"Machines Like Me occurs in an alternative 1980s London. Britain has lost the Falklands War, Margaret Thatcher battles Tony Benn for power, and Alan Turing achieves a breakthrough in artificial intelligence. In a world not quite like this one, two lovers will be tested beyond their understanding. Charlie, drifting through life and dodging full-time employment, is in love with Miranda, a bright student who lives with a terrible secret. When Charlie comes into money, he buys Adam, one of the first batch of synthetic humans. With Miranda's assistance, he co-designs Adam's personality. This near-perfect human is beautiful, strong, and clever--a love triangle soon forms. These three beings will confront a profound moral dilemma. Ian McEwan's subversive and entertaining new novel poses fundamental questions: What makes us human? Our outward deeds or our inner lives? Could a machine understand the human heart? This provocative and thrilling tale warns against the power to invent things beyond our control"--
Leave it to the master storyteller, McEwan, to make the not quite possible, possible and believable. As always, his exquisite word choice and use of just enough detail, and his knack for making us identify with his characters, even when their flaws and questionable choices create a sense of foreboding. One of the most original sci-fi books I've read in some time on AI and in an "alternative" 1980s setting to boot - amazing.
What does it mean to be human, can artificial intelligence create art now or ever? Saving Turing from his fate in our universe and version of history and changing events while leaving so many characters in place brings in a wry commentary about what it means to be human. If art is craft and machines are built by craft and become able to replicate themselves are they producing art - and is their version of fine art going to be about something else - a communication about what it means to be a machine as our fine art is a communication about what it means to be human.
Age does not seem to weary him, however, nor the years condemn, as his latest novel is as brimming with ideas as any of its predecessors. One mark of his dexterity as a writer has been his ability to straddle literary genres, and while some of his contemporaries might have been content to offer up new iterations of formerly successful themes, McEwan has continued to experiment. His previous novel, Nutshell, was narrated by an alarmingly sentient embryo slowly developing in the womb, and took the form of a recasting of the Hamlet plot. Sweet Tooth took us to the world of counterintelligence, with a delicious twist in the tale, while, bizarrely, Solar even secured the Bollinger Everyman Wodehouse Prize for comic writing. (One does have to wonder whether that decision constituted an act of comedy in itself – Solar was certainly a fine novel, and one that was amusing in parts, but I defy anyone to maintain that it was side-splitting or uproarious.)
In this latest novel he is once again straying into unfamiliar territory, combining an exploration of the world of artificial intelligence with the construction of an infinitely plausible alternative history. The novel is set in London in 1982, but a very different world to the one that I remember. While Margaret Thatcher is still Prime Minister, she is reeling from the crushing and humiliating defeat of the Task Force that she had despatched to counter Argentina’s invasion of the territory formerly known as the Falkland Islands. The Opposition is led by Tony Benn, and Jimmy Carter is halfway through his second term as President of the United States. Alan Turing is still alive, and has been knighted in recognition of his contribution to Britain’s victory in the Second World War, but also for his work in developing artificial intelligence, and for the host of advances that he has facilitated in the world of computers. The internet is already highly developed, and there is growing concern among trade unions about the serious threat to employment caused by the widespread use of robots to undertake a range of repetitive tasks.
Charlie Friend is living in a small rented flat in Clapham, and has just used the legacy from his recently dead mother to buy Adam, one of a batch of twenty-five androids (twelve males, called Adam, and thirteen Eves) that have just been released on to the market. They are designed to be highly lifelike and largely autonomous, capable of undertaking a range of household tasks, as offering companionship. They are internet-enabled, and have been designed to learn through a combination of observing and imitating human activity, and learning from the web. Charlie has some experience of artificial intelligence, having written a book about the subject a few years ago, and he is keen to see how far Adam’s capacity to learn can be pushed. Adam’s arrival also serves to cement Charlie’s relationship with his neighbour, Miranda, a postgraduate student living in the flat upstairs.
