Slice of Life - A Social Physics Game with Interactive Conversations using Symbolically Grounded LLM-Based Generative Dialogue - Paper

Aug 29, 2026

Uploaded by Liam

Slice of Life is a social physics game where players pursue social goals by selecting character interactions governed by Ensemble with Social Practices (ESP). Diverse choices create drastically different social worlds across playthroughs. The system leverages ESP’s detailed symbolic state to produce symbolically grounded prompts for a large language model, generating context-appropriate dialogue. Rather than using LLMs for novelty or cost, the approach enables nuanced conversation grounded in simulation logic while preserving deterministic gameplay and authorial control.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages11

Reading Time28 min

AccessFree to download

Introduction

Social physics game framing

Avoiding hallucinations via ESP anchoring

Role of social practices and prompt generation

Contributions and design principles

Related Work

Slice of Life: A Social Physics Game with Interactive Conversations using Symbolically Grounded LLM-Based Generative Dialogue Mike Treanor [treanor@american.edu](treanor@american.edu)[ ](treanor@american.edu)American University Washington D.C., USA Ben Samuel [bsamuel@cs.uno.edu](bsamuel@cs.uno.edu)[ ](bsamuel@cs.uno.edu)University of New Orleans New Orleans, Louisiana, USA Mark J. Nelson [mnelson@american.edu](mnelson@american.edu)[ ](mnelson@american.edu)American University Washington D.C., USA Abstract This paper describes the social physics game Slice of Life. In Slice of Life, the player strives to achieve various social goals by choosing social interactions for characters to engage in. These interactions are governed by a social simulation system called Ensemble with Social Practices (ESP) . The ways to achieve the player’s social goals are numerous and any given playthrough of the game will result in drastically different social worlds. Slice of Life also makes use of the underlying social simulation system’s detailed state to generate symbolically grounded prompts for a large language model (LLM) that generates context-appropriate character dialogue. Rather than using LLMs for novelty or for economic reasons, the underlying social simulation technology, we argue, necessitates this approach in order to make it feasible to have nuanced dialogue that reflects the many ways characters could have gotten themselves into particular social situations. The purpose of this paper is to provide a detailed account of Slice of Life’s design, how its social physics simulation enables interactive conversations based on social practices, and to illustrate how the generative possibilities of LLMs can be uniquely useful when applied as its natural language generation (NLG) system, without giving up authorial control of the gameplay or story. CCS Concepts • General and reference → Design; • Applied computing → Media arts; • Computing methodologies → Natural language generation. Keywords Social Simulation, Large Language Models, Playable Experiences ACM Reference Format: Mike Treanor, Ben Samuel, and Mark J. Nelson. 2025. Slice of Life: A Social Physics Game with Interactive Conversations using Symbolically Grounded LLM-Based Generative Dialogue. In International Conference on the Foundations of Digital Games (FDG’25), April 15–18, 2025, Graz, Austria. ACM, New York, NY, USA, 11 pages. [https://doi.org/10.1145/3723498.3723806](https://doi.org/10.1145/3723498.3723806) Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s) . FDG’25, Graz, Austria © 2025 Copyright held by the owner/author(s) . ACM ISBN /25/04 [https://doi.org/10.1145/3723498.3723806](https://doi.org/10.1145/3723498.3723806) 1 Introduction This paper presents a detailed system description and preliminary analysis of Slice of Life, a social physics game that leverages a combination of AI techniques to create a dynamic, symbolicallygrounded narrative world with fully realized character dialogue. Social physics games are characterized by player goals being achieved through interaction with the underlying simulation, where solutions are highly contingent on the complex and unique states created by the player’s choices. At its core, Slice of Life employs a social simulation system called Ensemble with Social Practices (ESP) [20, 25] to manage the simulation, and a Large Language Model (LLM) to generate context-appropriate character dialogue that is based on ESP’s simulation state. LLMs, while powerful, are prone to hallucination—generating information that may seem plausible but does not necessarily t

FAQs

How does Slice of Life let players achieve social goals?

Players select social interactions for characters, and those interactions are controlled by the Ensemble with Social Practices (ESP) social simulation system.

What role does ESP play in dialogue generation?

ESP maintains a detailed, symbolically grounded simulation state that is used to create prompts for the large language model so dialogue reflects the current simulation logic.

Why use LLMs in Slice of Life instead of purely deterministic dialogue systems?

The paper argues that LLM-based natural language generation is needed to produce nuanced, context-appropriate character dialogue while still keeping the simulation state deterministic and under authorial control.

Aug 29, 2026

Uploaded by Liam