The introduction of large language models (LLMs) and AI to social science research

The ESRA Podcast Series

In this episode, Tom Emery and Ethan Busby discuss the implications of the introduction of large language models (LLMs) and AI to social science research. Surveys on extremism, conflict and polarisation are hard to study using surveys due to high sensitivity and sample specificity, but can we use large language models (LLMs) to address these challenges? The discussion involves the following main ponts:

• What can LLMs and AI, in general, tell us about society?
• How easy was the collaboration of social and computational scientists?
• What is algorithmic fidelity? Why do social scientists using LLMs and AI in research need this?
• How can we work with such a rapidly growing tool?
• What are the potential uses of LLMs and AI within social science research? And how can it be used within survey research?

Affiliations:

Dr. Ethan Busby, Assistant Professor, Department of Political Science, Brigham Young University, Provo, UT, USA

Dr. Tom Emery – Director of ODISSEI, the Dutch National Infrastructure for Social Science; Associate Professor, Department of Public Administration and Sociology of Erasmus University, Rotterdam

Useful links:

Argyle, L. P., Busby, E. C., Fulda, N., Gubler, J. R., Rytting, C., & Wingate, D. (2023). Out of one, many: Using language models to simulate human samples. Political Analysis, 31(3), 337-351. DOI: https://doi.org/10.1017/pan.2023.2