McEwan writes with a deep plausibility. His alternative history is key as a background to the plot, but he paints it in very light strokes. For example, the politics of the day, which see Mrs Thatcher wavering in power against a resurgent Labour party led by Tony Benn (whose leadership of the party reflects many of the characteristics of Jeremy Corbyn’s rise to prominence, on the crest of a surge of younger voters who have flocked to the party, drawn by his palette of ambitious and inclusive policies) remain in the background, captured by fleeting references rather than sustained narrative.
His depiction of Adam’s development is also chilling. Once ‘awakened’, having been fully charged with power, and allocated his personality parameters (a selection that falls to Charlie, although he decides to let Miranda choose half of the settings), Adam quickly develops into am autonomous character in his own right. Programmed to be obedient, he is happy to undertake the various domestic chores that arise, but he is constantly learning, and frequently surprises Charlie and Miranda with unexpected conversational tangents.
Of course, this being McEwan, we know that mishap is near at hand, and that Fate is simply lurking around the corner, waiting to start flailing with its stuffed eel skin. In this instance, it is not the fruit of a forbidden tree but unfettered access to the internet that renders dangerous knowledge to Adam. Quite early in his development, before his personality ois fully formed, and before he has had a chance to learn the benefits of tactful reticence, he makes a chance remark to Charlie about an incident in Miranda’s past. Entirely unaware of this incident, Charlie finds himself plunged into torment and doubt, and his relationships with both Miranda and Adam are subject to strain.
As one would expect from McEwan, this is carefully plotted and thoroughly researched. The science implied by the development of Adam certainly more than satisfied my limited understanding of such things, and Adam is as triumphant a literary character as he is a technological achievement. Oddly, perhaps, my doubts about the book revolve around the character of Charlie, with whom I struggled to develop any empathy. The overall effect, however, is certainly powerful, and this is a strong addition to McEwan’s already impressive oeuvre.
The idea is fantastic and compelling, but it gets constantly bogged down in endless paragraphs about the political shenanigans of the alternate-history London McEwan has created for his characters. I skimmed pages and pages, trying to get to the meat of the story I actually cared about.
About a 2.5, for me.
Charlie is a former unsuccessful white collar criminal, living precariously from playing the markets, whilst lusting after Miranda, the much younger girl in the upstairs flat, who has secrets of her own. Having always had an interest in electronics and having written a book on Artificial Intelligence, when he inherits some money he spends it buying Adam, one of the first wave of manufactured humans. Adam has intelligence and the looks and motions to pass for human in the world, he is far more than a robot, he is a manufactured person. Whether or not he has a consciousness, or whether he has a mind as well as an intelligence, is the main narrative of the book.'
As a token of love to Miranda, Charlie allows her to set half the preferences for his "character"; in this way they become effectively Adam's co-parents, and their romantic relationship soon follows.
Adam is in most ways of course a child. He has the body of an adult - a good looking adult too, and the perfect lover, as Miranda wastes little time ascertaining, with an intelligence and learning capabilities any adult would envy. But he lacks a childhood, and the developmental experiences of a child in forming his character, and so is not an adult human. What do his expressions of love for Miranda, supported by juvenile haikkus, actually mean emotionally? By contrast, Charlie and Miranda find themselves (through a slightly convoluted plot device) becoming increasing responsible for Mark, an actual child, abandoned by his parents who they foster with an eye to later adoption. Mark expresses love for Miranda; so does Adam. So indeed does Charlie. In Charlie's tiny, cramped flat, this emotional quadrangle plays out. It would be a spoiler to go into much further detail, but Its to be hoped that Charlie and Miranda prove to be better parents to Mark, than they are to Adam
The main plot hangs around decisions that Adam makes that cause a crisis for Charlie, Miranda and Mark. These choices are in many ways logical, and unquestionably moral, but they would not be the same decisions a human would make. This,McEwan seems to suggest is the threat of non human intelligence. Not that it will in some way "take over" but that it will process information differently and come to different conclusions. This is the lesson of the AlphaGo victory over a human grandmaster at Go (an example discussed by the narrator). Whereas for a computer to win at Chess is essentially a maths problem (there are a limited number of moves possible at any stage - the computer just has to process which is best) to win at Go is an intelligence problem - the possible moves are limitless and intelligent strategy is required. AlphaGo taught itself the game, with no reference to human knowledge, and made moves that a human never would, that were none the less effective. In an AI driven future, McEwan suggests, this will happen all the time. "Correct" answers calculated in a completely different way
This is the morality question at the heart of the AI debate, and at the heart of the book. How can you teach a manufactured human to act as a human? Without a childhood.
