Book review: If our talent for sharing stories made our species into a global powerhouse, asks Yuval Noah Harari, how scared should we be of an alien entity – AI – that is getting better at telling stories than we are?
AI, in all its forms, is a hot topic, and many writers and researchers are publishing books wondering if it poses an existential threat to humanity. Harari’s latest contribution, the chunky 490-page Nexus, must rank among the worst.
Wild assumptions, truisms and hot reckons inflate his bloated tome. The underlying argument is based on his 2014 bestseller Sapiens, the publishing sensation that sold more than 25 million copies, making a star of its author.
An academic once known in Israeli university circles for his work in medieval and military history, Harari is now a darling of Silicon Valley, the guru of a modern literary genre that economist Michael Muthukrishna calls “The One Thing That Explains Everything”. In the case of Sapiens, “everything” is the mystery of human world domination, and the “one thing” is how human “social constructs” – myths and belief systems – are the glue binding large groups of humans together to farm, trade, make big cities and, yes, wreck the Earth.
“It is difficult to grasp,” writes Harari, “that the nation to which one belongs is an intersubjective entity that exists only in our collective imaginations.”
Harari’s atheism underpins much of his thinking – he often calls religion a “mass delusion”.
“In order to co-operate,” he writes, “Sapiens no longer had to know each other personally, they just had to know the same story.”
Although Sapiens’ readers responded to these ideas, the book was excoriated by experts for insisting, without evidence, that evolutionary changes in human brain structure and linguistic abilities amounted to a cognitive revolution around 70,000 BCE.
“I know of no book where expert criticism and popularity are so polar,” blasted British geneticist and science writer Dr Adam Rutherford.
In Nexus, Harari takes his thesis of shared stories further, suggesting that they led to information networks which moulded human society. And Harari really is talking about the importance of fantasy rather than fact. “What the people at the top know,” he offers, “is that telling the truth about the universe is hardly the most efficient way to produce order among large numbers of humans.”
In AI, he sees a dangerous storytelling competitor, one with the skill to sell more convincing delusions, “supercharge human conflict” and destroy human life.
Seeing animals, states, markets and human history as “information systems” is, to be fair, an interesting thought experiment. Although Harari is drawn to poetic-sounding but meaningless metaphors like “if humans aren’t careful, we might dissolve within the torrent of information like a gushing river”, many of his historic vignettes are thought-provoking, and some even help him to make a point. Misogynist Bible edits ruined women’s lives for centuries. European witch hunts amounted to 16th-century misinformation campaigns driven by a brand-new technology: Gutenberg’s printing press.
But often it is the author’s own arguments that “dissolve in a torrent of information”. Grand proclamations are made and followed up with examples that prove the opposite.
Democracy, Harari declares, “depends on information technology”, and thus, “for most of human history, large-scale democracy was simply impossible”. Without large-scale populations, it is true that “large-scale” democracy was impossible. But direct democracies, in which communities make decisions together in meeting houses and assemblies, have been known throughout recorded history. In the Western world, the concept of representative democracy – in which representatives are elected by the public – dates to the Roman Republic, a system Harari spends some time describing.
His elevated viewpoint often seems cold to the point of callousness...
When the author applies his “One Thing That Explains Everything” theory to art, he becomes seriously detached from reality. In one of the most inadvertently funny passages in the book, Harari insists that the workings of bureaucracy are suspect and obscure because human artists have evolved to prefer what he calls “biological dramas” which “copy their basic plotline from the handbook of evolution”. This is why, he says, poets and playwrights prefer sexy “boy meets girl” storylines to yarns about bureaucracy.
“The Ramayana [an important Hindu epic] is set within the context of large agrarian kingdoms but shows little interest in how such kingdoms register property, collect taxes, catalogue archives or finance wars.” If only those Sanskrit writers spent more time critiquing 7th-century BCE tax systems!
Harari concedes that bureaucracy is the central theme of classics like Kafka’s novel The Trial, and TV satires like Yes Minister, without appearing to understand that works like these do not reduce our suspicion of bureaucracy but exploit it.
Harari’s radically reductionist thesis, applied to every facet of life, needs far more evidence than he can supply. There are strange omissions – I kept waiting for him to mention language, surely a gold-plated example of a unifying human “information network”. But the only kind of language that interests him is computer language.
Harari can write succinctly about historic events he knows well. But his ambitious intellectual constructs are fragile. His elevated viewpoint often seems cold to the point of callousness, allowing for eyebrow-raising sentences like this: “Whatever conclusion we draw from the years 1939-41 [in regards to Nazi Germany and the Soviet Union], it cannot be that totalitarian networks necessarily function worse than democratic ones.”
Much of Nexus is poorly reasoned, pompous, deliberately provocative blather – a swollen balloon, in want of a needle.
NEXUS: A Brief History of Information Networks from the Stone Age to AI, by Yuval Noah Harari (Fern Press, $45), is out now.
