I am not the first to comment on Yuval Harari’s new book Nexus. A Brief History of Information Networks from the Stone Age. The reception has been mixed. I was curious to see if Harari would offer more than ‘big ideas’ and ‘big history’ that doesn’t require archival sources and close reading before drawing conclusions. In Sapiens, Harari proposed big ideas with abandon, such as gossip as the basis of the ‘cognitive revolution’ and the agricultural revolution as the ‘biggest fraud’ in history. Both may be big ideas, but they are also untrue. His generalizations paint beautiful, well-written and easy to understand anecdotes, but most of them are just that: anecdotes, mostly unconfirmed, unproven and without context.
Nexus is a disappointing read. The book could have been shorter than the 528 pages Harari expects us to read. There is much that is interesting, even brilliant. But most are anecdotes without context or truisms presented as original ideas. Harari sets out to write about information, switches to writing about information media and, in the end, about artificial intelligence, which is neither.
But perhaps these stories and truisms are at the heart of the book? Early in the book, Harari debunks the ‘naive view of information’, the idea that gathering more information leads to truth, wisdom and power. According to Harari, there is no correlation between information, truth and power. Information networks are not about truth, they are about stories. Think of religions. They are successful because they have a large information network and they tell the same story. While his point about religions may be true, information networks are not just about stories. There are proven links between control of information, control of (imagined) truth and power.
The second half of the book is about computers and artificial intelligence, which Harari believes could lead to Armageddon. Most of his concerns seem reasonable, such as generative AIs poisoning online discussions. He becomes a prophet when he discusses how ‘AI overlords’ might create a new pandemic, a new kind of money, and a flood of fake news or incitement to riot on information networks. Fake news and incitement are already true, and we don’t need advanced ‘AI overlords’ for that. And a solution? Regulation and ‘institutions with strong self-correcting mechanisms’. Answering with a truism to fight the ‘end of days’? Such a solution will not prevent a situation in which, as he writes in the prologue, ‘we might find ourselves nowhere at all, as Earth becomes inhospitable for human life’. He could have done better.
Nexus lacks focus. Anecdotes from all over space and time are juxtaposed and accompanied by truisms, such as, at the end of chapter 11: ‘The only constant in history is change’. Which is true, of course, but I did not need Harari to know that.
Perhaps I am a naive information theorist. And maybe that is why I do not recommend this book for reading.
Y.N. Harari (2024). Nexus. A Brief History of Information Networks from the Stone Age, Random House, New York, 528 pp.
Published in a post on LinkedIn, September 2024.
