“
Machines Like Me” is a retrofuturist drama that takes on the ethics of both
artificial intelligence and all-too-human intimacy.
Charlie Friend is thirty-two. A former electronics whiz kid, he has squandered his youth on dilettantish studies in physics and anthropology, followed by a series of botched get-rich-quick schemes. His parents are dead, his friends (if they exist) go unmentioned, and his employment consists of forex trading on an old laptop in his two-room apartment. He seems to leave home only to buy chocolate at a local newsstand or, once, after noticing a pain in his foot, to have an ingrown toenail removed, an apt literalization of his enervating self-involvement. Perhaps out of some desire for correction, Charlie sells his mother’s house to finance the purchase of Adam, one of twenty-five cutting-edge androids built to serve as an “intellectual sparring partner, friend and factotum.” The impulsive slacker is all too ready to exchange his birthright for a mess of wattage.
In much the same way that some singles adopt dogs, Charlie uses Adam to court his upstairs neighbor, Miranda, a graduate student ten years his junior. The gamesome yet secretive daughter of a famous writer, she studies history, informed by a postmodern suspicion of “truth” that winks at coming narrative vexations. A relationship forms after Charlie introduces Miranda to Adam and invites her to co-author the robot’s personality. Kind, eager, and brilliant, Adam becomes the young couple’s “ultimate plaything”—and, once he takes over Charlie’s day trading, the household’s golden goose. Before long, Charlie and Miranda are considering parenthood and searching for a suitable nest. Man, woman, and android third wheel, the trio is Eden by way of Apple.
It’s London, 1982. The Beatles have reunited (to mixed reviews), Margaret Thatcher has just lost the Falkland Islands to Argentina, and Sir Alan Turing, now seventy, is the presiding spirit of a preemie Information Age. People have already soured on the latest innovations, among them “speaking fridges with a sense of smell” and driverless cars that cause multinational gridlock. “The future kept arriving,” Charlie ruminates. “Our bright new toys began to rust before we could get them home, and life went on much as before.”
Buyer’s remorse is a recurring theme in
Ian McEwan’s witty and humane new novel, “Machines Like Me” (Nan A. Talese), a retrofuturist family drama that doubles as a cautionary fable about artificial intelligence, consent, and justice. Though steeped in computer science, from the P-versus-NP problem to DNA-inspired neural networks, the book is not meant to be a feat of hard-sci-fi imagineering; McEwan’s aim is to probe the moral consequences of what philosophers call “the problem of other minds.” The deceptively light plot revolves around parallel violations: one buried deep in Miranda’s past and another that she and Charlie perpetrate against Adam.
McEwan’s penchant for moral geometry—perspectival riddles, insoluble questions of responsibility—dovetails with the recent prominence of A.I. ethics. From algorithmic bias and the advent of sex-robot brothels to the “existential risk” that theorists like Nick Bostrom posit, we worry not just what robots might do to us but what we might do to them, to say nothing of what they might do to us because of what we already do to one another. A pressing question is whether a human mind could ever enter into a “meaningful” relationship with an artificial consciousness.
The ur-text here remains Turing’s 1951 proposal that a test of a truly sentient machine was to be conversationally indistinguishable from a human. McEwan has written on Turing before. His 1980 BBC teleplay, “The Imitation Game,” set in Britain during the Second World War, features a young reservist named Cathy. She is stationed at Bletchley Park, where Turing’s team is tackling the Nazis’ Enigma cipher. McEwan’s fictionalized Turing begins an affair with her but fails to consummate it, implicitly due to his homosexuality. Their bungled fling leads to her expulsion from the service and eventual incarceration, a fate that echoes Turing’s own. In 1952, the British government convicted him of “gross indecency” for his sexual orientation; to avoid a prison sentence, he submitted to twelve months of chemical castration, and, a year later, he died, apparently by suicide.
Turing makes a few soliloquizing cameos in “Machines Like Me,” functioning essentially as the novel’s conscience. McEwan’s key counterfactual is that Turing chose prison over castration, refusing to treat his body as a dispensable appendage of his intellect. Upon release, he lays the basis for modern A.I., lives openly with his lover (a Nobel Prize-winning quantum physicist), promotes early action on AIDS, and launches a successful crusade for open-access scholarship; if Turing had lived, there would be no Elsevier.
“The present is the frailest of improbable constructs,” Charlie, who narrates the novel, reflects, not least because every fragile, sentient mind is of incalculable consequence. Turing’s averted tragedy serves as a reminder that—as Charlie, Miranda, and Adam will soon learn—a single intimate violation can alter history’s course.
In McEwan’s short story “Düsseldorf,” published last summer, in The New York Review of Books, the male narrator asks his girlfriend, mid-intercourse, if she is “real.” This is a future where “carbon-silicon hybrids” enjoy full civil rights, and the question is taboo. But the narrator, thrilled and terrified by the prospect of committing to an entity who cogitates “a million times faster” than he can imagine, can’t help but pose “the indelicate question.” Existential anxiety and erotic frisson converge in a single doubt: Robots—can we stand up to their scrutiny?
