Last week, purely as an experiment, I put to ChatGPT some straightforward arithmetic to see how it would approach the question: “If something grows at an annual average rate of 2.8% over 7 years and reaches 155bn at the end of the 7 years, what level did it start at?” ChatGPT produced not only the numerical answer, but also a concise tutorial on the formula for compound growth. The experiment made me reflect that widespread application of AI seems likely to de-value learning, and knowledge itself: what is the point of learning anything, I mused, if an AI-powered chatbot can offer the needed ‘answers’? And if knowledge (or something that passes for it) becomes commoditised, how will it retain any value?

This, it transpires, is one of the main themes of Robert Skidelsky’s sprawling treatise ‘The Machine Age’. In the strongest section of the book, on the impact of technology on work and labour markets, he notes that “The world of work in the machine age is a world of lost skills”, and defends this assertion with numerous clear examples and thought experiments. Skidelsky demolishes the oft-touted reassurance that “With suitable education and training workers will be able to take on higher-level…tasks in any job, whilst the more menial tasks will be taken up by machines” by marshalling compelling arguments about the likely impossibility of the needed labour market transition, and by examining economic data that cast serious doubt on the idea that low unemployment rates in the rich world,  where take-up of technology has been highest, mean that fears of serious labour market disruption are unfounded.

The arguments would have been even stronger, however, without the inclusion of some broad generalisations. When Skidelsky writes “Most work is done both to earn money and to do something useful or worthwhile”, he comes down firmly on the side of social scientists who, he notes, link work with value—in contrast to economists’ view of work as a “disutility, the cost of living.” But are the economists definitely in the wrong? Indeed, as more jobs become lower quality, as Skidelsky argues is happening, is it not in fact natural for more people to see work simply as a disutility?

The central question posed in the section on work demonstrates the sweeping scope of ‘The Machine Age’: “Can an economic system in which the means of production are largely privately owned ensure that the gains of productivity are shared sufficiently widely to enable the future that Marx and Keynes both wanted? Or is a socialist economy required for this?” In the book’s final section, Skidelsky poses a similar critique of political rather than economic systems, noting “history does not suggest that a democratic system can deal with the scale of the problems now threatening the planet”.

One downside of this very extended scope is the necessarily cursory treatment of much of the material. Many sections of the book read as extended literature reviews on a particular topic, with pages of summary of other writers’ analysis of a topic rather than any new insights. A book that wears its “inter-disciplinary” label proudly, it references an astonishing array of scientists, economists, journalists, philosophers, novelists, playwrights, composers, and more, all the way from antiquity to the present day.

The author himself admits in the preface: “The book is over-ambitious. It is primarily about how western civilisation came to be captured by the dream of utopia through science and about the successive stages by which this dream turned sour”. However, it is far from a dispassionate analysis, for Skidlesky has a very definite point to make:  “The purpose of this book will have been accomplished if it dents the hubris of the engineers of the soul….The book’s message is that these ideas [of technological progress] will destroy the world we know.”

Skidelsky succeeds admirably in setting out several key thought-provoking and inter-related themes. Among them are that “progress turns all utopias into dystopias”; that “it is the reduction of humanness to the point of being controllable by automatic machinery that is the dominant tendency of information technology”; and the question of the extent to which machines bring about “liberation versus entrapment” (the title of the book’s penultimate chapter). The arguments are well-reasoned and written extremely clearly and succinctly. The book’s structure—based on “the application of the mechanical philosophy first to work and then to society”—works better in theory than in practice, since it lends itself to repetition of some key ideas.  And readers most interested in a detailed analysis of the implications of the machine age for our future will likely be frustrated by the sheer volume of historical, philosophical and literary contextualisation. ‘The Machine Age’ is therefore best approached as an informative, engaging and sweeping treatise on humanity’s millennia-long relationship with technology, and—for those who crave more detailed analysis—as a gateway to other reading.