Infrastructure for planetary intelligence

Planetary problems need a planetary intelligence. What if we built it together?

WhatIfWe’s mission is to create the space for it: a shared ground where science, indigenous knowledge and the world’s religious and philosophical traditions work on the same problem and reach what none of them could reach alone. Every great meeting of minds has needed a place. A meeting at the scale of humanity needs a space as wide as humanity itself, reaching every continent, every language and every way of knowing. Today, for the first time, that space can be created.

Our intelligence is narrower than our problems

The problems that matter most now cross every line we have drawn: the climate, the loss of living species, technologies more powerful than the institutions that oversee them, conflicts between whole peoples. Each framework we bring to them sees a real part and treats its part as the whole. Technical answers to ecological collapse leave out the relationship between people and land. Economic answers cannot hold meaning. Spiritual answers often have no mechanism. So each answer tends to create the next problem.

The people who have tried hardest to widen our thinking show where it stops. IPBES, the global body that assesses the state of nature, gave indigenous knowledge standing beside science. Its 2019 global assessment reported that lands managed by indigenous peoples and local communities hold a large share of the world’s remaining biodiversity, something science alone could not have documented. Yet how different ways of knowing actually work on something together is still an open problem. Europe’s research programmes have engineered collaboration across borders since 1984, and their own evaluations keep meeting the same wall: getting disciplines to integrate, as opposed to sitting side by side in one project, remains hard to produce and hard to measure.

The knowledge we need already exists, spread across traditions that have each worked on it seriously for centuries. What has never existed is the space for them to work on the same problem together, without one being forced into the other’s terms and losing the very thing that made it worth having.

Every great meeting needed a place

People from very different ways of knowing have created something new together before. In Renaissance Florence, classical learning, Islamic scholarship and Christian theology met, and a way of understanding appeared that none of them held. In New Orleans, African rhythm met European harmony, and jazz was born. At the Santa Fe Institute, physicists, biologists and economists built a new science of complexity.

Each time, the people involved shared some ground: a city, a nightly bandstand, a common mathematics. In the Axial Age, China, India, Persia, Israel and Greece shared no ground, and their new ideas never met. And the denser the ground, the faster something new appeared. The Renaissance took centuries, jazz took decades, complexity science took years.

A meeting between all of humanity’s ways of knowing, at the scale of its problems, needs ground of its own, as wide as humanity itself. It reaches people where they live, across every continent, language and way of knowing, and it keeps what each meeting reaches. That ground has never been built on purpose. Creating it is WhatIfWe’s mission, and why it can work explains why it has to be built deliberately.

What we set out to build

The tools to build that space now exist. WhatIfWe is designed around an AI system for a job no system does yet: letting genuinely different ways of knowing work on the same problem, at scale, and reach approaches none of them would reach alone.

  • Tradition-grounded agents. Sixteen AI agents, each built only from one tradition’s own material and judged by people formed in it. They bring each tradition into the conversation and carry every contribution into the terms of the others.
  • An open global forum. Thousands of people, and in time far more, take part from their own tradition and in their own language, in many groups at once.
  • One record that grows. Every agreement, open disagreement and new approach is kept, with who contributed what, so each new piece of work starts where the last one finished.
  • An open benchmark tests the agents first. They work through thousands of real problems, and practitioners judge blind whether each tradition is still recognisable.

People who have gone furthest in their traditions work with the system for months at a time. They calibrate the agents and show what happens at the deepest level of each tradition. What we’ll build sets out the design, the principles and the questions we want to settle with partners.

AI is already shaping how the world thinks. It can widen it instead

A systems biologist in California and a Yup’ik elder in Alaska can be on a video call in minutes. Connection was solved in our lifetime. What still fails is understanding: to be heard at all, one of them has to put what they know into the other’s words, and most of it is lost on the way.

The machinery that decides what billions of people see already shapes how they feel about each other. Twitter’s engagement ranking served more anger and more hostility toward the other side than a plain chronological feed (Milli and colleagues, PNAS Nexus, March 2025). Reranking the live feeds of 1,256 people for ten days during the 2024 US election moved how they felt about the other party, in either direction (Piccardi and colleagues, Science, November 2025). The lever is real, and it can be pointed at understanding.

