Home Archive Shop About us
©2023 Dispatches Media
Prisoners of Logic

The MANIAC, by Benjamín Labatut

Emmerich Anklam
December 1, 2023
von Neumann (L) and J. Robert Oppenheimer in front of the IAS Computer in Princeton, New Jersey.

“Anything that comes out of a writer,” Benjamín Labatut has said, “is fiction.” Born in Rotterdam and now living in Santiago, Labatut is best known for his genre-defying When We Cease to Understand the World, which received near-unanimous acclaim when its translation from the Spanish appeared in 2020. The fictional characters in his books are real-life scientists, people whose conceptual breakthroughs have led to death and destruction at scales that seem almost metaphysical, but he is not a historian of science. Written in English, The MANIAC, Labatut’s latest book, is a dreamlike examination of the tormented visionaries to whom he is drawn like cobalt to a magnetic field. It is also a mediation on the ways that certain forms of reason and extreme mental power can verge on madness, atavism, and apocalypse.

Formally, The MANIAC is a nested triptych: a three-part inquiry whose middle splits into three further segments. Labatut, who refuses to call the book a novel, focuses on three exceptional figures: the theoretical physicist Paul Ehrenfest (1880–1933), the mathematician and computer scientist John von Neumann (1903–1957), and the master go player Lee Sedol (b. 1983). His default narrative voice is suitably godlike, aloof but grandiose. One of his epigraphs to the second part of The MANIAC is a long quotation from Adam Curtis’s 2021 documentary series Can’t Get You Out of My Head, and his style shares much with Curtis’s stern end-times sermonizing. When he’s not careful, his sensationalizing approach obscures the fact that his subjects were actual human beings. Take his introduction of John von Neumann, which is split into italicized fragments across a dozen pages and sounds like the voiceover to an action-movie trailer from the nineties: “He was the smartest human being of the 20th century.” [turn the page] “An alien among us.” Four pages later: “His name was Neumann János Lajos.” [turn the page] “A.k.a. Johnny von Neumann.” (By which flawed measure of intelligence is Labatut measuring “smartest,” anyway?)

The brief first part of The MANIAC follows the life and horrific death of the Austrian physicist Paul Ehrenfest, beginning with a sharp shock, hair-raising and nausea-inducing. It’s especially unsettling to read that Ehrenfest, who in September 1933 murdered his 15-year-old developmentally disabled son, nicknamed Wassik, before committing suicide with the same pistol, was no run-of-the-mill sociopathic genius but rather, in Labatut’s characterization, a man of uncommon empathy. “[The] only thing that really sparked joy in Paul,” Labatut writes, “was giving himself to others.” He was an inspiration to the budding physicists who came to study with him in Leiden and remained his devoted admirers, including J. Robert Oppenheimer. Einstein called him “the best teacher in our profession.”

How could such a generous person be driven to such an awful act? Ehrenfest loathed fascism, violence, cold logic. He was terrified by the ascent of quantum mechanics—how it replaced “real, physical intuition” with “brute-force artillery,” how it was training his students to become human calculators. In his eyes, physics’ new direction threatened to dehumanize the world, to transform people into automata. (By this point, of course, many branches of science had adopted numerous dehumanizing tendencies essential to war, slavery, and conquest.) Ehrenfest raged against the fate he predicted, his desperation increasing as his career hit the skids and Hitler took power in Germany. No longer, he believed, could he financially provide for or protect his son, who was institutionalized in Amsterdam, nor did he want to burden his wife or other children with Wassik’s expenses after his own suicide. Perhaps Labatut’s cautionary tale implies that Ehrenfest, who “was not a computer,” so deeply absorbed the violence and fatalism roiling everything around him that he ended up becoming what he fought so desperately to avoid. In a scene painting the last day of Ehrenfest’s life, Labatut describes him, the murder-suicide weapon in his pocket, as “machinelike, propelled by a force he neither recognized nor understood . . . his legs as stiff as an automaton’s.”

This is where we leave Ehrenfest, begin the central part of the book, and meet John von Neumann. In Labatut’s telling, von Neumann’s singular mathematical abilities stem from the same source as his “almost childlike moral blindness.” Von Neumann detests fascism just as Ehrenfest did, yet when he sees a parade of Reichswehr tanks he “[begins] to salivate like a dog who hears his master emptying leftovers into his bowl.” Among his contributions to the Manhattan Project, he calculates the height at which the atomic bombs would cause the most destruction when detonated (the bombers use his calculation). His childhood love of a mechanical textile loom becomes one of his primary inspirations when he eventually builds the device after which Labatut names this book—the Mathematical Analyzer, Numerical Integrator, and Computer (MANIAC). In von Neumann’s final months, lying in a guarded room at Walter Reed Army Medical Hospital that he orders to be filled with toys, he responds to a question from his daughter, Marina, with unsettling solipsism:

I asked him, point-blank, how he used to be able to contemplate, with total equanimity, the killing of hundreds of millions of people in a nuclear first strike against the Soviet Union, and yet he could not face his own mortality with any sense of calm or dignity. “That is entirely different,” he replied.

