NarrativeLoom: Enhancing Creative Storytelling through Multi-Persona Collaborative Improvisation

2026-03-07

AI Summary

This work proposes a multi-agent AI collaboration framework for narrative generation, grounded in Campbell’s theory of “blind variation and selective retention.” Addressing the limited novelty and diversity of existing AI storytelling tools, the system employs multiple specialized large language models to produce a rich set of narrative alternatives, while the user acts as a creative director who selects and refines outputs. By integrating a computational model of creative evolution with human-in-the-loop curation and dynamically adjustable levels of AI support, the approach significantly enhances both the originality and structural complexity of generated stories. Empirical results demonstrate markedly higher user-perceived novelty and diversity, with expert evaluations showing statistically significant improvements over baseline systems across fluency, flexibility, originality, and elaboration—particularly benefiting novice writers.

Technology Category

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

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Large Language Models show promise for AI-assisted storytelling, yet current tools often generate predictable, unoriginal narratives. To address this limitation, we present NarrativeLoom, a multi-persona co-creative system grounded in Campbell's Blind Variation and Selective Retention theory. NarrativeLoom deploys specialized AI personas to generate diverse narrative options (blind variation), while users act as creative directors to select and refine them (selective retention). We designed a controlled study with 50 participants and found that stories co-authored with NarrativeLoom were not only perceived by users as more novel and diverse but were also objectively rated by experts as significantly better across all Torrance Test creativity dimensions: fluency, flexibility, originality, and elaboration. Stories are significantly longer with richer settings and more dialogue. Writing expertise emerged as a moderator: novices benefited more from structured scaffolding. This demonstrates the value of theory-informed co-creative systems and the importance of adapting them to varying user expertise.

Problem

• AI-assisted storytelling
• narrative originality
• creative co-creation
• predictable narratives
• storytelling creativity

Innovation

• multi-persona collaboration
• blind variation and selective retention
• co-creative storytelling
• large language models
• creativity enhancement