At a faculty retreat a few weeks ago, a colleague showed the video of Sal Kahn, in a TED talk, in which artificial intelligence (AI) advised students on a wide range of topics, answering a variety of questions. It was puzzling to me. If AI can conduct a Socratic conversation on any topic, with the participation of the other party, what are we left with as teachers who practice the case method? Are we easily replaceable by chatbots?

To answer these questions, I reflected on my own experience as an observer of the case method in action. Classroom observation, as Professor Jim Austin states, has benefits both for the observed who receives useful feedback, and for the observer, in being able to appreciate other styles. One of the suggestions I made most often in these observations was that the instructor should break the dialogue with the student and promote "multi-dialogues" among class members. In such discussions, one student told me, "I discovered what I thought about the subject."

Why don't these debates between students happen more often? It is possible that for some, disagreements are uncomfortable. There are more and less confrontational cultures. And there is a risk that a difference of opinion can turn into an exchange of insults. But the ability to understand different points of view without necessarily accepting them, to "put yourself in the other's shoes" as recommended in principled negotiation, is an essential management skill.

This skill cannot always be acquired in the Socratic-style question-and-answer dynamic between teacher and disciple (see "The Case Method is not Socratic," No. 20 of this blog). Both methods- Socratic and Case Discussion- may begin with an open-ended question, such as "What is the problem?" or "What does justice mean?" And in both, the goal is to resolve a dilemma, although in one version known as "elenchus" (Greek for "irony"), it sought to reveal the lack of intelligence of the wise. This version did not please the wise men and was what ultimately led to Socrates' death.

We can identify this version in the teaching style of some professors, and I do not recommend it for case discussion-even if the consequence is less fatal. Rather, I suggest relying more on the thinking of Aristotle, who did not seek to set traps to reveal the ignorance of the wise, but appreciated the diversity of disciplines: science for physics and mathematics, and logic for critical thinking and good judgment. The latter was based on three criteria: deep analysis of facts, maturity, and the ability to explain one's reasoning to others in a convincing manner.

Aristotle's criteria, although not very precise, can serve as guidelines for teaching good judgment among human persons, beyond the capabilities of the IA. In case method teaching, it translates into a style that can be called guided discovery. It is neither open-ended, aimless, nor is it a minefield of planned tricks. There are learning objectives, but there is no single truth. There can be surprises and discoveries, both for the students and the instructor.