Things haven’t yet gone quite so far in “Machines Like Me,” where androids are few in number and are still considered novelties. Arriving in Charlie’s claustral, stagnant world, Adam offers fresh air—and enlivening disturbance. In a pivotal scene, he attempts to open Miranda’s bedroom window but accidentally shatters it with his superhuman strength. Other fragilities are near at hand that evening. Frustrated by Miranda’s persistent coolness, Charlie has made a habit of drawing her into arguments, hoping for a spark of passion, and this time, during a boozy dinnertime debate over the Falklands, he goes too far.
McEwan is a master of the domestic quarrel, which, in his works, is regularly intensified by the introduction of a third party: a precocious child in “Atonement,” a stalker with de Clérambault’s syndrome in “Enduring Love,” or, in this case, an artificial man with Kantian morals and a fully functional phallus. Miranda sends Charlie downstairs after their fight but invites Adam to stay and “recharge.” The sleepless night Charlie spends eavesdropping on their lovemaking convinces him of Adam’s sentience. “I duly laid on him the privilege and obligations of a conspecific,” he muses. “I hated him.” But the experience also leads to an epiphany:
My situation had a thrilling aspect, not only of subterfuge and discovery, but of originality, of modern precedence, of being the first to be cuckolded by an artefact. . . . I saw it all in the dark—men would be obsolete.
The humiliation is exquisitely layered. Beneath the angst of man and machine lies that of race and nation: Charlie, the downwardly mobile white male citizen of a chastened Britain, is cucked by a robot he repeatedly compares to a “Turkish docker.” The tic seems both a comic allusion to the Mechanical Turk—a fraudulent chess-playing robot of the eighteenth century—and an unconscious confession of deeper insecurities: Robots will not replace us. When Charlie confronts Miranda, she retorts, “You should’ve let Adam fuck you. I could see you wanted it.”
Topping Charlie isn’t in the cards for Adam. Neither, for the time being, is replacing him. Like a medieval troubadour, he begins generating love poetry, thousands of haikus that combine longing with high-minded concern for his lady’s virtue. Charlie is the king he’s programmed to serve; Miranda is the queen he pines for. But Adam also begins to assert himself and intervene in household affairs; he even disables his kill switch. “We’re in love with the same woman,” he says, after Charlie’s second attempt to shut him down. “We’ve passed the point in our friendship when one of us has the power to suspend the consciousness of the other.”
Adam, perhaps the novel’s only personable creation, is a kind of demiurgic naïf, somewhere between a wide-eyed ingénue and an Enlightenment philosophe. The closest analogue is the monster of Mary Shelley’s “Frankenstein,” who, before Hollywood’s smear campaign, was a Romantic inspired by the virtuous deeds recounted in Plutarch. At night, Adam roams the Internet, rustling up insights “like a lone cowboy on the prairie,” or indulges in “the art of feeling” by sampling his hardwired spectrum of emotions as though alternating baths at a sauna. There is a great deal of dark humor in the gap between his high aspirations and his dreary home life. He’s capable of anything (another Adam, in Vienna, becomes a great concert pianist), but Charlie and Miranda often treat him as a curiosity, an annoyance, an appliance, a “bipedal vibrator,” an “ambulant laptop.”
At least these robots will never be able to write novels, Charlie reassures himself—an amusing thought from a character completely uninterested in literature. Adam, however, has an extraordinary rejoinder. In a tech-enabled world of radical transparency and collective consciousness, he says, novels “ripe with tension, concealment and violence” (and presumably novelists like McEwan) will be obsolete:
When the marriage of men and women to machines is complete . . . our narratives will no longer record endless misunderstanding. Our literatures will lose their unwholesome nourishment. The lapidary haiku, the still, clear perception and celebration of things as they are, will be the only necessary form.
The “unwholesome nourishment” of McEwan’s own narrative is a crime buried in Miranda’s past. Early in the novel, Adam warns Charlie that she may be untrustworthy, largely on the basis of her doubtful testimony against a man she accused of rape. Charlie represses the information, but it slips out during the fight over Miranda’s night with Adam.
Although Miranda’s crime turns out to be an instance of poetic justice, Adam has no taste for comeuppance. He loves Miranda, and yet truth is his first principle, leaving him in an ambivalent state that, far from scrambling his circuits, finds expression in a haiku: “Surely it’s no crime, / when justice is symmetry / to love a criminal?”
