Few shot learning with large language models has the potential to give individuals without formal machine learning training the access to a wide range of text to text models. We consider how this applies to creative writers and present Story Centaur, a user interface for prototyping few shot models and a set of recombinable web components that deploy them. Story Centaur’s goal is to expose creative writers to few shot learning with a simple but powerful interface that lets them compose their own co-creation tools that further their own unique artistic directions. We build out several examples of such tools, and in the process probe the boundaries and issues surrounding generation with large language models.

The reader of a choose your own adventure novel and the user of a modern virtual assistant have a subtle similarity; both may, through the right lens, be viewed as engaging with a work of Interactive Fiction. This literary form emerged in the 1970s and has grown like a vine along the branch of modern technology, one guided by the advances of the other. In this work we weave together threads from the Interactive Fiction community and neural semantic parsing for dialog systems, defining the data model and necessary algorithms for a novel type of Interactive Fiction and open sourcing its accompanying authoring tool. Specifically, our work integrates retrieval based semantic parsing predicates into the branching story structures well known to the Interactive Fiction community, relaxing the relatively strict lexical options of preexisting systems.