Description
Chroniqueur is an emergent narrative experimentation platform and deep social simulation engine that generates an entire human history, complete with a comprehensive archive of events. It aims to provide players with stories that arise organically from the dynamic interactions of simulated characters in a unique, ever-evolving world. Without pre-written narratives or AI-generated scripts, Chroniqueur lets stories emerge naturally, offering players the opportunity to explore or actively engage with the history of the simulated worlds. Each game generates a new world populated by thousands of characters, whose lives and actions have contextual reasons and tangible consequences, forming a rich tapestry of interwoven narratives. Chroniqueur serves as a versatile backdrop for various genres, such as life simulation, survival, management, and politics. Its standout feature is an extensive in-game documentation system—a growing “Wikipedia” of the simulation—that allows players to delve into the histories, relationships, and secrets of the simulated world.
Process:
As a developer on Chroniqueur, I contributed to the development of this complex narrative-driven world simulation through the following tasks:
- Simulation Development: Worked on the core world simulation narrative system using the Unity engine to enhance the emergent storytelling mechanics.
- LLM Integration: Developed a pipeline for integrating Large Language Models (LLMs) to analyze in-game data and augment the narrative depth.
- Portrait Creation Tool: Designed and implemented a rule-based portrait creation tool, enabling artists to generate character visuals dynamically based on predefined requirements. Added a portrait customization feature to the existing UI system for player interaction.
- Codebase Optimization: Refactored the codebase by decoupling the main simulation system from Unity's rendering system, improving code management and enabling better game optimization.