Game Development as Human-LLM Interaction

TL;DR

This paper presents ChatGE, an LLM-powered game engine enabling users to develop custom games through natural language interaction, simplifying game development for non-experts.

Contribution

It introduces a novel framework for human-LLM interaction in game development, including a data synthesis pipeline and a three-stage training strategy for ChatGE.

Findings

Effective interaction quality demonstrated in poker game case study

High code correctness in generated game scripts

Successful transfer of dialogue-based LLM to game engine

Abstract

Game development is a highly specialized task that relies on a complex game engine powered by complex programming languages, preventing many gaming enthusiasts from handling it. This paper introduces the Chat Game Engine (ChatGE) powered by LLM, which allows everyone to develop a custom game using natural language through Human-LLM interaction. To enable an LLM to function as a ChatGE, we instruct it to perform the following processes in each turn: (1) $P_{script}$: configure the game script segment based on the user's input; (2) $P_{code}$: generate the corresponding code snippet based on the game script segment; (3) $P_{utter}$: interact with the user, including guidance and feedback. We propose a data synthesis pipeline based on LLM to generate game script-code pairs and interactions from a few manually crafted seed data. We propose a three-stage progressive training strategy to transfer the dialogue-based LLM to our ChatGE smoothly. We construct a ChatGE for poker games as a case study and comprehensively evaluate it from two perspectives: interaction quality and code correctness.