A new book suggests that Alan Turing's foundational assumption for artificial intelligence might be flawed. The author argues that crucial aspects of human intelligence like common sense and intuition cannot be encoded, making true human-level AI impossible.