We ask again: Is AI a science?

Thomas Kuhn’s *The Structure of Scientific Revolutions* provides a framework for evaluating the rapid evolution of artificial intelligence. While traditional scientific progress relies on community-defined paradigms and falsifiable puzzles, current AI development functions as an engineering-led endeavor driven by capital investment and scaling laws rather than pure scientific inquiry. This shift mirrors the Industrial Revolution, where technical and procedural breakthroughs fundamentally altered societal conditions without necessarily adhering to the norms of established scientific disciplines. AI’s trajectory is currently locked into a path-dependent, market-driven logic that prioritizes the rapid deployment of powerful computational tools over the systematic, objective study of reality. Consequently, the field is less a scientific revolution and more a transformative force that is actively remaking the conditions of human cognition and social interaction, potentially rendering traditional Kuhnian models of scientific progress obsolete in the face of such rapid, non-scientific disruption.