Designing embodied generative artificial intelligence in mixed reality for active learning in higher education

Generative Artificial Intelligence (GenAI) technologies have introduced significant changes to higher education, but the role of Embodied GenAI Agents in Mixed Reality (MR) environments is still relatively unexplored. This study was carried out to develop an embodied GenAI system designed to facilitate active learning, self-regulated learning and enhance human-AI shared regulation in educational settings. The study also aimed to understand how adult learners engage with and perceive these anthropomorphic agents in an immersive MR setting, with a particular focus on their effects on active learning and cognitive load. Using an echeloned Design Science Research (eDSR) approach, we developed an MR learning experience incorporating an Embodied GenAI Agent. The application was demonstrated with 26 higher education learners through questionnaires and observational recordings. Our study contributes to the ongoing design and development of AI-based educational tools, with the potential to afford more active and agentic learning experiences.