I am a Fellow of the Openmind Research Institute and an Adjunct Professor of Computing Science at the University of Alberta.

My goal is to understand the computational principles behind both natural and artificial minds. I want to bring definition to the core phenomena we associate with intelligence. This means explaining how agents solve complex problems, acquire and represent knowledge, interact with their surroundings, and learn from their experiences. I draw inspiration from my peers across computer science, the cognitive studies, and philosophy—to develop new algorithms and mathematical theories with meaningful empirical impact.

Much of my research frames quesitons within the reinforcement learning paradigm. I am currently thinking about goal specification, theoretically-sound algorithms for translating preferences to rewards, as well as how computation imposes limits on interacton and agency. Over the years I have studied a broad range of topics, including meta-learning methods for efficient search, flexible architectures for organizing unstructured sense data, strategies for risk mitigation using distributional RL, as well as various applied problems in chip design, robotics, and aviation. Long ago I helped build the world’s first full-scale autonomous helicopter.

NEWS

March, 2025: Steering committee for the second edition of Finding the Frame.
March, 2025: Talk at Openmind workshop in Edmonton, Alberta.
January, 2025: New paper led by Fatima Davelouis: [pdf].
January, 2025: Co-organizing an RLDM workshop: Saving the Phenomena of Minds.
December, 2024: Spoke about embodiment at the Openmind retreat in Singapore.
December, 2024: Started as a Research Fellow at Openmind.
October, 2024: Talk at the Openmind retreat in Banff, Alberta.
October, 2024: Seattle Minds and Machines Talk: The Methodological Tangle of AI Research.
August, 2024: UofA Teatime Talk: The Methodological Tangle of AI Research.
August, 2024: Attending RLC in Amherst, Massachusetts.
July, 2024: New paper on Meta-Gradient Search Control: [pdf].
June, 2024: Talk at Cohere for AI.
March, 2024: Co-organizing an RLC Workshop: Finding the Frame.
July, 2023: Attending ICML in Honolulu to present Settling the Reward Hypothesis.
May, 2023: Attending Upperbound in Edmonton, Canada.
April, 2023: Settling the Reward Hypothesis, accepted to ICML 2023 as an oral!
March, 2023: Two ICLR workshop papers, led by Rafael Rafailov: [pdf-1], [pdf-2].
February, 2023: In Barbados for the RL Workshop on Lifelong Learning.