Fillor

Fillor is a fictional software framework described in this article to illustrate how automated content completion tools can be integrated into interactive storytelling. It is not a real product, but a conceptual model used for discussing methods of narrative augmentation.

Fillor aims to provide a modular system for filling in missing narrative branches in story states produced by interactive fiction engines, emphasizing authorial intent, stylistic consistency, and user control over generated material.

The architecture comprises several components: a fill engine that generates narrative fragments; a story state repository that tracks characters, places, and active plot threads; a policy layer that enforces pacing, tone, and safety constraints; and a plugin interface for external AI models or rule-based generators.

Algorithms and use. It may implement techniques such as constraint-based generation, retrieval-augmented generation, and coherence scoring to produce plausible continuations, supporting authoring workflows that let creators approve, edit, or veto automatically generated text.

History and reception. Conceptualized by a fictional studio, Aurorae Labs, in 2022, Fillor has appeared in design discussions as a case study of narrative augmentation rather than a deployed system. It is often cited in debates on AI-assisted storytelling and ethics.

Limitations and considerations. As a hypothetical model, Fillor highlights issues of narrative integrity, user agency, and bias in automated content. Realistic deployment would require strong governance, clear licensing, and safety controls to prevent harmful outputs.

See also related concepts include procedural generation, narrative AI, and data imputation.