Advances in digital technology are changing the methods of teaching and learning. In the course of this stream of changes, the emergence of artificial intelligence (AI) has become a topic of great interest and concern. The development of teaching AIs (artificial intelligences that teach), which can provide adaptive and personalized dialogue with students, has shifted the relationship between teachers and students, and has raised new questions regarding the role of teachers. The purpose of this paper is to rethink the meaning (...) of the act of teaching, without denying its variety, and to reflect on the role of teachers in the relatively unfamiliar situation created by the emergence of AI teaching machines. To this end, I attempt, first, to re-examine the nature of teaching by considering two possible orientations: teaching as a return to homogeneity and teaching as exposure to heterogeneity. My purpose is to consider the characteristics and limitations of the teaching implemented by AI and to discuss the role of human teachers in overcoming these limitations. Second, I seek to reveal what it is for the teacher to speak, which is often understood as the most important aspect of teaching, again by way of two characterizations: the closed conversation, which converges on a conclusion, and the open conversation, oriented towards ethical responsibility. In doing this, I shall examine critically some characteristics of language used by AI teaching machines and suggest alternatives that might constitute openness to an ethical conversation. In furtherance of this, I shall consider the account of teaching that is central to the philosophy of Emmanuel Levinas. (shrink)