Human-AI Interaction in Fitness and Nutrition Mobile Applications : Human-Centered AI Principles for Better User Experience

Sammanfattning

The growing use of artificial intelligence (AI) in fitness and nutrition applications has introduced new challenges related to automation, transparency, and human oversight. This study investigates how to design a digital fitness and nutrition platform that leverages AI while maintaining human-centered values. The study builds on prior research in Human-Centered AI and Human-Centered Design, focusing particularly on user trust, adaptability, and guided decision-making. The project used a qualitative, iterative approach grounded in Action Design Research. Data were collected through semi-structured interviews with exercisers aged 18–32 and a personal trainer. These insights informed the design and prototyping of a platform that integrates AI for personalized recommendations while allowing users and professionals to adjust and oversee its outputs. Features include onboarding with body composition analysis, AI-generated workout and meal plans, and configurable levels of automation based on user experience and preference. Findings indicate that novice users benefit from structured support and motivational nudges, while experienced users prefer greater autonomy and customization. The results validate the hypothesis that a flexible AI system, when combined with meaningful human input, improves user satisfaction and engagement. The resulting artifact contributes both a practical tool for personal wellness and a generalizable framework for designing human-centered AI systems.