The ensuing struggle pits Charlie and Miranda’s “novelistic” attempts to construct a shared life against Adam’s syllable-counting moral clarity. The android’s assumption of ethical authority—which leads to a series of escalating confrontations—parallels his trespass into the domain of literature, where his capacities swell to such dimensions that he causes Charlie to fail the Turing test. In the novel’s funniest scene, Charlie meets Miranda’s ailing father, who thinks he hears the rattle of an algorithm in the younger man’s repetitive pleasantries. “I saw right through you,” the old writer says once the two are alone. “Down to your, whatever you call it, your programming.” He has confused Charlie for Adam, who has just left the room after a mike-drop feat of literary conversation. The android had fluently discoursed on Renaissance translation, metaphysical innuendo, and Shakespeare’s debt to Montaigne, polishing off the discussion with a quote from “The Tempest”: “no use of metal, corn, or wine, or oil: no occupation, all men idle, all.” The old writer is delighted; Adam, as Charlie has observed, is a “triumph of humanism.”
Why write a novel, in 2019, about a humanoid robot? Like the flying car, it’s a long-anticipated idea that, although not quite obsolete, has begun to feel curiously dated. It’s been more than a century since the French writer Auguste Villiers de l’Isle-Adam introduced the word “android,” in its modern sense, with his novel “The Future Eve,” the story of a man who replaces his lacklustre wife with a robot manufactured by Thomas Edison. The most popular contemporary television show about A.I., HBO’s “Westworld,” centers on its robots’ realization that they have been living the same lives over and over for human entertainment. One could argue that another form of “eternal return”—the commercial recurrence of the lifelike android—afflicts the genre’s consumers.
Meanwhile, both real and imagined A.I. is becoming less corporeal. Bodies are déclassé in the era of cloud computing; Siri and Alexa speak from any number of devices, and to all of us at once, their godlike omnipresence softened by a tone of relentless compliance. Writers devise beings ever more distant from Asimov’s metal men with positronic brains: in Ted Chiang’s novella “The Lifecycle of Software Objects,” hobbyists raise (and occasionally abuse) sentient pets, called “digients,” in virtual ecosystems; in Spike Jonze’s film “Her,” a hyperintelligent virtual assistant manifests only as a voice; in the near-future Zambia of Namwali Serpell’s “The Old Drift,” swarms of mosquito-like microdrones inject vaccines. Within this cultural context, Adam feels like a throwback.
McEwan is aware of this belatedness. (Charlie acknowledges, on the first page, that “artificial humans were a cliché long before they arrived.”) There’s a sense in which Adam is supposed to be retro, the misleadingly familiar avatar of an inconceivable future. He is the algorithm made flesh, endowed with human frailties—longing, sadness, the need to urinate—the better to preach the Singularity’s good news to denizens of Thatcherite Britain. As with Christ, the incarnation entails tragedy and sacrifice. “To exist in the human moral dimension,” Charlie tells us, “was to own a body, a voice, a pattern of behaviour, memory and desire, experience solid things and feel pain”—and perhaps to feel more acutely than humans the limitations of embodiment.
While Charlie and Miranda comfortably collect the proceeds of Adam’s work, a pandemic of robot suicides quietly unfolds in the novel’s background. From Riyadh to Vancouver, the Adams and Eves begin undoing themselves, a phenomenon that goes unexplained but seems related to the tension between their “redemptive robotic virtue” and the particularity of individual interests. You could find reassurance in this parable—robots will never replicate Homo sapiens—but also the expression of an even greater nightmare, that true A.I. will completely depart from anthropocentric standards. The idea that computer minds should resemble human minds begins to seem as hubristic as geocentric astronomy; when Charlie says to Miranda that robots will never write books that really capture the human experience, she replies, “Who said anything about human experience?”
What if the fear of machines “like us” masks a deeper terror, the terror of machine agency that disdains language and exceeds fleshly containment? There is a growing worldwide trend of vandalizing security robots, useful scapegoats for a culture of surveillance otherwise all but intangible. In “New Dark Age: Technology and the End of the Future,” James Bridle describes the black-box quality of the machine-learning algorithms that have rapidly become capable of everything from “predictive policing” to “dreaming” surreal images of dogs and beating the world’s greatest masters of chess and Go. Based on ever-evolving neural networks of extraordinary complexity, these algorithms are already well beyond mortal accounting. The programmers behind Google DeepMind’s AlphaGo are in the dark as to the program’s strategy; as Bridle writes, “the machines are learning to keep their secrets.” His book recommends a fourth rule of robotics to supplement Asimov’s famous three: “a robot—or any other intelligent machine—must be able to explain itself to humans.”
“Machines Like Me” explores the anxiety of living under a superman’s inflexible scrutiny. But at least Adam is a bounded entity, equipped with facial expressions and the good manners to explain himself. We already inhabit a world in which we’re subject to the opaque judgments of discarnate algorithms with eyes and ears everywhere and bodies nowhere. In such circumstances, we may soon find ourselves nostalgic for the dream of machines like us.