Design Techniques for LLM-Powered Interactive Storytelling: A Case Study of the Dramamancer System
Tiffany Wang, Yuqian Sun, Yi Wang, Melissa Roemmele, John Joon Young Chung, Max Kreminski
TL;DR
The paper addresses the challenge of balancing author intent with player agency in interactive narratives by leveraging LLMs. It introduces Dramamancer, which uses author-defined story schemas and two LLM-based modules (instantiation and interpretation) to create player-driven playthroughs, supported by separate author and player interfaces. The core contributions include a schema-to-playthrough design, a prompt-driven instantiation/interpretation pipeline, and evaluation perspectives for authors and players. This work demonstrates a scalable approach to dynamic storytelling with LLMs and outlines practical evaluation criteria for future LLM-enabled narrative systems.
Abstract
The rise of Large Language Models (LLMs) has enabled a new paradigm for bridging authorial intent and player agency in interactive narrative. We consider this paradigm through the example of Dramamancer, a system that uses an LLM to transform author-created story schemas into player-driven playthroughs. This extended abstract outlines some design techniques and evaluation considerations associated with this system.