Catalyst Seminar Series: AI and Society

The Alliance Programme is a transatlantic academic partnership that unites Columbia University with three leading French institutions: École Polytechnique, Paris 1 Panthéon-Sorbonne, and Sciences Po. Over the past year, the programme has organised a series of seminars dedicated to the theme of "AI and Society", cultivating research collaborations and advancing dialogue between scholars from both sides of the Atlantic.

On June 27th, Sciences Po, in partnership with École Polytechnique, hosted the third and final seminar of the series, titled “Exploring the Role of Large Language Models (LLMs) in Social Science Research.” This event brought together researchers from diverse disciplinary backgrounds - including sociology, political science, history, and economics - to share innovative approaches and perspectives on how LLMs are transforming research in the social sciences and humanities.

A notable aspect of the seminar was the active participation of students from the Summer Institute in Computational Social Science (SICSS Paris). These students, who attended as part of their summer school programme, contributed valuable questions and insights during the discussions, enriching the dialogue with fresh perspectives from emerging scholars in the field.

Through insightful presentations and vivid discussions, the seminar underscored the growing impact of artificial intelligence - and LLMs in particular - on the way social scientists and humanities scholars explore and interpret the complexities of the social world. The event marks a significant step forward in fostering cross-institutional partnerships at the intersection of AI and society.

The seminar featured presentations from four distinguished speakers listed below :

• Emma Bonutti D'Agostini (Sciences Po / Institut polytechnique de Paris) : Mediated Voices: A CSS Investigation into Journalists’ Portrayal of the Political Sphere
• Bart Bonikowski (NYU) : National Identification on Twitter, or How to Find a Needle in a Haystack with LLMs
• Matthew Connelly (Columbia University) : America's Top Secrets: Using AI to Decipher Official Secrecy
• Thomas Renault (Paris 1 Panthéon-Sorbonne) : Forecasting inflation with Large Language Models: A Multilingual, News-Based Approach