“What do you like to read?” asks the interviewer. “I like a very specific kind of book,” the writer answers, “one that gathers information, beauty, and horror.” With his debut novel, Chilean author Benjamín Labatut wrote exactly the kind of book he likes to read. When We Cease to Understand the World, translated by Adrian Nathan West, pieces together dream-like stories about some of the most haunting scientific discoveries of the 20th century, from Werner Heisenberg’s uncertainty principle to the development of the Zyklon-B compound used in Nazi extermination camps. Some reviewers of the novel were critical of Labatut’s fusion of biographical detail with fictional storytelling to foreground the beauty and horror wrought by the human mind over the past few centuries. In the wake of W.G. Sebald and Roberto Bolaño, both major influences on Labatut’s writing, that kind of critical dismay seems puzzling. When Sebald was once asked about the pollination of the historical with the fictional that defines his work, he responded: “You adulterate the truth as you try to write it. There isn’t that pretense that you try to arrive at the literal truth. And the only consolation when you confess to this flaw is that you are seeking to arrive at the highest truth.” In an interview for Physics Today, Labatut gives an echoing response to a similar question: “You have to be willing to pervert and distort your raw materials, because you’re trying to reach a truth that is very particular to fiction, that is concerned with what is mysterious, incomprehensible, and dark.” Indeed, the authors he admires—among them, Jorge Luis Borges, J.A. Baker, Bruce Chatwin, Philip K. Dick, Eliot Weinberger, and of course Sebald and Bolaño—all practiced this kind of distortion of factual reality to give voice to what “the literal truth” alone cannot tell us.
The MANIAC is Labatut’s second novel and his first to be written in English; like its predecessor, it seeks to re-enchant our imagination with the very literal but mysterious truths revealed at the frontiers of scientific inquiry. The novel is a kind of triptych, presenting us with the conception, painful birth, and exponential growth of the digital computer and its own disquieting offspring, artificial intelligence, with each section focusing on a different man: one who foresaw, one who laid the foundations for, and one who got to compete with, AI. The brief opening panel, titled “Paul, or The Discovery of the Irrational,” relates the tragic story of Austrian physicist Paul Ehrenfest, who was among the first to intuit the consequences of the intellect-defying world revealed by quantum mechanics and who foresaw the advent of what he called “an inhuman form of intelligence.” (He would eventually die by suicide after killing his mentally ill son.)
The book’s sprawling middle section, “John, or the Mad Dreams of Reason,”, traces the inception of AI through a series of fictionalized first-person testimonies of those who knew the inhumanly clever Hungarian mathematician John von Neumann, among them his close friend and Nobel laureate in physics Eugene Wigner; renowned American physicist Richard Feynman; and members of von Neumann’s family, not least Klára Dán, his second wife and lead programmer of The MANIAC I, an early computer based on von Neumann’s blueprint and built to crunch the numbers needed for the creation of the hydrogen bomb. Feynman’s off-the-cuff monologues shine in the book, making you feel as if the avuncular physicist were right by your side, shooting the breeze about building the world’s first nuclear bomb at Los Alamos and beating the MANIAC at chess. Dán’s expressions of outrage and affection for the infantile Johnny are worthy of comparison with the consciousness-torrent that is Molly Bloom in James Joyce’s Ulysses
. And Wigner, a true friend to the man many considered to possess the most powerful intellect of the twentieth century, comes across as lucid, humane, and perhaps the most eloquent of all the book’s characters.
Some of the novel’s other first-person narrators, however, feel more mannered. In a couple of passages one is reminded of the ominous tones of H.P. Lovecraft, whose influence Labatut has acknowledged and about whom he wrote in an earlier book. The mathematician Theodore von Kármán, for instance, at one point warns that “uncovering the foundations [of all we know] is always dangerous, for who can tell what lies in wait among the fault lines in the logic of our universe, what creatures sleep and dream amid the tangle of roots from which human knowledge grows?” Elsewhere, Sebald’s probing, tastefully academic style is discernible. We hear it, for example, in the story told by economist Oskar Morgenstern, co-author with von Neumann of the seminal 1944 text Theory of Games and Economic Behaviour
: “I myself suffer from a morbid sense of despair, and even now, decades after I worked with von Neumann, I still find myself questioning our central tenet: Is there really a rational course of action in every situation?” Perhaps the most stylistically mannered testimony, though, is symbiogenesis pioneer Nils Aall Barricelli’s long lament of complaint, which opens with a paragraph that could be straight out of Edgar Allan Poe: “I’m not a madman even though they’ve called me one many times. But I’m not mad. In all my troubled years […] I did not abandon my wits, or let dismay drive me past folly and into madness. But I could have. I could. Because I know madness.” Such mannerisms, however, do not distract from the novel’s pervading mood of a kind of creeping, Lovecraftian unease. If anything they deepen it, as the writing casts the spell of a story channeling the unthinkable will, ineffable force, unconscious drive, demon, or whatever it is that guides our species to pursue technological progress as if by default: as if our minds were wired to seek the limits of all things, not least our own intellect, and to sound the depths of the cosmic void with elegant formulas whose ultimate meaning we cannot even conceive. “Writing,” as Labatut has said, “is the last remaining dark art.”
