Since the release of Sapiens and Homo Deus, Yuval Noah Harari has gained recognition as both phenomenon and a go-to “Big Ideas” public intellectual. However, in Nexus, his new book, Harari promises a grand history of information networks, but ultimately demonstrates that big ideas aren’t always sound, especially when applied to areas beyond the author’s expertise.
Nexus promises readers “a brief history of information networks from the Stone Age to AI”, an ambitious premise that falls short in several key aspects. The book feels like two seperate works stitched together: the first part attempts to present a history of information networks, while the second, slightly more inspired half, reads like an extended discussion on artificial intelligence.
In the initial chapters, Harari defends the thesis that the primary distinction between democracies and dictatorships lies in how they manage information. Over roughly 200 pages, he offers a disjointed and haphazard historical overview, cherry-picking examples from Mesopotamian society, the rise of the Christian Church, and totalitarian regimes such as those of Hitler and Stalin to support his argument.
Unfortunately, this historical overview feels chaotic and unfocused. Harari’s writing is intellectually shalow, lacking the rigor of expert critique, and tries to cover too much. Throughout the book, the author struggles with conceptual clarity, frequently oversimplifying complex topics. The examples he provides often feel obvious and not very interesting, giving the impression that the book doesn’t respect the reader’s time or intelligence. Additionally, there are numerous counterarguments and objections to be made regarding Harari’s reasoning.
More concerning, Nexus fails to develop a coherent or accurate theory of information or networks. As a result, Harari’s central claim — that the “naive view of information” is flawed, and that more information is not always better — becomes a catch-all that constantly shifts to accomodate his various points. For a more in-depth and expert analysis of the historical relationship between networks, hierarchies, and political power, readers might turn to The Square and the Tower by Niall Ferguson instead.
In the second half of Nexus, Harari shifts to discussing the risks and policy concerns surrounding AI. While this part of the book contains more engaging insights, it still fails to add anything new to the established discourse on AI. Harari once again lacks the technical depth and conceptual clarity needed to address the subject meaningfully. In fact, Nexus offers no more insights into the risks of AI than Nick Bostrom’s Superintelligence, which was published a decade ago.
At times, Nexus reads like doomsday prophecy about AI, even though most of the many issues that Harari tackles are unrelated to artifical intelligence, let alone information networks. The so-called “Silicon Curtain” — a global division akin to the Iron Curtain that would prevent Western and Eastern computers from communicating — appears to be a (geo)political concern rather than a technological or AI-related risk, for example.
This overly pessimistic focus on artificial intelligence overlooks the real challenges posed by the technology. By stacking cliché upon cliché, Harari’s warnings about AI come off more as an apocalyptic sermon, positioning him as an oracle rather than a historian or academic.
Ultimately, Harari excels at story-telling, but even here Nexus falters. At nearly 450 pages, the book suffers from excessive verbosity, repeating itself and echoing more original authors too often. Nexus feels like an attempt to force history into a a poorly developed framework of information networks, revealing the limitations of grand historical narratives in explaining complex societal and political realities. Harari’s latest work feels incomplete and lacks the intellectual rigor that marked his earlier successes.
Leave a Reply