Vojtěch Kovařík

Vojtěch Kovařík (also known as Vojta Kovarik) is a Czech AI safety researcher and postdoctoral fellow at the Artificial Intelligence Center, Czech Technical University in Prague. His academic background is in mathematics, having completed a PhD at the Faculty of Mathematics and Physics (MFF) of Charles University Prague, with a focus on analysis, topology, and Banach spaces, as well as game theory with imperfect information games. He subsequently held a postdoctoral position at Carnegie Mellon University's Foundations of Cooperative AI Lab (FOCAL), collaborating with Caspar Oesterheld and Vincent Conitzer on game-theoretic approaches to AI alignment. His AI safety research focuses on systemic risks and consequentialist reasoning, including work on Goodhart's Law applied to AI (formulating "Extinction-level Goodhart's Law"), cooperative AI, AI debate frameworks, and how AI systems may reason strategically during evaluations. He has co-authored papers at IJCAI, NeurIPS, and the SafeAI@AAAI workshop, and received Long-Term Future Fund grants to support his transition to full-time AI safety research.

No endorsements yet.

Personal Website

-

LinkedIn

-

Twitter / X

-

LessWrong

vojtakovarik

EA Forum

-