As such this is a great narrative, more interesting that I probably, in the interests of not having spoilers, have made it sound.
Where I have some problems is with the timeline. Because this is not set in the future, or the present day, but in 1980s Britain. A slightly different 1980s Britain with not just mobile phones, and the Internet, but autonomous vehicles, and artificial intelligence. And I am not sure why this is necessary. I am even less sure why Alan Turing needs to be alive, and a character in the book as the father of Artificial Intelligence and much else (no reason why not I suppose - but why?). Other counterfactuals include Turing's collaboration with Demis Hassabis (who would have been a child in the 1980s), the reformation of The Beatles, Britain losing the Falklands War, Tony Benn as Prime Minister, and the list goes on. There is nothing wrong with such playfulness of course - and as McEwan himself rightly says on p64, "The present is the frailest of improbable constructs, any part of it, all of it, could have been different" . True enough - but it can be jarring, and I am not sure what structural purposes it serves
But if you can forget the occasional jarring, this is a very rewarding, and very well researched book. There's plenty to get out of it, because there's so much going on
McEwan ventures into the alternate history genre here to explore some of the what-ifs of the accelerating pace at which we are introducing artificial intelligence and robotics into our everyday world. The novel is set in a 1980s version of the world very different from the one recorded by the history books. Margaret Thatcher is driven from office in disgrace after badly losing the Falklands War; John Lennon is alive and well and the Beatles are still a band; and the Brighton hotel bombing this time does manage to kill a British prime minister (Thatcher’s successor). Oh, and Jimmy Carter wins a second term, John Kennedy survives his trip to Dallas, and novelist Joseph Heller finds fame with a book he titles Catch-18. You get the idea.
Charlie Friend, thirty-two years old and single, takes great pride in the fact that he doesn’t have to answer to any boss. Charlie lives alone in a London apartment where he sits in front of his computer all day long buying and selling stocks, earning just enough to cover his day-to-day needs. He is not the most ambitious guy in the world, and when he learns that what he earns from day-trading stocks is just below the wage of the average Londoner, Charlie is proud that he is doing that well without having to answer to anyone. He is not the type to worry much about his future. Now, though, Charlie is falling in love with Miranda, the student who lives in the flat above his - even though she does not seem to feel the same way about him. But after blowing all the money his recently deceased mother left him on one of the world’s first synthetic humans, Charlie may have just stumbled onto a way of binding Miranda to him. He lets her help him design the personality of Adam, the near-perfect physical specimen who will now be sharing Charlie’s flat.
Miranda, as it turns out, has secrets of her own, secrets that she can’t hide from someone like Adam who never sleeps and spends all of his spare time researching and learning about the world into which he has so suddenly been thrust. And after Adam warns Charlie that Miranda is not really who she seems to be, things begin to get tricky – especially after Adam declares his own love for Miranda.
Machines Like Me explores whether or not artificial intelligence can ever understand human emotions, motivations, and reasoning. Will it be possible for such a created consciousness to grow beyond the black and white rules it has initially been designed to follow? And if not, how will the inevitable conflict be resolved? What is to be done when our synthetic humans decide that they know what’s good for us better than we do. Which of us crosses the line first?
This quote (page 370 of the Large Print edition) should give all of us, researchers included, something to think about: “They couldn’t understand us, because we couldn’t understand ourselves. Their learning programs couldn’t accommodate us. If we didn’t know our own minds, how could we design theirs and expect them to be happy alongside us?”
Bottom Line: Machines Like Me is a bit frustrating at times because of the long, detailed digressions that McEwan strays into that do not always do much to advance the “discussion” of the potential conflict between artificial intelligence and human intelligence - but the patient reader will be well rewarded for his patience. I suppose that Machines Like Me will be most easily appreciated by science fiction and alternate reality fans, but it is a thought provoking philosophical novel as well.