Generative AI and Interdisciplinary Transformation: Navigating the Challenges of Embodied Knowledge in Higher Education Physical Education

Abstract

The integration of Generative Artificial Intelligence (GenAI) into higher education presents a unique opportunity and challenge for disciplines centered on embodied knowledge, such as Physical Education (PE). This study examines how GenAI can enhance interdisciplinary competencies among PE instructors while navigating the inherent tension between propositional knowledge generated by AI and the embodied, tacit knowledge central to physical skill acquisition. Through a mixed-methods approach involving 328 survey respondents and 12 in-depth interviews from diverse Chinese universities, we identify a significant positive correlation between GenAI proficiency and interdisciplinary competency (r = 0.72, p < .01), yet adoption remains limited by pedagogical, ethical, and institutional barriers. Key challenges include a lack of tailored pedagogical training (71% of respondents), data privacy concerns (58%), and professional liability anxieties related to AI-assisted instruction. The study proposes a tripartite framework emphasizing structured professional development, discipline-specific pedagogical toolkits, and institutional policy reform. By situating the discussion within broader cultural and epistemological contexts, this research contributes to ongoing debates about technology-mediated transformation in embodied learning environments and offers practical pathways for aligning AI innovation with the human-centered values of physical education.