Old dog, new tricks? Exploring the potential functionalities of ChatGPT in supporting educational methods in social psychiatry

Abstract

Background: Artificial Intelligence is ever-expanding and large-language models are increasingly shaping teaching and learning experiences. ChatGPT is a prominent recent example of this technology and has generated much debate around the benefits and disadvantages of chatbots in educational domains. Aim: This study seeks to demonstrate the possible use-cases of ChatGPT in supporting educational methods specific to social psychiatry. Methods: Through interactions with ChatGPT 3.5, we asked this technology to list six ways in which it could aid social psychiatry teaching. Subsequently, we requested that ChatGPT perform one of the tasks it identified in its responses. Findings: ChatGPT highlighted several roles it could fulfil in educational settings, including as an information provider, a tool for debates and discussions, a facilitator of self-directed learning and a content-creator for course materials. For the latter scenario, based on another prompt, ChatGPT generated a hypothetical case vignette for a topic relevant to social psychiatry. Conclusions: Based on our experiences, ChatGPT can be an effective teaching tool, offering opportunities for active and case-based learning for students and instructors in social psychiatry. However, in their current form, chatbots have several limitations that must be considered, including misinformation and inherent biases, although these may only be temporary in nature as these technologies continue to advance. Accordingly, we argue that large-language models can support social psychiatry education with appropriate caution and encourage educators to become attuned to their potential through further detailed research in this area.