Designing a Hybrid Intelligence System for the Facilitation of Group Collaboration

This paper introduces a Hybrid Facilitation System (HFS), combining human and AI strengths to enhance remote collaboration. Hereby, tasks are distributed between a human and an AI facilitator based on their complementary skills. The study follows a design science research approach, utilizing a task allocation framework, an underlying study (with half of the teams moderated by a chatbot and others by humans), and a systematic literature review to develop a prototype HFS iteratively and evaluate it through a pilot study with three groups (n=12) collaborating on a creative task in Slack. Feedback from surveys and interviews with both end-users and facilitators demonstrated that the hybrid system fosters a trusting and efficient collaborative environment. The paper contributes a framework for task allocation in hybrid systems as well prescriptive knowledge to a nascent design theory of Hybrid Facilitation.