Ray Bradbury said that a function of science fiction is not only to predict the future, but to prevent it. Citing Nineteen Eighty-Four, he proposed that this particular future was less likely to happen because Orwell had so vividly imagined it. Frank Herbert made a similar observation, also with reference to Nineteen Eighty-Four. In her 2000 essay A Few Rules for Predicting the Future, Octavia E Butler turned to this question of cautionary dystopias and proposed a caveat: futuristic fictions are really deep examinations of the past, with its ‘repeating cycles of strength and weakness, wisdom and stupidity, empire and ashes’. Trying to foretell the future without studying history, Butler continued, would be ‘like trying to learn to read without bothering to learn the alphabet’. In The Rise and Fall of the Artificial State, Jill Lepore argues that this is precisely what Big Tech companies and aligned governments are doing: they’ve become somewhat future-fixated and have neglected their studies of the past. Furthermore, they appear to have ‘misread’ various dystopian novels, taking them not as warnings but as ‘technical writing, a set of assembly instructions’. Lepore explains what this means, how she understands the concept of the Artificial State and why taking inspiration from dystopian novels is probably not a good idea.

Lepore is a Harvard professor who teaches both history and law, and who has previously written on questions of myth and reality, evidence and truth. Her book If Then, for example, tells the story of Cold War data mining and the Simulmatics Corporation, a strange precursor of our automagical tech companies (the word ‘Simulmatics’ is a clunky merging of ‘simulation’ and ‘automatic’). In The Rise and Fall of the Artificial State, Lepore turns to contemporary data mining and predictive algorithms. Governments and corporations, she writes, increasingly ‘automate political behaviour and public discourse’, replacing ‘democratic deliberation’ with ‘prediction by calculation’. Pre-empting charges of neo-Ludditism, she points out that she is not anti-technology: artificial things can be ‘beautiful, wondrous, hallowed’ (she lists pianos, foetal heart monitors, suspension bridges, computers and digital libraries). Also, verbal language is ‘artifice’ too: mysterious reality is not intrinsically formed of words. The trouble arises, she suggests, when ‘perfectly sensible automation’ is overtaken by chaotic, market-driven ‘government by machine’ and the replacement of humans by artificial intelligence. This is the Artificial State, Lepore argues, and it is a ‘fantasy half built’, a ‘precipitate’ and ‘heedless’ transformation that imperils ‘civil society … equality of rights, freedom of thought [and] freedom of expression’.

Lepore presents a history of ideas which, she argues, are fundamental to the Artificial State. These include two assumptions: first, that ‘the entire universe … is a machine that operates according to deductible laws, rules subject to quantification, calculation, and prediction’ and second, that ‘machines determine the course of history