Bridging the Intelligence-Control Gap: A Case Study in Syntegrative Curriculum Design for AI Imaging

This paper examines the emerging intelligence-control gap in generative AI imaging and its implications for curriculum design. Through a comparative case study of two scaffolded modules, World Building and Storytelling, it analyses the shift from controllable open-weight diffusion pipelines to LLM-orchestrated multimodal systems. The comparison reveals a structural division: multimodal intelligence accelerates staging and narrative alignment, while controllable pipelines remain essential for continuity, editability, and print mastering. Creative agency did not correlate with system intelligence but with visual judgement and documented intervention. The subsequent redesign operationalises a syntegrative curriculum model that bridges intelligence and control through method-led practice.