Simulating Persuasive Dialogues on Meat Reduction with Generative Agents

2025-04-07

AI Summary

Reducing meat consumption faces resistance from entrenched social norms—such as the centrality of meat in shared meals—and constraints imposed by social costs. Method: This study introduces a low-cost, high-ecological-validity approach to exploring persuasive communication strategies, innovatively deploying large language model–driven generative agents to simulate multi-turn persuasion dialogues. Grounded in the Theory of Planned Behavior (TPB) and integrated with a social cost assessment framework, the method concurrently evaluates psychological acceptability and real-world feasibility. It enables population-specific strategy customization and validates efficacy through strong alignment between simulation outputs and existing empirical data. Contribution/Results: The paradigm overcomes key limitations of traditional experimental designs—namely, scalability, cost-efficiency, and contextual authenticity—offering an extensible methodological toolkit for dietary behavior interventions.

Technology Category

• Multiagent Systems: Agent Communication
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Application Domains: Humanities & Computational Social Science

Application Category

• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Meat reduction benefits human and planetary health, but social norms keep meat central in shared meals. To date, the development of communication strategies that promote meat reduction while minimizing social costs has required the costly involvement of human participants at each stage of the process. We present work in progress on simulating multi-round dialogues on meat reduction between Generative Agents based on large language models (LLMs). We measure our main outcome using established psychological questionnaires based on the Theory of Planned Behavior and additionally investigate Social Costs. We find evidence that our preliminary simulations produce outcomes that are (i) consistent with theoretical expectations; and (ii) valid when compared to data from previous studies with human participants. Generative agent-based models are a promising tool for identifying novel communication strategies on meat reduction-tailored to highly specific participant groups-to then be tested in subsequent studies with human participants.

Problem

• Simulating dialogues to promote meat reduction strategies
• Reducing social costs in meat reduction communication
• Validating generative agent outcomes against human data

Innovation

• Simulating dialogues with Generative Agents
• Using LLMs for meat reduction strategies
• Validating outcomes with psychological questionnaires