Harari’s subject is information networks in general, with AI as their latest and strangest branch. He argues that large-scale human cooperation has always turned less on the truth of shared information than on its ability to connect people — writing, print, and bureaucracy each reorganized power and social order long before anyone grasped what they’d built. AI, in his account, is the first such technology able to generate its own content and make its own decisions, rather than simply carrying decisions humans have already made.
For readers following AI, the book’s use is in this reframing: a model isn’t just a tool added to human networks, it’s a new kind of participant in them. That shift complicates familiar questions about truth, trust, and who a society’s institutions ultimately answer to.
