An Experimental Method to Study Opinion Diffusion in Human-AI Hybrid Societies

2026-05-09

AI Summary

This study investigates the mechanisms underlying opinion diffusion and polarization in hybrid human–artificial intelligence societies. By constructing a 5×5 grid-based social network, the authors conduct a controlled experiment featuring eight iterative rounds of opinion expression and revision, incorporating human participants alongside large language model (LLM) agents. The research uniquely integrates prompt engineering to modulate LLM behavior and employs a multi-condition experimental design comparing fully human, fully AI, and mixed networks. Results reveal that mixed networks exhibit the lowest levels of polarization, whereas purely human networks show heightened polarization and reduced local consensus among neighbors. Furthermore, the prompt framework demonstrates preliminary efficacy in shaping collective convergence patterns, suggesting its potential as a tool for influencing group-level opinion dynamics.

Abstract

As artificial intelligence increasingly mediates public discourse, it becomes important to understand how human-AI collectives shape opinion formation, deliberation, and democratic outcomes. We present a novel experimental method for studying opinion dynamics in hybrid human-AI social networks. Participants, human or AI, were embedded in $5\times5$ grid lattice networks and iteratively asked to select and revise statements on a given polarizing topic over eight rounds. We compared three conditions: human-only, AI-only, and hybrid networks with equal proportions of human and AI participants. Hybrid human-AI networks achieved the lowest final polarization while, in contrast, human-only networks exhibited higher polarization with lower neighbor agreement. We also ran additional experiments varying Large Language Model (LLM) prompt framing to explore whether instruction design might influence convergence patterns. Although these early findings are preliminary and cannot yet support broad generalizations, they highlight the potential value of experimental social networks for understanding opinion dynamics in human-AI hybrid societies.

Problem

• opinion diffusion
• human-AI hybrid societies
• polarization
• social networks
• collective deliberation

Innovation

• human-AI hybrid networks
• opinion diffusion
• experimental social networks
• polarization dynamics
• LLM prompting