Theatrical Language Processing: Exploring AI-Augmented Improvisational Acting and Scriptwriting with LLMs

2025-05-08

AI Summary

Traditional improvisational theater suffers from high predictability, limited interactivity, and insufficient cognitive stimulation. This paper introduces the novel paradigm of Theater Language Processing (TLP) and presents Scribble.ai—an AI-powered creative assistant designed to enhance performers’ real-time responsiveness and spontaneous creativity through the generation of unconventional improvisational scenarios. Methodologically, it integrates large language model–driven interactive scenario generation, dynamic prompt engineering, and a human-AI co-training framework, validated via a mixed qualitative study with 14 professional actors. Key contributions include: (1) the first formal definition of TLP as a domain-specific computational paradigm for performing arts; (2) empirical identification of AI-induced unpredictability as a positive catalyst for improvisational problem-solving; (3) discovery that over-specified scripts impede subtextual exploration; and (4) formulation of a design principle balancing script granularity with artistic “negative space.” Results demonstrate that AI-generated anomalous scenarios significantly expand creative boundaries and improve interpretation of unfamiliar contexts.

Technology Category

• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The increasing convergence of artificial intelligence has opened new avenues, including its emerging role in enhancing creativity. It is reshaping traditional creative practices such as actor improvisation, which often struggles with predictable patterns, limited interaction, and a lack of engaging stimuli. In this paper, we introduce a new concept, Theatrical Language Processing (TLP), and an AI-driven creativity support tool, Scribble.ai, designed to augment actors' creative expression and spontaneity through interactive practice. We conducted a user study involving tests and interviews with fourteen participants. Our findings indicate that: (1) Actors expanded their creativity when faced with AI-produced irregular scenarios; (2) The AI's unpredictability heightened their problem-solving skills, specifically in interpreting unfamiliar situations; (3) However, AI often generated excessively detailed scripts, which limited interpretive freedom and hindered subtext exploration. Based on these findings, we discuss the new potential in enhancing creative expressions in film and theater studies through an AI-driven tool.

Problem

• Enhancing actor improvisation with AI to overcome predictable patterns
• Developing AI tools for interactive scriptwriting and creative spontaneity
• Balancing AI-generated script detail with interpretive freedom for actors

Innovation

• Introducing Theatrical Language Processing (TLP) concept
• Developing AI tool Scribble.ai for interactive acting
• Enhancing creativity with AI-generated irregular scenarios