and jobs have been transferred from land-based tasks to agro-food industries and agricultural tertiarisation (the ‘pie’ has also changed).
The result is that we have lived and are living in the ‘age of work’, which has seen the most routine tasks automated to leave more time for humans to devote to creative or empathetic tasks, for which machines are not yet ready.
The ALM hypothesis – the acronym stands for the initials of its authors – states that it would be enough to keep inventing non-routine tasks to be safe from the professional intrusiveness of machines. This is an optimistic approach and, for the time being, the dominant one among economists.
Against this, Daniel Susskind argues that, as the capabilities of machines increase, that potential pool of tasks reserved for humans is shrinking by leaps and bounds. Not even artistic creation is spared, as demonstrated by many of the artistic exhibitions that, using AI, we have organised at the Zaragoza City of Knowledge Foundation.
This great development of AI would not have been achieved if the initial focus on imitating human behaviour had been maintained. Instead, the current second ‘wave’ of AI is already capable of surpassing human capabilities in many areas of thought, precisely because it works with mechanisms that are totally different from those we use to reason.
The Threat
All three types of human skills – manual, cognitive and affective – are increasingly threatened by machines. This is what Susskind calls ‘task encroachment’, a phenomenon that is not homogeneous, but operates at different speeds depending on the sector and the place.
This invasion of tasks by technology causes, also in a heterogeneous way, a series of frictions in the labour market that lead to pockets of unemployment resulting from three types of misalignment between what jobs require and what people can offer:
- skills: the person does not have the technological level required for the new jobs,
- geographic: the technology skills are not where they are needed,
- and identity: many people without technology skills find the jobs at the bottom of the system demeaning and are unwilling to apply for them.
That the system has no incentive to stop this ‘technological unemployment’ is shown by the fact that US industry today produces 250 times more than it did in 1950 with 40% fewer workers.
However, it is not only at the lower end of the labour market that there are problems. Susskind shows how, especially since 2000, people with a university education are increasingly performing less creative tasks and are therefore more likely to be taken over by machines, thus increasing the pressure on the upper end of the labour market exerted by technology. Today, the new giants of the digital economy such as Google, Amazon, Microsoft, or Apple are turning over more than their industrial predecessors with much less skilled labour.
Curiously, within large companies in the digital age, wage inequality is rampant. While the ratio of CEO salaries to average salaries in 1980 was 18:1, today it is 350:1. No one is surprised, therefore, that while productivity in most sectors has soared over the last 75 years, hourly wages have remained more or less constant for the last 50 years.
And quite logically, this wage disparity within companies is passed on to society as a whole. In many of the most advanced countries, the top 0.1% have as much wealth as the bottom 90%. Inequality – and Thomas Piketty, whom Susskind quotes extensively in this part of the book, argues this better than anyone else – will be (is) already our main social problem.
The Response
Susskind has it, and it has three fronts: the educational, the political-economic, and the personal.
At the educational level, and based on the recognition that the education system mainly teaches us the skills for which machines are better prepared than we are, he proposes reinforcing the teaching of values, empathy and care. School teaches us more to be accountants than social workers. Education should avoid concentrating on the early stages and accompany people throughout their life cycle – at least in the active working period – and should stop being point-multipoint (one teacher for 20 pupils), and become more personalised.
In the political-economic sphere, Susskind advocates without hesitation raising taxes: on capital, inheritance and, of course, on robots. This would then be redistributed across the board through a Minimum Living Income, conditional, of course, on voluntary work. It also proposes strengthening the role of the state as a ‘holder’ and manager of funds for the common good, as Norway does with its sovereign wealth fund.
The chapter on political-economic measures closes with special attention to Big Tech, which Susskind proposes to control through a ‘Political Supervisory Authority’, not because of the danger of economic monopoly – something that has always existed – but fundamentally to limit its political power. (For those sceptical on this matter, the concerning role of Ellon Musk in the Trump Administration should dispel any doubts).
Daniel Susskind ends ‘A World Without Work’ by reflecting on the meaning of our life without work, a false dilemma, as it is about time to recognise the price of voluntary work – which includes running a home and family, raising children and caring for our elders. And Susskind also proposes better preparation for the enjoyment of the vast amount of ‘leisure time’. In a future leisure economy, we will need to educate ourselves to use it responsibly. If in the age of work the state was geared to teaching us how to work, it may be time for it to be geared to helping us enjoy ourselves outside of traditional work.
How To Live
Susskind warns us before concluding, not without some irony, that we should not take the title ‘a world without work’ literally. The work that will be scarce will be paid work, and there will be more opportunities to work at what we like, and perhaps the very act of writing this article is an indication of that.
And here the book links, perhaps unconsciously, with one of the first philosophers of modernity: Michel de Montaigne. Equal parts politician, intellectual, and enjoyment-seeker, Montaigne wanted to investigate the purpose of our individual lives on the skin of this planet. Want to find some highly topical reflections on how to live in a world in which work is not the main protagonist? Read Susskind, but then read Montaigne.
And don’t forget to take a look at the ideas put forward in the mid-1950s by the Parisian situationists led by the indescribable Guy Debord. They were among the first to point out – I was going to write seriously, but nothing in early situationism was really serious – that the aim of automation could be none other than to exchange work for enjoyment. Since then, nothing really propositional had fallen into my hands in relation to how to take advantage of Artificial Intelligence to live better.
Article published under a Creative Commons free culture licence. Some rights reserved
