There is something almost quaint about a book that warns you the building is watching, when the building has since learned to finish your sentences. Shoshana Zuboff's *The Age of Surveillance Capitalism* — shelved here under a different title, an artifact of cataloging drift that feels oddly appropriate for a book about the slippage between what things are called and what they do — arrived in 2019 with the force of a treatise and the length of one too. It proposed a new species of capitalism, one that feeds not on labor or nature but on human experience itself, rendered as behavioral data, refined into prediction products, and sold to parties whose interests are not your own. Seven years on, the architecture she described has not been dismantled. It has been furnished.

What Zuboff got right is staggering in its specificity. The migration from virtual to real-world extraction she mapped — from search queries to smart thermostats to urban infrastructure — has accelerated past her examples into ambient AI systems that don't merely observe but generate, interpolate, and preempt. Her concept of "Big Other," a distributed computational infrastructure that shapes behavior without traditional coercion, reads now less like theory and more like product documentation. The instrumentarian society she warned of — where group pressure and computational certainty replace politics — is visible in algorithmic content curation that has demonstrably altered elections, public health outcomes, and the interior lives of adolescents. Her analysis of the "human hive" and social media's effect on the young was, in 2019, still contested. By 2026, it is the consensus position of the U.S. Surgeon General, the EU, and most parents of teenagers. She was not early. She was on time.

And yet the blind spots are instructive. Zuboff's framework is relentlessly structural, focused on the corporate architectures of Google, Facebook, and their satellites, and it treats the state as a junior partner or passive enabler. She could not have fully anticipated the degree to which governments themselves — not just in China but in the EU, India, and the United States — would become active architects of surveillance infrastructure, sometimes in competition with the platforms, sometimes in uncomfortable symbiosis. The rise of generative AI, large language models, and their integration into the prediction-and-modification pipeline she described is the most conspicuous absence. She wrote about a world that watched and nudged. We now live in a world that watches, nudges, and speaks back — fluently, persuasively, in your preferred register. The instrumentarian power she theorized has acquired a voice, and that changes the phenomenology of control in ways her framework, rooted in behaviorism and B.F. Skinner, does not fully accommodate. She also underestimated the degree to which users would not merely tolerate but actively desire the exchange — trading sanctuary for convenience with something closer to enthusiasm than resignation.