Yuval Noah Harari’s latest book, Nexus: A Brief History of Information Networks from the Stone Age to AI, is a narrative treatment of history, anthropology, political theory, and artificial intelligence. It’s a continuation of his ambitious style that began with Sapiens: A Brief History of Humankind.
One of the main arguments of Nexus is that information doesn’t represent truth; it creates connections that form new realities. Harari contrasts his perspective on information with what he calls “the naïve view” that equates information with truth. Instead of communicating facts, he emphasizes the role of information in constructing inter-subjective realities—like contracts, nations, currencies—as well as inter-computer realities governed by automated systems.
Harari explores bureaucracies and algorithms as ways of organizing information and power. He also points out that power usually doesn’t result from individual action alone. Rather, it relies on networks of cooperation, which themselves depend on information and practices of maintaining it. His warnings about the dangers of centralized, algorithm-driven systems rest on these assumptions. Here, examples like social credit systems and AI’s influence on politics, make his arguments both more concrete and more pressing.
I admire Harari’s project, especially for its wide scope. I am, however, uncomfortable with how tightly he constructs a narrative around the pieces he curates in his book. This is, ironically, one of the problems he points out with organizations and traditions—The way we curate information imposes an order, rather than discovering what is actually the case.
We could also critique Harari’s omission global capitalism and populist movements in the discussion. The book raises practical questions about technology, the balance between flexibility and control in networks, the potential biases of AI, and the importance of self-correcting systems like democracy, but it doesn’t directly address these in terms of contemporary political economy.