“He was the father of game theory, and an early harbinger of a technology we may come to fear every bit as much as the atom bomb: artificial intelligence.”
In the summer, as Oppenheimer blazed its way across our cinema screens, we were invited to revisit an archetype: an old narrative of scientists driven by passion but manipulated by power; of discoveries made in the spirit of cumulative understanding but employed to the most nullifyingly monstrous ends. But what if that same passion served some malign power from the start? And what if manipulation and monstrosity undergird the very nature that scientists set out to strip bare?
Another figure present for the Manhattan Project was Johnny von Neumann. He was no archetype, as we learn from Benjamín Labatut’s The Maniac, but a new kind of man. He was so prodigious that a full enumeration of his contributions proves beyond the scope of this book. We can safely say, however, that he was the father of game theory, and an early harbinger of a technology we may come to fear every bit as much as the atom bomb: artificial intelligence.
Von Neumann forms the nucleus of this intensely diffuse novel. But it opens with the story of the physicist Paul Ehrenfest, who in 1933 murdered his intellectually disabled son Vassily before turning the gun on himself; it concludes with the story of the computer program AlphaGo, which in 2016 defeated the world’s best Go player in a feat many considered impossible. Even in the central chapters, we never encounter von Neumann directly, but through the first-person (and therefore biased) accounts of those who knew him – his friends, his teachers, his colleagues, his wives…
In attempting this approach, the writer must be dextrous enough to ventriloquise many voices convincingly. Conveniently for Labatut, von Neumann’s peers turn out to be stultifyingly unanimous on two points: his unique brilliance and the unsettling, “nonhuman” quality of that same brilliance. Labatut goes on making this point long after he has proven it, and so we are treated to such tiring banalities as this: ‘I once saw him take two books to the toilet, for fear that he might finish the first one before he was done.
And Labatut does not always capture the idiosyncrasies of voice so pivotal to the kind of polyphonic novel he has written. Most sound like variations on each other, down to their penchant for hyperbole, for the frenetic, and for the mystical. In other words, they almost all sound like Labatut. On those rare occasions when he tries to inflect a voice with the personality of its speaker, the results are mixed. The account of Richard Feynman reads like the product of a large language model (complete with American colloquialisms circa 1950). There is one clearly dissenting voice, stylistically speaking: that of Nils Aall Barricelli, who – in a breakthrough whose influence von Neumann never acknowledged – simulated the evolution of populations of digital organisms.
This story was a throwback to Labatut’s debut, When We Cease to Understand the World, in that it comes vividly alive through the kind of writing at which he excels: vignettes of scientists at work at their savage, obsessive vanities, ‘understanding by way of destruction. A lunatic’s insight.’ Although Labatut shows little here of the cold versatility he sees in his subject, he may yet secure his breakthrough.
Words: Diarmuid McGreal