The Kate Chopin Experiment: Comparing Generative AI and Human Creative Writing Fictional Narrative Processes

Abstract

[Abstract]: This project aims to understand the underlying process by which humans and Large Language Models (LLMs) undertake complex creative writing tasks. Using a series of semi-structured interviews with ten human writers who write a specific story, and an analysis of how four generative AI LLMs ‘create’ the same story, this project seeks to explore similarities and differences in how humans research and produce creative works, and how AI LLMs generate simulacra of creative works, in order to further understand how human creative processes and AI generative production operate. The results show that there are fundamental differences in AI and human creative writing processes, and that anthropomorphic linguistic agency is falsely attributed to LLMs. Human creative processes are varied and eclectic, sometimes illogical and non-linear, but humans could use some of the techniques used by LLMs as labor-saving shortcuts to what is sometimes a laborious and time-consuming procedure.