No one in higher education will escape the urgent debate about the role of AI in teaching and learning. But what, specifically, does it mean for teaching conversation analysis? Is there anything about the peculiarities of generative AI that could actually be useful? Bogdana Huma, Nthabiseng Shongwe, Borbála Sallai, all from the Vrije Universiteit Amsterdam, take us through an intriguing experiment which tries to answer that question. Along the way, Bogdana links to an intriguing AI-generated podcast of a ROLSI article, which is a fascinating artefact in its own right, and a novel teaching resource.
In academia, the availability of generative Artificial Intelligence (genAI) through platforms such as ChatGPT, Claude, or NotebookLM has started to transform how we work, how we teach, how we learn, and even how we think. When we seek information, it is tempting to turn to these platforms, even though we know that their answers can be unreliable. Referred to as “hallucinating”, “falsifying”, or “fabricating”, making up facts is genAI’s well-known Achilles’ heel for which there is no remedy in sight.
Continue reading








