Talking about political topics can be difficult, and it’s even harder to do this online! Researchers at BYU & Duke University wanted to see if using AI could help make these chats kinder. Learn more in our newest adapted article: sciencejournalforkids.org/articles/how-c… Lisa Argyle Ethan Busby Chris Bail (chris_bail_duke 🧵)

This is a fantastic team - excellent to work with, and providing such a cool service to K-12 education!

My paper with Patrick Liu on attitude polarization has been conditionally accepted at the APSR. In it, we develop a novel design that uses AI to target 'core issues' on the fly. We unearth key scope conditions that we believe move the literature forward bit.ly/3VAoYtp

My new piece in Harvard Business Review describes our work using AI to perform conflict mediation on social media, and how it inspired a new intervention by NextDoor which resulted in a 15% decrease in toxic content! hbr.org/2024/03/genai-…

I'm co-editing a special issue on listening in political conversations. If you might be interested in submitting an abstract in early Sept. '24 (& paper Jan '25), please contact me ASAP! Katy Coduto Molly Scudder Michael Neblo Guy Itzchakov Lisa Argyle shorturl.at/rESW3

Honestly - Same!!

Plus I have serious jealousy of her title.

Thanks for the invitation! It was a great audience, and a great chance to share (way too many of) my thoughts!

🚨 Language Model APSA Preconference🚨

Applications now open for the UPenn/Princeton LM APSA preconference (hosted by me and Nick Pangakis). Deadline is April 19. Please circulate and apply to present/discuss/attend!

arthurspirling.org/LMiSS/

We are excited to share the agenda for the IAS Seminar Series in the spring of 2024 and welcome everyone to join!

Among the speakers:
Kristian Karlson Martin Kolk Markus Strohmeier Lisa Argyle Naja Hulvej Rod Abdullah Almaatouq | عبدالله المعتوق Stephan Lewandowsky Julie Vallée

I missed this a few days ago, but thanks for highlighting our work!

For Day 7, we join Politics and Religion Journal with an #OpenAccess article from Lisa Argyle, Rochelle Terman and Matti Nelimarkka, hypothesizing that politics plays a key role in how religious practices are transformed into demarcations between “us” and “them.”

cup.org/3NjdIfP

#12DaysofReligion

Join us for the PNOC on Wednesday, November 8 at 12pm ET with Erin Rossiter & Taylor Carlson on 'To Discuss or Not to Discuss? How Selective Exposure to Political Discussion Conditions Experimental Findings on Polarization' (discussants: Ethan Busby and Lisa Argyle).

🤖 AI can improve the quality of political conversations without manipulating policy positions 🤖

Great AI-Mediated Communication study from Lisa Argyle Chris Bail (chris_bail_duke 🧵) Ethan Busby Joshua Gubler Howe, Chris Rytting Taylor Sorensen
David Wingate

doi.org/10.1073/pnas.2…

Everyone's focused on chatbots, but...
Can AI improve our difficult conversations with other people?
In new work PNASNews, we find that receiving AI suggestions improves mutual respect in divisive conversations, without influencing views. (1/n)
pnas.org/doi/10.1073/pn…

My oh my, new PNASNews !

We have LMs help foster understanding in convos between people who disagree on a divisive issue (gun control). This increased conversation quality and democratic reciprocity/tolerance, without changing policy positions!
🔗pnas.org/doi/10.1073/pn…
🧵1/7

In this #APSR FirstView , Lisa Argyle & Mike Barber introduce a new machine learning influenced method to correct for misclassification in Bayesian Improved Surname Geocoding (BISG), reducing the misclassification error by up to 50 percent.
#APSR
ow.ly/POTg50PLLRc

Working in AI is just pinging 'hey did you see they launched this' back and forth every day until you retire

If you like this paper, please make sure to check out first author Aidan Combs amazing dissertation that looks at how people infer identities through language... and for potential employers, she's on the job market!!! aidancombs.netlify.app

🚨New Polarization Lab Paper🚨How does gender shape political influence on social media? We randomly assigned female and male avatars to democrats who used a social media platform we created to let them discuss their preferred nominee for the 2020 Democratic Primary.