Introduction：Decreased motivation is a major barrier to activity continuation in elderly life support. Flow Theory, a concept in positive psychology, describes a mental state of deep engagement and fulfillment, facilitated by a balance between challenge and skill (Challenge-Skill Balance: CS Balance). While widely studied in sports and the arts, its application to daily life support remains limited. The principal investigator developed Adjusting the Challenge-Skill Balance (ACS) in rehabilitation, showing its potential to enhance QOL. A preventive remote life support program is now planned, but conventional occupational therapy processes (e.g., assessment, trust-building, individualized planning) are challenging in remote settings. AI-driven ACS support may address these challenges.
Objective：This study examines whether generative AI can enhance remote life goal support and contribute to a practical support model.
Methods：Six participants (two per age group: 10s–20s, 30s–40s, 50s–60s) set two life goals. Challenge and skill levels were assessed using a 7-point Likert scale. Participants received ACS-based support from an occupational therapist (OT support) and AI (AI support), in a randomized crossover design. AI support used a GPT-based program. Support was evaluated using rubrics, comparing assessment discrepancies and time efficiency.
Contribution to the occupational therapy profession：AI support outperformed OT support in response speed, question quality, and overall effectiveness. It promoted autonomy and continuous engagement. Given the private nature of life support, AI-integrated ACS support may reduce psychological burden. Future research will explore hybrid AI-OT support models.