Uploaded June 2025 | Updated September 2026, 3 weeks ago
By John Horton, Massachusetts Institute of Technology. This talk will discuss what kind of software infrastructure we would want for "normal" social scientists to take advantage of LLMs in their work for piloting, exploration/ideation and feedback. It will also discuss how to get the existing body of social science findings and data into formats useful for LLMs, and how to potentially combine LLM and human evidence in fruitful ways.
By John Horton, Massachusetts Institute of Technology. This talk will discuss what kind of software infrastructure we would want for "normal" social scientists to take advantage of LLMs in their work for piloting, exploration/ideation and feedback. It will also discuss how to get the existing body of social science findings and data into formats useful for LLMs, and how to potentially combine LLM and human evidence in fruitful ways.
