The increasing use of AI-powered running coaching applications in marathon training has raised questions about whether AI can replace the tacit knowledge of human coaches. This study explores how tacit knowledge changes as coaching shifts from human coaches to AI-powered coaching systems. An exploratory qualitative approach was adopted through semi-structured interviews with eight marathon runners and two running coaches who had experience using AI-powered running coaching applications. The interview data were analyzed using Braun and Clarke's six-phase reflexive thematic analysis. The findings show that AI is effective in supporting data driven coaching tasks. However, AI still has limitations in understanding athletes' emotional conditions, personal circumstances, and contextual factors that influence coaching decisions that remain difficult for AI to replicate. These findings suggest that future AI-powered running coaching systems should be designed to support rather than replace human coaches by considering athletes' contextual and emotional conditions, providing features that assist coaches in decision-making, and encouraging collaboration between AI and human coaches to improve marathon training.
Copyrights © 2026