The field · Who else is trying
Every serious attempt at collective intelligence, from CERN to vTaiwan, reaches the same wall. We set out to build the layer each of them is missing.
What others have already achieved, and the exact point where each one stops. Each limit marks where the next attempt can begin.
Four institutions
These four cases run from collaboration between nations who share one way of knowing, to collaboration between ways of knowing that do not share much at all.
CERN
Countries that could agree on almost nothing else have built, together, the most complex scientific instruments in history. Its council holds scientific independence and political authority side by side, deliberately unmerged. Where it stops: every scientist at CERN works within one tradition, Western empirical science. Read the CERN case
IPBES
The global biodiversity panel gave indigenous and local knowledge formal standing alongside peer-reviewed science, and wrote procedures for how the two would be weighed. Where it stops: it exists to inform governments, and how different ways of knowing actually work together in real time is still unsolved. Read the IPBES case
Sant’Egidio
A lay Catholic community ended a sixteen-year civil war in Mozambique in 1992, where diplomats and the United Nations had failed. It could do so because it was a faith community, a mediator and a long-standing humanitarian presence all at once. Where it stops: it grew into this over decades, it was never designed, and it works within one religious culture. Read the Sant’Egidio case
EU Horizon
Europe’s research programmes are the most deliberately engineered collaboration system ever built, refined over nine generations since 1984. Where it stops: its own evaluations keep returning to problems that funding cannot fix. Our reading is that the limit is people’s capacity to hold real difference without collapsing it. Read the Horizon case
What technology already does
A growing body of work uses technology to help groups think together. Each strand below works well within its own terms, and each runs into the same limit. Figures are from our survey, The Technology Landscape (v2.0, May 2026), and the sources named.
Finding agreement in large groups
Taiwan’s vTaiwan and Polis tools have involved more than 200,000 people and fed into 26 pieces of legislation by 2025. Google DeepMind’s Habermas Machine tested AI-written group statements with 5,734 people in the UK. Participants preferred them to statements by human mediators, and independent raters found them clearer and more informative (Science, 2024). Where they stop: both find agreement among people who already share a framework. When the framework is what differs, as between a Quechua knowledge holder, a systems scientist and a Sufi philosopher, neither has a way in.
AI for collective intelligence
Research shows AI can help groups remember, focus and reason together. One experiment found something troubling as well. AI help made individuals more creative and made the group’s ideas more alike (Doshi and Hauser, Science Advances, 2024). Where it stops: nearly all of this work has been done with groups from one culture, usually Western, academic and scientific.
Knowledge that lives in relationships
Nora Bateson’s Warm Data Labs, run in more than 40 countries, work with knowledge that exists only in the relationships between things and is lost when pulled out of context. Where it stops: when the session ends, the conditions reset. Nothing accumulates.
Plurality
Glen Weyl and Audrey Tang’s Plurality (2024) argues that technology should turn social difference into progress instead of conflict. Where it stops: it deals with differences of values and interests. A Daoist philosopher and a systems ecologist may share democratic values entirely. What differs between them is what each is able to perceive.
Wiser AI
Contemplative AI tries to build principles such as non-duality and care into AI systems. Abundant Intelligences, an Indigenous-led programme, works to rebuild AI on Indigenous knowledge. Where they stop: both work on the AI itself. Our question is different. It is how people who already carry living traditions can meet each other in a way that produces something new.
What the map shows
Every one of these efforts, followed far enough, meets the problem of deep difference in how people know things, and finds it cannot handle that problem inside its own design. We are not offering an alternative to any of them. We are trying to build the piece each of them identifies as missing.
Know a project that belongs here?
This map is incomplete. If you know of work in any field that shows what becomes possible when very different ways of knowing work together, please tell us.