Machines Like Me by Ian McEwan

Reviewed by Anna Hollingsworth

My first reaction was a desperately deep sigh when I heard that Ian McEwan would be taking on human-like artificial intelligence as the topic for his new novel. AI is standard science fiction fodder, and human-machine relations have been written about, filmed, and otherwise imagined so many times before – from the ancient Greeks’ Hephaestus and his Golden Maidens to Channel 4’s Humans – that ‘cliché’rather than ‘exciting innovation’ springs to mind; what more could McEwan possible bring to genre? A lot, it turns out, in the form of an absolutely brilliant book: Machines Like Me dives into neural networks and buzzes in a way that is electrifyingly clever, deeply touching, and irresistibly funny with no short circuit in sight.

The setting is an alternative 1980s London, where Britain has lost the Falklands war, Margaret Thatcher fights over No 10 with Tony Benn, and Alan Turing is continuing his groundbreaking work after opting for a spell in prison instead of hormone injections. Turing is a big believer in open-access science (more than a nudge to some very heated debates in academia – researchers, take note) and the availability of his solutions to computational problems means that AI has moved forward in leaps and bounds, with the world’s first synthetic humans – twelve male “Adams” and thirteen female “Eves” – coming to the market. One of the Adams ends up in the humble abode (“two damp ground-floor rooms in the dull, no-man’s-land of Edwardian streets between Stockwell and Clapham”) of Charlie, a struck-off tax lawyer, AI enthusiast, and struggling online investor because “my hundred hours of community service convinced me that I should never have a regular job again.” Partial to extravagant buys, Charlie blasts away his recent inheritance on the deceptively human-looking machine. But the robot is there to build a human connection, too: Charlie finds himself in love with his upstairs neighbour Miranda, an enigmatic doctoral student writing a paper on 19^th century Corn Law, and Adam soon becomes their joint project. The pair share the task of picking Adam’s personality traits, but it soon turns out that there is more to the AI than pre-determined features, as Adam shows his true feelings and the trio find themselves in a love triangle like no other.

On the surface, Machines Like Me is a novel about AI, but, as clichéd as it may sound, deep down it’s all about what it means to be human. McEwan’s take on the Big Question is nowhere near a cliché, though, and delivers a sharp treatment without trying to be too clever, and is gloriously everyday without being too obvious. The philosophical ponderings take a very human form in the mundanity of Charlie and Miranda’s far-from-perfect flats, where two people and one machine have to come to understand each other as well as themselves. For example, Charlie is faced with the question of what Adam means by “I see”: can a machine see? Who is doing the seeing? When Adam expresses his feelings for Miranda, can they be dismissed as electric currents, or can they be as real as Charlie’s? At first, Adam is given the role of a domestic servant, but as he starts to recite haikus that he has written, the justification for menial tasks falters. In a poignant turn of events, Adam takes some of Charlie’s money that he has earned by working his computational magic on online trading, because he sees it as only fair that he should have his share of the profits. Charlie is left with no reason to disagree.

McEwan is a master of taking kitchen table discussions to head-spinning heights. One of the intellectually finest scenes is when Charlie, Miranda and Adam discuss what the meaning of art is, and whether it is just an incomplete way of humans trying to express the content of their minds to others. Adam argues that if AI is taken to completion and human cognition is merged into a shared network, there will be no need for art: everyone’s mind will be readily accessible to everyone else without imperfect art as an intermediary.

As serious and mind-bending as the themes are, McEwan does not abandon his trademark wry wit and laconic remarks. Without giving too much away, after a particularly climatic series of events, Miranda throws up on the carpet of her long-time nemesis:

She said later that she was helpless, out of control, but I always thought, or preferred to think, that here, at our feet as we left, was the avenging angel’s parting shot. It was tricky, stepping over it.

McEwan does not pass the ample opportunities for social commentary, either. There are obvious parallels between his alternative 80s political discussions and today’s Brexit debates, and between the alternative Thatcher and Benn and today’s May and Corbyn. It would be easy to make the mistake of overdoing the commentary, but McEwan doesn’t succumb to temptation, and instead does just enough to land some heavy and well-placed blows on today’s politics.

As much as I’d love a direct view of McEwan’s cerebral computations, I hope that Adam is wrong and that literature and art won’t become superfluous. I have faith that life won’t be reduced to neural networks, for Machines Like Me is deeply human story with such tenderness, insight and wit that I struggle to see how AI could ever recreate it.