The fact that AI has not yet passed a Turing Test has not prevented it from being sold to the public as a superior kind of intelligence capable of handling vast amounts of data and therefore capable of making “evidence-based” decisions about human behavior. There is no basis for this claim. AI uses advanced statistics to fine-tune generalizations; it is a glorified actuary table, not an intelligent agent. At the time of his death in 1952, Alan Turing was exploring the differences between biological intelligence and his initial conception of AI. This paper focuses on those differences and sets limits on the uses to which current AI can legitimately be put.