Meanwhile AI is narrowing how the world thinks. A 2026 study of language models on ethical dilemmas from Indigenous contexts found that they adopt a tradition’s persona on the surface while defaulting to individualistic conclusions (Merrikhi, preprint). Another found that an average American was a closer match to Indian cultural values than any model trained for India, and that better prompting did little to change that (Agarwal and colleagues, 2025). The more the world thinks through AI, the more that matters.

The parts needed to change this already exist. DeepMind’s Habermas Machine tested AI-written group statements with 5,734 people. vTaiwan has drawn in more than 200,000 people and contributed to 26 laws. Running a language model at a given quality becomes about ten times cheaper every year (a16z, November 2024). What WhatIfWe adds is the aim: synthesis, the new approaches to real problems that appear only when different ways of knowing work on them together.

What changes if it works

  • Approaches no tradition would reach alone. Real problems worked on by many ways of knowing at once, producing approaches that each participant can recognise as partly theirs and none of them could have produced alone. We call this synthesis, and it has already appeared, once at a time and by accident.
  • A reading of any hard question, from many traditions, when it is needed. Today “we need a cross-tradition reading of this” is a request nobody can answer. With a record that grows, it becomes one with an answer, and anyone can see who contributed what.
  • Each piece of work starts further on. Understanding is kept, so it does not have to be produced again from nothing every time.
  • The layer every deliberation platform is missing. The systems that already reach hundreds of thousands of people work within one frame. This would let them work across frames.
  • From understanding to action. In time, the community that forms around the record can choose which problems to take on next and coordinate the work on them. How a community this diverse decides what comes first is something the system will have to invent.
  • A test AI has never faced. Can a model reason from inside a tradition, and not only borrow its vocabulary? The answer matters to everyone who builds AI.

That is what we mean by planetary intelligence: a way of thinking together as wide as the problems we face, producing what no single way of thinking can.

Why it can work: the seed is already in every tradition

Every tradition has asked two questions. The first is personal: how does a person grow? The second is collective: how do people who see the world in completely different ways grow together into something none of them could become alone? Buddhists call part of it the Bodhisattva vow. Jews call part of it Tikkun Olam, the repair of the world. In southern Africa it is Ubuntu: a person is a person through other people.

Each tradition tried to act on the second question, and each stopped at its own edge. The Rimé movement brought Tibetan Buddhist lineages together and stayed Buddhist. The ecumenical movement brought churches together and stayed Christian. The Santa Fe Institute brought physicists, biologists and economists together, and all of them were scientists. Nobody failed for lack of seriousness. There was no way across.

This is what makes the system adoptable. If the traditions had nothing to say about the second question, the only way forward would be a new belief system to replace them all, which nobody would adopt. They do have something to say. Each one, followed far enough by its own logic, points beyond itself. So WhatIfWe works from inside each tradition, asks nobody to give up what formed them, and lets each tradition own what it helps produce.

We have written an account for each of sixteen traditions, from inside it, covering the traditions that have formed most of the world’s people. Each shows where its tradition already reaches out, and each is a way in for the people who hold it. Each is a starting point for the people who hold it to take deeper.

Our problems are wider than any one tradition.

The knowledge they need already exists, worked out apart.

Planetary intelligence is what may appear when those ways of knowing meet.

Nothing built so far lets them meet on purpose.

WhatIfWe exists to build the infrastructure that lets them meet.

Where we stand, and what it takes

This runs against the current. Today’s AI is built to answer in one voice, and today’s platforms to hold attention. A system that widens how humanity thinks will not appear by default. It will exist because people decide to build it, and that is why we want to build it in the open, with partners.

Nothing is built yet. What exists is the full argument on this site, sixteen accounts written from inside each tradition, a survey of what others have built and where each one stops, the design of the system, and a list of people to talk to, already tested against our own standard. The open benchmark comes first, because it needs only published texts and could show that the idea does not work. We will publish what it shows.

Each kind of partner builds a different part of the whole:

  • AI labs and researchers: models, compute, evaluation, and the design of the agents and the forum.
  • People inside the traditions: judgement on whether each tradition is represented faithfully, and doors into the traditions most of the world belongs to, which still have no guest.
  • Platforms that already work at scale: to test this layer alongside tools that already reach very large numbers of people.
  • Funders who want to know whether AI can widen how humanity thinks.

Partners sets out each role and what we would ask of it, and AI labs and funders will find a short brief written for them.