AI for Institutional Design

How can we use AI to help us design and build better institutions, which enhance human capabilities for collaborating, governing, and living together?

• Where are the best opportunities for deploying such techniques?

• What bottlenecks are there to current approaches in this vein, and how can we unblock them?

The Collective Intelligence Project and the Cooperative AI Foundation convened around 25 participants from around the world, from institutions including Google DeepMind, Demos, Harvard, MIT, NESTA, OpenAI, Protocol Labs, Oxford, and UCL, for two days of participant-driven presentations and discussion. Participants submitted ideas in advance, which the organisers clustered into themes (Composition, Information, Representation, Simulation, Delegation, and Support). Theme by theme, everyone gave a lightning talk on their idea, before breaking into small groups to work on specific ideas, which were peer reviewed at the end. The outcomes of this process were formed into 'project cards’, which are hosted on a new website. The workshop’s directors hope to support people interested in working on these projects, or in the area more generally, including the potential for funding, mentorship, or collaboration.