Martín Soto

Martín is a graduate in Maths and in Physics, and finishing his MSc in Mathematical Logic, where he’s pursuing research in Computational Complexity with Albert Atserias. He’s spent the last year exploring theoretical and conceptual research avenues in AI Safety, with collaborators from OpenAI and London’s Center on Long-term Risk, among others. He’s now decided to focus on multi-polar failure modes, tackling them both from theory (Cooperative AI) and governance. That’s why, at CHAI, he’ll be working with Niklas Lauffer on multi-agent training dynamics. In the past, he’s also researched Algorithmic Decision Theory, and AI Safety cause prioritization.