Research

The Digital Cooperation Lab studies how humans and AI agents, in varying combinations, reach good collective decisions together. We bring computational social choice, AI, game theory, and combinatorial optimization to the foundations and tools of cooperation at scale — with an emphasis on the blockchain and DAO ecosystem.

A recurring constraint runs through everything we do: bounded attention. No agent — human or AI — can attend to every decision. So decisions must be configured, delegated, or handed off — and our research is the science of doing that well.

How we work

We build mathematical models of cooperation and stress-test them in simulation, exposing the tradeoffs, strengths, and limits of design choices before they reach the real world. In particular, we build a principled environment for tuning governance parameters and comparing design choices across every direction above. (Partially funded by IOG / Cardano.)

What we study

We organize our work by who cooperates with whom:

• Humans, with AI. Personal AI agents that help people find one another and form grassroots coalitions. Concrete projects: grassroots federation formation; AI-Agora coalition formation. (Submitted for EU — EIC Pathfinder funding.)

• AI, reviewed by humans/AI. Decision-making for content creation: evaluating machine-generated work through mechanisms for scarce evaluative attention. Concrete projects: token-based peer review; reviewer markets; perpetual retroactive funding. (Submitted for NSF-BSF funding.)

• Humans, with humans. Scaling direct democracy under limited attention through delegation and randomization. Concrete projects: topic-based liquid democracy; delegation with voter commitment; liquid democracy vs. lot. (Funded by the EU — PERYCLES.)