What We Study:
Why do groups fail to cooperate even when everyone would be better off? From climate action to AI safety, this puzzle is at the heart of many of the biggest challenges we face. We use game theory, simulations, and AI to study how cooperation emerges and breaks down — from algorithms that must learn to collaborate with strangers, to societies navigating collective action, to the scientific system itself and whether its incentives help or hinder the search for truth.
Media:
- Visible punishment key to large-scale cooperation — A nice summary of our work on pro-social institutions to enable large-scale cooperation. phys.org
- 🇨🇴 Cooperación y Competencia en Inteligencia Artificial — Catedra Colombia, National University of Colombia. Outreach talk on AI and cooperation. In Spanish (English subtitles). YouTube
- Journals and Scientists' Incentives — talk — A talk on how journal incentives shape the quality of peer review and scientific publishing. YouTube
- Evolution of cooperation and cooperative artificial intelligence — Press release about our work on picking strategies in games of cooperation. Max Planck Institute for Evolutionary Biology
- Repeated games and direct reciprocity — perspective in Science — Our work on direct reciprocity in structured populations was featured in a perspective by Thom Sherratt and Gilbert Roberts. Science