Abstract

In this paper we discuss the notion of interaction between humans and robots. For that we place the Turing test in an epistemological argument of artificial intelligence, as well as with respect to recent announcements claiming that it was passed. We argue that this claim is false, especially with respect to the difficulty of talking (in the sense of communicating) with robots. Thus, we raise some interrogations on what would intelligence be for robots. We return to the concept of behaviour portraying similarities to humans or complementary to human faculties. So if robots have already invaded our environment, the projection of humanity both in their skills and in their appearance is still a theoretical investigation.