Clearly, the story told in The MANIAC is not after verisimilitude for its own sake. To criticize the novel for lacking in developed characterization—in this context, this would all too easily translate into a prose-parlor of professional bickering, psychological probing, and well-crafted gossip about the vices of Johnny—is to not give due weight to the “mysterious, incomprehensible, and dark” pulse that informs its every line. One can only be grateful to the author for not fattening what needed to stay lean, so as not to distract attention from what matters thematically: that is, the novel’s attempt to express, in the troubling words of the doomed Ehrenfest, the birth of “the strange new rationality that was beginning to take shape […] a profoundly inhuman form of intelligence that was completely indifferent to mankind’s deepest needs.”
The book’s polyphonic middle leads us on to the triptych’s concluding panel: “Lee, or The Delusions of Artificial Intelligence.” Here, we are treated to compelling reportage as Labatut relates the legendary Go match between Lee Sedol, perhaps the greatest human player to play the ancient game in modern history, and AlphaGo, a computer programme conceived by Demis Hassabis, co-founder of DeepMind. Labatut has cited the 2017 documentary AlphaGo as one of the sources of inspiration for his book, and watching it certainly deepens one’s appreciation of the precision of the writing in recreating this AI milestone on the page. (AlphaGo beat Sedol four games to one.)
By the time you reach the novel’s end, Johnny von Neumann has stolen the show; you just might be convinced that his mind is as close as our species is likely to get to thinking with the speed, retentive power, and amorality of the digital computers that he made possible. What is certain is that von Neumann’s design of the modern computer helped spread the spores of code that have grown into the present global rhizome of algorithmic social networks and artificial intelligence servers.
Toward the end of his life, when he was ill with cancer and before his prodigious brain began to deteriorate, von Neumann was asked what it would take for an intelligent machine to resemble a human mind. He said that it would have to grow on its own. That it would need to learn to use language freely. “And he said that it would have to play,” Labatut writes, “like a child.” Today, we are witnessing the growth of such an entity; it’s still very far from resembling us, but it’s learning fast. The question remains as to who, given enough time, will come to set the rules of the game.
Labatut reflects on this question in La piedra de la locura
(The Stone of Madness), a slim volume of two essays published shortly after his first novel. The book takes its title from a small Hieronymus Bosch painting hanging unostentatiously in the Prado Museum in Madrid. The painting shows a man sitting on a cumbersome wooden chair with a medieval doctor standing behind him; to the man’s left, there’s a nun with a red book balanced on her head and a tonsured monk holding a silver jug. The doctor is in the process of removing the proverbial stone from the man’s opened skull with a scalpel. If we look closely, though, what at first seemed to be a stone starts to resemble a sickly flower; what if, Labatut asks in La piedra de la locura, the doctor was not removing the stone of madness from the man’s head but putting something inside? The medieval surgeon who, judging by his demeanor and blank expression, might as well be an android experimenting on the human organ responsible for its creation, could be planting the fatal flower of progress inside the mind of an unresisting fool.
After reading La piedra de la locura, it occurred to me to ask one of the many AI image generating systems now freely available online to create, if that is the right word, an image. I type in the prompt “Doctor inserting a flower inside the head of a man sitting on a chair in the style of Hieronymus Bosch.” The screen loads for a few seconds and, just like that, there’s a hodgepodge of pictures. The strange images look like dream-flotsam of some-Thing from another place, outside the compass of human thought. Most of them are wrong and all are awry in a way hard to explain. But there was one that stood out.
This strange juxtaposition of two figures produced by the text-to-image model has distorted the sense—in Sebald’s phrase, “the literal truth”—of my words. Instead of a doctor, we see an eerie replicant holding a taskmaster’s cane. It is looking straight into the eyes of an old scholar with an open book on his lap, a warped finger keeping his place in the text. Above the replicant’s head there’s a floating crimson flower-like shape: The flower of progress, once planted in the head of a hapless fool, has borne startling fruit.
In The MANIAC, von Neumann’s story closes with a lyrical paraphrase of the closing pages of “Can We Survive Technology?”, an essay the great mathematician wrote for Fortune magazine in 1955, in which he proclaimed: “For progress there is no cure.” In Labatut’s recreation, von Neumann calls technology “a human excretion [that] should not be considered as something Other. It is a part of us, just like the web is part of the spider.” These words are not von Neumann’s own, yet the spirit of his original essay is all there in Labatut’s artful distortion. By an uncanny coincidence, bound to become routine, the AI model’s hallucinatory distortion of what I asked it to depict happened to capture the dilemma at the heart of the novel. This distorted dreamscape may not be artful, but it does show us our current predicament, which von Neumann foresaw before his progeny had come into its own.
