GPTeammate: A Design Fiction on the Use of Variants of the GPT Language Model as Cognitive Partners for Active Learning in Higher Education
Abstract: This paper explores the potential for variants of the GPT language model, such as ChatGPT and GPT-3, to participate as cognitive partners in active learning scenarios with higher education students. Using the ‘design-fiction’ approach to speculate on future scenarios in human-computer interaction (HCI), this paper explores the potential for a hypothetical AI application- “GPTeammate” - to participate as fully realized collaborative social entities in active learning processes and projects in graduate-level learning, design, and technology (LDT) courses. In exploring this future design, I argue that AI-inclusive pedagogical approaches like GPTeammate have the potential to create effective cognitive partnerships for active learning, particularly in supporting critical thinking and problem-solving skills. Included are a discussion of implications and directions and suggestions for future experimentation and research.
Presider: Natalie Bolton, The University of Missouri St. Louis