Optimizing Interpersonal Communication by Simulating Audiences with Large Language Models
Lists
Only you see this
Abstract
How do we communicate with others to achieve our goals? We use our prior experience or advice from others, or construct a candidate utterance by predicting how it will be received. However, our experiences are limited and biased, and reasoning about potential outcomes can be difficult and cognitively challenging. In this paper, we explore how we can leverage current Large Language Models (LLMs) to help us communicate better. Specifically, we propose the Explore-Generate-Simulate (EGS) framework, which takes as input any scenario where an individual is communicating to an audience with a goal they want to achieve, 1) explores the solution space by first producing a diverse set of advice relevant to the scenario, 2) generates potential candidates conditioned on subsets of the advice, and 3) simulates the reactions from various audiences, selecting both the best candidate and advice to use. We evaluate the framework on eight scenarios spanning the ten fundamental processes of interpersonal communication. For each scenario, we collect a dataset of human evaluations across candidates and baselines and showcase that our framework's chosen candidate is preferred over popular baseline generation mechanisms including Chain-of-Thought. We also find that audience simulations achieve reasonably high agreement with human raters across $5$ of the $8$ scenarios. Furthermore, we demonstrate the generality of our framework by applying it to real-world scenarios described by users on web forums. Viewing LLMs as a library of shared experiences and opinions, our approach draws on this library to integrate cultural and individual experience and ultimately help people communicate better.
BibTeX
@misc{liu2024optimizing,
author = {Ryan Liu and Howard Yen and Raja Marjieh and Thomas L. Griffiths and Ranjay Krishna},
title = {{Optimizing Interpersonal Communication by Simulating Audiences with Large Language Models}},
year = {2024},
url = {https://openreview.net/forum?id=ZBEs9CJiWs},
howpublished = {OpenReview},
note = {Submitted to ICLR 2024},
}