A LLM-Based Chatbot for Mindfulness Practice with Older Adults: A Development and Usability Study

Area of Science:

• Gerontology
• Human-Computer Interaction
• Artificial Intelligence

Background:

• Mindfulness interventions show promise for older adults' well-being.
• Existing digital mindfulness tools may not fully meet the needs of this demographic.
• Older adults require user-friendly and engaging digital health solutions.

Purpose of the Study:

• To develop and evaluate a Large Language Model (LLM)-based AI mindfulness coach tailored for older adults.
• To explore user-centered design principles for AI-driven health support in gerontology.
• To assess the usability and potential of an AI coach informed by qualitative user feedback.

Main Methods:

• A pre-study involved 15 older adults using a commercial mindfulness app.
• Qualitative feedback from the pre-study identified key design requirements (voice, personality, interactivity, ease of use).
• A prototype AI mindfulness coach was developed by fine-tuning OpenAI's ChatGPT on custom mindfulness scripts and evaluated via participatory design workshops.

Main Results:

• Qualitative feedback highlighted critical design elements for an effective AI mindfulness coach for older adults.
• Participatory design workshops provided insights into the usability and acceptance of the AI coach prototype.
• The study successfully integrated user feedback into the iterative development of an AI mindfulness coach.

Conclusions:

• LLM-based AI chatbots can be effectively developed to support mindfulness practices in older adults.
• User-centered design and participatory methods are crucial for creating accessible and engaging digital health tools for gerontology.
• Further research should explore the long-term efficacy and scalability of AI mindfulness coaches for older populations.