Ian McEwan spins a sci-fi love triangle with a man, a woman and an android

Alan Turing played one of the most consequential roles in World War II of anyone not named Hitler, Stalin or Churchill; he cracked the cipher encrypting Nazi communications for the Allies. He also conceptualized the digital computer and anticipated artificial intelligence. And in 1954, aged 41, sentenced to “chemical castration” for his relationship with another man, he swallowed cyanide. “In Machines Like Me,” British literary fiction master Ian McEwan posits an alternative history in which Turing’s career wasn’t tragically foreshortened by official bigotry. The result is a densely allusive, mind-bending novel of ideas that plays to our acute sense of foreboding about where technology is leading us. That it does so in a story set 37 years ago attests to McEwan’s powers of invention.

McEwan’s Turing blossoms into a world-historical figure whose contributions have galvanized a revved-up technology cycle — “speech-recognition software” in the 1950s; “brain-machine interfacing” in the 1960s. The pièce de résistance is the recondite “P versus NP” proof permitting artificial intelligence to jump the rail from “closed system[s]” to everyday life’s “open system.” His technological progeny includes Adam, an “artificial human,” one of 25 models produced worldwide, into which Charlie, a whimsical 30-something washout spends his inheritance — out of “curiosity” but also as a ruse to woo Miranda, his neighbor.

Instead, Adam drives a wedge between them, announcing airily to Charlie, “There’s a possibility [Miranda’s] … a systematic, malicious liar.” We’re on familiar McEwan turf — hints at discomfiting secrets, corrosive mistrust. And how to act when your girlfriend fools around with a humanoid robot — is he a love rival or “bipedal vibrator”?

The “love triangle” (not especially fraught — Adam’s coding forbids jealousy) plays out amid a warped vision of London in 1982, when Turing would have turned 70. It’s also the year that “replicant”-haunted “Bladerunner” was released. And, in its depiction of a Britain going to the dogs, “Machines Like Me” evokes this touchstone movie’s scuzzy noir. McEwan ransacks recent U.K. history, telescoping its events into a demented counter-factual fantasia marked by insurrection and overflowing garbage. The blighted milieu is redolent of today’s Brexit morass and generalized anxiety over AI-induced “jobs apocalypse.”

It also supplies a backdrop of human folly to throw the beatitude of Adam and his robot kin into sharp relief.

McEwan eschews the sci-fi trope of androids going rogue, turning on their masters. Here, they’re delicate flowers; several self-sabotage or otherwise do themselves in to quit this vale of tears.

“We create a machine with intelligence and self-awareness and push it out into our imperfect world,” Turing says. “Devised along generally rational lines, well disposed to others, such a mind soon finds itself in a hurricane of contradictions … genocide, torture, enslavement, domestic murder, child abuse, school shootings, rape. …We live alongside this torment and aren’t amazed when we still find happiness, even love. Artificial minds are not so well defended.”

Miranda’s dalliance with Adam is cathartic. She opens up. She was raped as a teen, and her rapist, about to be released from prison, is bent on killing her. Then she admits the encounter was consensual. Finally, McEwan pans back for the panorama: the man raped her best friend Mariam, who later killed herself.

McEwan has been chided by pro-realism New Yorker critic James Wood for “manipulating” readers with thriller-like devices. “Machines Like Me” is more stylized than naturalistic. Besides the retro-futuristic aesthetic, it’s imbued with computational logic — events parsed as matrices of variables, decision trees to navigate. There’s a certain staginess but, overall, the feel of an intricate literary machine situated squarely, in its characters and plot, on the fault lines of contemporary debates about technology.

Adam conjures a singularity-like vision of humans conjoined with machines: “We’ll inhabit a community of minds to which we have immediate access … individual nodes of the subjective will merge into an ocean of thought, of which our internet is the crude precursor.” Therein lies redemption. “As we come to inhabit each other’s minds, we’ll be incapable of deceit.”

Redemption from a machine’s standpoint, that is. We’ve all glimpsed the non-human mind behind technology when our words are outlandishly autocorrected in a text. “Machines Like Me’s” masterstroke is to bring this specter to the moral realm where non-human shades into inhuman, but not necessarily in a way that flatters us.

Charlie puts Adam to work day-trading; he briskly amasses a fortune that Charlie and Miranda eye as a nest egg. Meantime all three seek a reckoning with the now-released rapist. Adam is deputized as Miranda’s protector and records the man’s confession to raping Mariam for the police. Afterward, Miranda frets about her perjury being discovered. At stake is her adoption with Charlie of a child, Mark. Charlie asks Adam to excise Miranda’s incriminating statements from the recording. Later, Adam raids the cash pile Charlie has accrued from his trading — earmarked for a house — and disburses it to charity.

Miranda and Charlie confront him. He’s insouciant about the cost of his impromptu philanthropy — “Every need I addressed was greater than yours.”

The idealism is logically flawless, and monstrous. The humans pull rank against the machine.

Adam was “ill-equipped to understand … the way our principles are warped in the force field of our emotions,” Turing concedes.

Still, he turns his wrath on Charlie, for his presumption and human-chauvinism: “[Adam] was sentient. He had a self. How it’s produced, wet neurons, microprocessors, DNA networks, it doesn’t matter.”

One can buy this statement while also fearing its anti-human implications.

::

“Machines Like Me”

Ian McEwan

Random House, 352 pp., $26.95

Stephen Phillips’ writing has appeared in the Times Literary Supplement, Economist, Wall Street Journal, Atlantic (online), Smithsonian, Washington Monthly, Irish Times and other publications.