The article proposes that deep-learning technologies have the capacity to produce artificial super-intelligence, which makes possible one of Dworkin’s most controversial and tantalising theories: that there is in fact one right answer to any legal question.