Skip to main content
Menu

Computational Social Science Seminar: Tina Eliassi-Rad, Northeastern University

CSS Seminar announcement for Tina Eliassi-Rad, Northeastern University. The Seminar is titled "The Truth of the Matter in the Age of Generative AI" and will be held on April 16, 2026 at 12:00 - 1:00pm at the Leinweber Building, Room 4320. RSVP is requested. Sponsored by the John Seely Brown Technology and Society Lecture Gift Fund.

Lecture

Location: Leinweber Building, Dow Room (4320)
Thursday, Apr 16, 2026 Noon - 1:00 p.m.
Mode: In Person

The Truth of the Matter in the Age of Generative AI

Abstract:

We live in an age where algorithmic and human behaviors are deeply intertwined. Generative AI tools, such as ChatGPT, Claude, and Gemini, have seen widespread adoption, yet their effects on us and our influence on them remain poorly understood. These tools often contribute to epistemic instability in society. For example, generative AI tools are not experts in any field and are prone to falsehoods (also known as hallucinations) and adversarial attacks, yet they are often treated as experts. The relationship between a person and a generative AI tool is further complicated by the sense of familiarity users experience. The lack of effective oversight and accountability for this technology exacerbates these problems. How can we govern a technology that is evolving so rapidly? A digitally savvy public is an essential part of the solution.


Papers:

[1] C. Wagner, M. Strohmaier, A. Oltearnu, E. Kıcıman, N. Contractor, T. Eliassi-Rad. Measuring Algorithmically Infused Societies. Nature, 595: 197-204, 2021. https://doi.org/10.1038/s41586-021-03666-1

[2] G. Savcisens, T. Eliassi-Rad, L.K. Hansen et al. Using Sequences of Life-events to Predict Human Lives. Nature Computational Science, 4: 43–56, 2024. https://doi.org/10.1038/s43588-023-00573-5

[3] Challenging Systematic Prejudices: An Investigation into Bias Against Women and Girls in Large Language Models. UNESCO, March 2024. https://unesdoc.unesco.org/ark:/48223/pf0000388971

[4] D. Pedreschi, L. Pappalardo, E. Ferragina et al. Human-AI Coevolution. Artificial Intelligence, 339: 104244, 2025. https://doi.org/10.1016/j.artint.2024.104244

[5] G. Savcisens, T. Elaissi-Rad. The Trilemma of Truth in Large Language Models. In NeurIPS 2025 Mechanistic Interpretability Workshop, December 2025. https://arxiv.org/abs/2506.23921

[6] S. Dies, C. Maynard, G. Savcisens, T. Eliassi-Rad. Representational and Behavioral Stability of Truth in Large Language Models. arXiv, 2511.19166v3, January 2026. https://arxiv.org/abs/2511.19166 

About the CSS Seminar Series: 

The University of Michigan School of Information’s Computational Social Science seminar series brings together a vibrant and diverse community of scholars whose cutting-edge research in information science, computer science and the social sciences aims to broaden our understanding of important social and technological issues. 

The organizer for the Winter 2026 series is UMSI assistant professor Sabina Tomkins.

This seminar is sponsored by the John Seely Brown Technology and Society Lecture Gift Fund. 

Please RSVP for attendance and lunch planning.

Contact: [email protected]