Marianne Aubin Le Quéré presentation

Marianne Aubin Le Quéré is a final-year Information Science PhD candidate at Cornell University advised by Mor Naaman. Marianne’s work traces how AI and other emerging technologies impact online news and civic information ecosystems. She leverages mixed-method techniques to drive insights across the Social Computing, Computational Social Science, and Communication fields. Her research has been published in general science venues and top computer science conferences including CHI, CSCW, and ICWSM. Marianne’s work has received a Top Paper award at the International Communication Association and she was named Facebook Fellowship Finalist. Before her PhD, Marianne received a BSc from Brown University in Computer Science and Non-Fiction, and worked as a Product Manager at Microsoft.

Abstract:
The societal adoption of generative AI is reshaping the way we produce, access, and understand civic information. In this talk, I will present my work tracing how AI and emerging technologies impact our local online information ecosystem. This talk draws on mixed-method approaches from social computing, computational social science, and communication to provide an overview of how technology shapes people’s exposure to local civic information through platforms, social media, and now AI-based technologies.

I first demonstrate how people use technology as a filter to interact with local civic information through studies conducted on Facebook and Reddit. These studies establish the importance of place in online information ecosystems, and how local online spaces are tied to civic participation. I then show how AI can be useful in analyzing local information ecosystems by demonstrating the application of a qualitative codebook at scale using GPT-4. This LLM-augmented content analysis reveals how community surveillance posts relate to a neighborhood’s gentrification patterns. Finally, I contrastingly also show how AI can exacerbate information inequities, through preliminary findings from an audit study of Google’s new “AI Overview” feature. This study finds that the search company relies on local news to produce high-quality AI summaries, even as these summaries may compromise the future of the very sources they rely on. I conclude with my argument that the use of AI technologies presents new opportunities for civic data analysis, but risks entrenching divisions in access to information.

Question? Contact Christine Hopper.