Jeremy FooteMar 16

New paper!

Conversations with AI chatbots have been shown to be persuasive across a bunch of contexts. However, most of these studies are short-term and track only changes in attitudes or beliefs.

We wanted to look at the long-term impact of chatbot conversations on behaviors, so we built the Reddit Conversation Library.

Show threadJeremy Foote

This is a set of software tools designed to do conversational AI field experiments. It recruits and consents Reddit users, assigns them to conditions, manages and store their conversations with chatbots, and tracks their public behavior before and after the intervention.

This allows researchers to track longer-term changes to behavior in a real-life setting.

Mar 16 at 9:10pmWeb
Show threadJeremy FooteMar 16

The paper describing the tool and making an argument for its benefits (and delineating some of the risks and drawbacks to this approach) is out in Computational Communication Research:

https://www.aup-online.com/content/journals/10.5117/CCR2026.2.5.FOOT

The open source GitHub project is at:

https://github.com/jdfoote/Reddit-Conversation-Laboratory

We (me, Loizos Bitsikokos, Hitesh Goel, and Deepak Kumar) hope that this paper and library make it much easier for others to pursue this kind of research. Contributions and pull requests to the project are very welcome!

Reddit Conversation Laboratory: Field experiments with conversational AI agents | Amsterdam University Press Journals Online

LLM-based generative AI agents are the first autonomous technologies that can act as true conversational partners. Communication researchers and others have already begun to explore the influence of AI conversations on their interlocutors. We present a methodological framework and software tool for conducting field experiments with AI agents on Reddit. The Reddit Conversational Laboratory is Python-based software that identifies potential participants, messages and consents them, and conducts conversational experiments with researcher-designed AI chatbots. In addition to storing all conversations, the software can also record participant behavior before and after conversations. In this paper, we outline design principles, best practices for using the tool, and possibilities for future extensions.