AI Chat Assistants can Improve Conversations about Divisive Topics
Abstract: A rapidly increasing volume of conversation occurs online, but divi-
siveness and conflict can fester in digital interactions. Such toxicity increases
polarization and corrodes the capacity of diverse societies to cooperate in solv-
ing social problems. Scholars and civil society groups promote interventions
that make conversations less divisive or more productive, but scaling these ef-
forts to online discourse is challenging. We conduct a large-scale experiment
that demonstrates how online conversations about divisive topics can be im-
proved with artificial intelligence tools. Specifically, we employ a large lan-
guage model to make real-time, evidence-based recommendations intended to
improve participants’ perception of feeling understood. These interventions
improve reported conversation quality, reduce political divisiveness, and im-
prove the tone, without systematically changing the content of the conversation
or moving people’s policy attitudes.
0 Replies
Loading