Payne has produced an engaging and accessible guide to the development of artificial intelligence (AI) as applied to war. He shows how an initial push stalled in the 1970s; what made the difference later was the remarkable increase in computing power, the amount of data being generated on the Internet, and the way engineers came to appreciate how machines can learn. Warbots are AI-enabled platforms that can make their own decisions: in principle, they can identify targets, as well as maneuver and fire, independently. From Payne’s analysis, three main conclusions emerge. First, profound ethical issues arise once machines can decide which humans to kill, but the technology is now too varied and too far advanced to be banned. Second, AI favors the offense, owing mainly to the ability of AI-enabled weapons to swarm. Third, and perhaps most important, if AI receives a lot of data and a narrow goal, it will be tactically brilliant in ways that human commanders could never match—but AI will never be a true strategist.