Writing Is Coding: When Writing Becomes the Starting Point for AI Cross-Media Learning and Narrative

Writing in the digital age has transformed from a mere communicative practice into a computational interface for multimodal creation—establishing a new paradigm where natural language functions as executable code for artificial intelligence systems. This research examines the pedagogical implications of this “Writing is Coding” framework, focusing on how learners navigate cross-media AI co-creation processes. Through a comprehensive three-round experiment, 102 undergraduate students engaged in progressive AI collaboration across textual, visual, and video modalities, using natural language as their primary creative instrument. Employing grounded theory methodology, the researchers analyzed data from student reflection journals, teacher observation logs, and semi-structured interviews to identify emergent patterns in AI-human creative partnerships. Five interconnected themes emerged: (1) prompt engineering as algorithmic thinking, (2) multimodal literacy development, (3) collaborative negotiation of creative agency, (4) emotional engagement through iterative refinement, and (5) metacognitive awareness of human-AI boundaries. Findings suggest that the “Writing is Coding” framework facilitates a deeper understanding of computational processes while enhancing creative expression. The three-round progression demonstrated increasing student comfort with AI collaboration, though tensions around authorship and creative identity persisted. This research contributes to emerging pedagogical approaches that position writing as a form of programming in AI-enhanced learning environments, with implications for curriculum design and digital literacy development.