Introduction

I am a philosopher and cognitive scientist working on AI. I am a co-founder of Sophron Research, a funded non-profit developing evaluations for models and their ability to help us make better decisions.

Background

I received my PhD from the University of Michigan in 2026. My research focused on how to explain motivation using empirical and computational methods, and how to understand the relation between concepts like desires, preferences, and values, and their relation to rational behavior. I also have interests in consciousness and decision theory.

I’m a mentor for SPAR, a research affiliate with the MINT lab, and I have previously been an AI for Human Reasoning Fellow with the Future of Life Foundation. My research has been supported by the Cosmos Institute, Longview Philanthropy, Open Philanthropy, the Institute for Humane Studies, and the Clarendon Foundation. Before Michigan, I completed a BPhil in Philosophy and a BA in Philosophy, Politics, and Economics, both at Oxford. Before that, I grew up in a town known for having the world’s largest straw goat.

Interests

When I’m not thinking about my primary work on evaluating AI models, I tend to get curious about other things. Some questions that I’m currently thinking about include:

• How do our preferences change? Can an agent rationally bring about such changes?
• What is the relation between reward in reinforcement learning and preferences in economic decision theory?
• How do incommensurable values feature in decision-making? Do they?
• What are the necessary conditions of consciousness? How can we design tests for consciousness that are informative across substrates?
• What would it mean for the world to be made out of computations? Might it be?

Aside from my academic research, I (aspirationally) write on my Substack, The Internal World. You can also find me on X @PReaulx or reach me at pauldfr@umich.edu.