In this episode we explore what AI-generated fiction can teach us about human storytelling, student writing, and the value of literary messiness.

We look at a new study on StoryScope from researchers at the University of Maryland and Google DeepMind, which compares the narrative structure of human-written short stories with stories generated by several major AI models.

Topics covered:

• What the StoryScope study reveals about AI-generated fiction
• How human stories differ in ambiguity, time, and unresolved meaning
• Tim Requarth’s concern that AI may homogenize imagination before writing begins
• How English teachers can help students notice narrative complexity
• The risk of “closed-loop” thinking in both storytelling and education technology

Sources:

https://arxiv.org/pdf/2604.03136

https://substack.com/home/post/p-200818672