Multi-agent Negotiation
Date:
Imagine a world where AI agents not only assist individuals in decision-making but also assist groups in resolving conflicts and reaching agreements.
What if agents could identify common grounds in a high-stake diplomatic negotiation, and propose solutions that neither side had considered? What if they could prevent mistrust, miscommunication, or bias during negotiation?
My research explores how AI agents, specifically LLM agents, can improve collective negotiation outcomes and foster human collaboration.
We’re going to use Concordia, a social experiment framework, to identify the best collaboration strategies across different scenarios.