AI is often presented as a model that has become a competitor of human intelligence. But if it is characterised as an artificial way of solving problems that require intelligence from humans, there is no need for intelligence. In this article, we address this issue at two levels: on an ontological level, is the machine as intelligent as humans, or humans as humanely as a machine? at a sociological level, is the machine not being called upon to replace humans due to our modern societies, whose computational rationality confines people to machine and repetitive tasks? We argue that the machine remains ontologically questionable, unlike human beings, with the three enigms of the blade of grass, the caress and the eye marking an irreductable difference, but that the real question is to think of a modern society based on human intelligence where the machine is its complement and not its competitor.