Study on Learning Experiences of Generative AI Tools in Stereoscopic Composition Courses: Usage Patterns, Impacts, and Challenges

Abstract

This study investigates the integration effects of generative artificial intelligence (GenAI) in three-dimensional composition courses and its implications for design education. Through a mixed-methods approach (N=104) combining quantitative surveys and qualitative analysis, findings reveal that AI tools significantly enhance creative efficiency (inspiration stimulation mean=3.49, solution screening mean=3.452) and optimize workflow processes (efficiency perception mean=3.356), while simultaneously exposing technical limitations in 3D model practicality (mean=3.337) and tool compatibility (mean=3.337). Students demonstrated cautiously positive attitudes (overall mean=3.2–3.5) but exhibited skill degradation anxieties (mean=3.337, SD=0.820) and divergent ethical risk perceptions (copyright concerns SD=0.878). The research identifies three core contradictions constraining AI efficacy: 1) tension between technical generality and disciplinary specificity, 2) conflict between efficiency gains and capability preservation, and 3) balance between creative freedom and ethical regulation. Students strongly advocate for structured support systems, particularly prompt engineering training (demand mean=3.49), reverse-engineering functionality optimization (mean=3.529), and integration guidance for traditional techniques (mean=3.452). The study proposes a tripartite strategy: 1) developing domain-specific AI tools with cross-platform data integration, 2) establishing a "technology-ethics-methodology" triadic curriculum module, and 3) implementing dynamic competency assessment mechanisms through blended learning to balance technological empowerment with traditional skill transmission. It warns that unchecked tool-centric pedagogy risks homogenizing design thinking and eroding critical creativity. Future research should expand cross-disciplinary comparisons and longitudinal tracking to comprehensively map AI's evolving impact on design education ecosystems.