Can AI simulate the Fed? - with Tara Sinclair
What happens when AI takes a seat at the world’s most powerful monetary table? Find out in this eye-opening podcast.
.jpg%3Frev%3D796e78b76223434aa7b491ad0156c72c&w=3840&q=75)
Duration: 35 Mins
Date: 09 Oct 2025
The discussion reveals how digital avatars can model real-world debates, test “what if” scenarios, and potentially transform group dynamics far beyond monetary policy.
Key highlights:
AI personas of the Fed: Professor Sinclair’s team trains AI avatars to closely mimic real Federal Open Market Committee (FOMC) members, using speeches, biographies, and district data for realism.
Simulating policy debates and ‘counterfactuals’: The research models FOMC meetings, letting these AI personas deliberate and vote on interest rates, including simulations of the July 2025 meeting with factors like political pressure and labour market shocks.
Findings on dissent: Increased political pressure led to more dissent, while revised employment data shifted voting patterns – mirroring real-life complexities.
Behavioural wedge and broader applications: The podcast explores the “behavioural wedge” – the gap between AI-driven deliberation and traditional voting models. It also examines how AI personas could transform how boards and committees evaluate complex scenarios, alongside key ethical considerations.
Labour market insights: The discussion highlights optimism about AI’s ability to improve job matching and productivity, while emphasising the importance of retaining human judgment in the process.
Listen to the latest episode of our Macro Bytes podcast and discover how AI could change the future of economics and policy setting.
