Everyone is arguing about whether the machines can think. Almost no one is asking the more urgent question: what are they consuming? In The Metabolism of Machines, the answer turns out to be nearly everything—electricity, groundwater, rare minerals, farmland, capital, and the intimate data of hundreds of millions of people. This is the story of an industry that has stopped building tools and started cultivating an appetite.

Drawing on a sweep of recent developments—a chipmaker committing five billion dollars to a single AI lab, retirees blockading data centers in humid Florida parking lots, lawmakers drafting a literal 'kill switch,' and a chatbot quietly ingesting your medical records—this book reframes the entire AI boom as a metabolic event. Each chapter follows a different nutrient into the maw: the power grid, the mineral supply chain, the flood of investor money, the harvest of human behavior, and finally us. What emerges is a portrait of a system that grows not by getting smarter but by getting hungrier.

With the investigative depth of the best tech-culture writing and a biologist's eye for how organisms feed, grow, and defend themselves, the book takes the industry's own metaphors at their word. If executives call their models 'living,' if they speak of 'scaling' and 'compute' as though describing a creature's diet, then let us measure the creature honestly—by what goes in and what comes out.

The Metabolism of Machines is not a doom prophecy about superintelligence. It is something stranger and closer: a natural history of an appetite we built, mistook for a mind, and are now rearranging the world to satisfy.