arxiv:2410.18975

Unbounded: A Generative Infinite Game of Character Life Simulation

Published on Oct 24

· Submitted by

flavoredquark on Oct 25

Upvote

Authors:

Jialu Li ,

Yuanzhen Li ,

Neal Wadhwa ,

David E. Jacobs ,

Mohit Bansal ,

Nataniel Ruiz

Abstract

We introduce the concept of a generative infinite game, a video game that transcends the traditional boundaries of finite, hard-coded systems by using generative models. Inspired by James P. Carse's distinction between finite and infinite games, we leverage recent advances in generative AI to create Unbounded: a game of character life simulation that is fully encapsulated in generative models. Specifically, Unbounded draws inspiration from sandbox life simulations and allows you to interact with your autonomous virtual character in a virtual world by feeding, playing with and guiding it - with open-ended mechanics generated by an LLM, some of which can be emergent. In order to develop Unbounded, we propose technical innovations in both the LLM and visual generation domains. Specifically, we present: (1) a specialized, distilled large language model (LLM) that dynamically generates game mechanics, narratives, and character interactions in real-time, and (2) a new dynamic regional image prompt Adapter (IP-Adapter) for vision models that ensures consistent yet flexible visual generation of a character across multiple environments. We evaluate our system through both qualitative and quantitative analysis, showing significant improvements in character life simulation, user instruction following, narrative coherence, and visual consistency for both characters and the environments compared to traditional related approaches.

View arXiv page View PDF Add to collection

Community

flavoredquark

Paper author Paper submitter 3 days ago

This work defines a new class of interactive experience that we call generative infinite games, essentially video games where the game mechanics and graphics are fully subsumed by generative models .

We are making a bet that in the future most games will be entirely generated, and that generative game engines allow for things that traditional game engines do not. We know that LLMs exhibit emergent behavior, and we think this translates to games. Our first generative infinite game is called Unbounded and is a game of character life simulation. It's inspired by games like Tamagotchi and The Sims. We have scientific contributions in the language and vision domains to allow for generative LLM and vision game engines that achieve gameplay at interactive speeds.