『Life 3.0』要約

In Life 3.0, MIT physicist Max Tegmark examines how artificial intelligence may reshape humanity’s future, framing the discussion through a simple but striking definition of life’s stages. He calls biological life “Life 2.0”: organisms whose hardware and software are both shaped by evolution, so humans can learn but cannot redesign their bodies or minds. “Life 3.0,” by contrast, would be able to redesign both its hardware and its software, making it the first life form capable of fully controlling its own destiny. Tegmark uses this framework to organize a wide-ranging exploration of possible AI futures, from a benevolent “enlightened” superintelligence to dystopian outcomes such as an oppressive singleton or humanity’s marginalization. The book moves from near-term concerns like automation, lethal autonomous weapons, and the “robustness” problem of aligning AI with human values, to far-future questions about consciousness, meaning, and cosmic expansion. Tegmark introduces scenarios and thought experiments, including the “Omega Team” vision of an AI that gradually takes over more tasks, and a detailed “Prologue” set in a future where an AI system named Prometheus helps a group achieve astonishing control over matter and energy. He also explores the “coherent extrapolated volition” idea from philosopher Eliezer Yudkowsky and discusses the “orthogonality thesis,” the claim that intelligence and final goals need not correlate. Throughout, Tegmark argues that the most important question is not whether superintelligence will arrive, but what values it will pursue, and who decides. He urges researchers, policymakers, and the public to treat AI safety as a serious, urgent project rather than a science-fiction fantasy. The result is less a prediction than a map of choices, intended to help readers think clearly about the stakes of creating minds that might one day surpass our own.