Can Generative AI Facilitate Consensus Building?

Reactions to LLM involvement in the decision-making process of complex issues and a proposed method for evaluating these reactions

Abstract

ChatGPT, a generative AI that emerged in 2022, is being used in a wide variety of fields, including customer support, content generation, translation, and educational support, beyond the framework of a simple search engine. In this study, we will verify that ChatGPT can reduce the time required for decision making by using it in consensus building situations through gaming. In addition, we examine the bias effects of expressions on episodes, advice, and reasons for decision-making when LLMs intervene in the decision-making process for complex episodes through machine learning and evaluation by LLMs (LLMs as a judge).