Macro Musings with David Beckworth Tara Sinclair on Building a Synthetic FOMC Through AI
6 snips
Nov 10, 2025 Tara Sinclair, a professor and chair at George Washington University and former U.S. Treasury official, dives into her innovative work on simulating Federal Open Market Committee (FOMC) meetings using AI. She discusses the significance of public economic data, the potential of AI in reshaping forecasting roles, and the creation of simulated FOMC personas, reflecting real committee dynamics. Insights on institutional design experiments and the implications of AI for monetary policy are also highlighted, showcasing a future where AI informs economic strategies.
AI Snips
Chapters
Transcript
Episode notes
Data Gaps Blind Policy At Turning Points
- Missing official data can blind policymakers at potential turning points in the economy.
- Tara Sinclair warns gaps (jobs, CPI) risk policy decisions made in ignorance and hinder later research.
Treat Private Big Data As Complementary
- Use private-sector real-time data as a complement, not a substitute, for government statistics.
- Tara Sinclair advises caution because private sources lack representativeness and long historical series for turning points.
LLMs Create A FOMC 'Flight Simulator'
- LLMs can create a sandbox to run counterfactual FOMC experiments that we cannot do in the real world.
- Sinclair frames the model as a flight simulator to test meeting frequency, timing, and decision processes.