Had Labatut chosen to narrate all two hundred pages of von Neumann’s part in his usual godlike style, The MANIAC might have collapsed under its own portentousness. Instead, he makes a narrative decision that opens up the book and makes it richer and more rewarding: He tells von Neumann’s story through the voices of more than a dozen people who knew him. Sometimes the author is visible through the mask; Eugene Wigner in particular, who receives the most space, often lapses into sounding just like Labatut. But the chapters narrated by Richard Feynman, Gábor Szegő, and above all Klára Dán von Neumann are thrilling.

Klára Dán and her spouse, John von Neumann. Courtesy Los Alamos National Laboratory

Labatut’s Feynman is chatty, irreverent, an evocative scene-painter, alive to the role of accidents and chance in history. He pinpoints the absurdity of life at Los Alamos, especially against the enormity of their task. Szegő, who relates the story of von Neumann at the Reichswehr parade, is a contrarian, too, but he is also the religious conscience of the book. When he recounts von Neumann’s dreams of a mathematics reduced to axioms without contradictions, he says, “And what kind of an Eden would that be? I wondered. Surely not a place where plants and trees could grow.”

The chapters narrated by Dán—the first woman to write modern computer code, and John von Neumann’s wife from 1938 until his death—are the highlights of The MANIAC. Dán plays an essential role in the history of twentieth-century technology, but she resists her husband’s moral tunnel vision. And she does the most to expose the scientists around her as people, not gods. She wanders the woods surrounding Princeton’s Institute for Advanced Study, rifle in hand, knowing that the men of science nearby were “easy prey for any hunter, because they were so caught up in themselves and completely indifferent to the world outside.” When John can’t be bothered to open the garage door, Dán locks John out of his study, threatening to lock his papers on fire as he rages outside like a toddler parted from his favorite plaything.

After Labatut follows John von Neumann’s early work on mathematics and his involvement with the atomic bomb, he focuses on von Neumann’s formative contributions to computers and AI. Obsessive work on the MANIAC, theoretical leaps, and broken relationships ensue. By the time of his death—and the MANIAC’s retirement around the same time—von Neumann had laid much of the theoretical groundwork for computers as we know them today. It’s astonishing to recall how powerful computers have become in the intervening six or seven decades—and especially how adept they are at playing games.

The third and final part of the book recounts the 2016 go match between 9-dan player Lee Sedol and Google DeepMind’s AI-driven AlphaGo program, which AlphaGo won four games out of five. Labatut acknowledges a debt to the 2017 documentary AlphaGo, and his voice here sounds so much like documentary narration, it’s sometimes unclear what sets his version apart. His description of AI’s escalating gameplay abilities is disturbing, recalling his description of humanity’s helplessness before the power of the atomic bomb. But does the story of AlphaGo and Lee Sedol say anything about the world outside the board? In an earlier chapter about John von Neumann’s cofounding of game theory, Labatut has the theory’s co-creator Oskar Morgenstern ask:

Is there really a rational course of action in every situation? Johnny proved it mathematically beyond a doubt, but only for two players with diametrically opposing goals. . . . Life is so much more than a game. Its full wealth and complexity cannot be captured by equations, no matter how beautiful or perfectly balanced.

AlphaGo’s victory lies in the realm of two players with diametrically opposing goals, and because it does nothing to defy Morgenstern’s pronouncement, I wonder whether this final part adds anything more to The MANIAC. Throughout the book Labatut meditates on games and their role in society; he has Feynman describe go as “really intractable to computation” before we learn otherwise, and Mariette Kövesi wonders to what extent war simulations have fed the human appetite for real wars. Still, if Labatut is interested in the growing power of computers over people, a tightly bound two-player game is not a good model of the actual world.

Penguin Press, 2023

The games of go and chess work only because their players choose to adhere to the rules of play. The future, however, is not a closed game whose rules we must accept wholesale from people who act as though the future is theirs to master. Klára Dán is a compelling character in The MANIAC precisely because she refuses to play by the rules of the game. She reminds us that sometimes it is in our power to reject the mad dreams of reason.

When the brilliance of a scientist is the starting point for a book, it can obfuscate the book’s search to understand science as a shared human pursuit. When We Cease to Understand the World sometimes succumbed to this problem. The MANIAC does a better job resisting it, albeit an imperfect one. Other novels, like those of Kim Stanley Robinson or Richard Powers, acknowledge that modern science requires a wide range of people who understand the significance of cooperation and collaboration if they wish to get anything significant done. The best books about scientists working together also offer portraits of brilliant individuals in great depth. Robinson, Powers, and Yaa Gyasi, in her 2020 novel Transcendent Kingdom, do this, and so does Richard Rhodes in his superb history The Making of the Atomic Bomb. They can make graceful pivots from the many to the one. Going the other direction is harder.

If we hope to undo the notion of individual genius that so many power-hungry tech leaders exploit, we must pierce the aura that surrounds many legendary minds, the aura Labatut both upholds and undercuts throughout The MANIAC. Like the historical figures it centers, Labatut’s latest fiction does not deserve to be encased in a fog of reverence. Let us be skeptical here, at the very least, so we may better scrutinize the godlike narratives that accord today’s titans of STEM. Worship does no justice to science.

The MANIAC
by Benjamín Labatut
Penguin, 2023
$28.00

Emmerich Anklam is managing editor at the independent publishing house Heyday. He grew up in Santa Rosa, California and lives in Berkeley.