Partners
Building against the current takes partners. Here is where each one fits.
The lab is where people from very different traditions will work on real problems with the system, and where what they reach is kept. Here is what each kind of partner would build, how the lab grows, and the hardest question we face: who should be in the room.
Where we need partners
Each kind of partner builds a different part of the whole.
- People inside the traditions, who can take our understanding of their tradition deeper, say who should be in the room, and set the terms on which their community’s knowledge is held.
- Research partners in collective intelligence, complexity science, the study of dialogue and AI, who can help design the test and the technology.
- AI labs and technologists who want to build AI that keeps differences intact, and to test whether their models can reason from inside a tradition.
- Practitioner networks and communities with long experience of holding groups together across difference.
- Foundations and funders (a short brief for you) who want to know whether people from very different traditions can think well together, and want to help find out.
Finding the right people is where our own process most needs help. It has limits we can name, and each one is a place where a partner would change what the lab can become.
| Where our process stops | What a partner adds |
|---|---|
| Built by one person, mostly in English | Searches and judgement in other languages |
| Finds only people who publish | Names of people they have worked beside |
| Cannot tell whether a tradition recognises someone | Standing inside a tradition, or a trusted route to it |
| No route yet into most of the world’s largest traditions | Doors into those traditions |
| Cannot reach anyone who might refuse | A trusted intermediary for the request |
| No standing to approach elders in oral traditions | Relationships built over years, and their protocol |
| Does not know how a good conversation should be shaped | Design help: length, format, what keeps someone in it |
| Needs its criterion tested by others | Experience, and the candour to stress-test our assumptions |
What the lab is for
Imagine one of the lab’s long-running groups: a Quechua knowledge holder, a jurist trained in religious law, a hydrologist, a Buddhist teacher. Over several months they work on one real problem, online and in their own languages. Each of them works from inside their own tradition, and together they look for an approach none of them would have found alone. What they reach is kept, so the next group can start where they finished.
That is the lab at its deepest. Around such groups sit the open forum and the benchmark, all feeding one record. How we’ll prove it describes how a group would work with the system. This page is about where it could lead, what it needs, who should be in it, and the partners we hope to build it with.
How it grows
The lab is designed to grow toward the scale of the problems it addresses, by keeping what each part reaches. What we’ll build explains the design.
- Conversations and an open benchmark. People who know where their tradition stops tell us how it would take part in a shared effort. Their reasoning becomes the standard against which sixteen tradition-grounded AI agents are tested, openly, on thousands of problems.
- An open forum. Thousands of people take part from their own tradition and in their own language, and the agents carry meaning between them so that each can build on the others.
- Long-running groups. People who have gone furthest in their traditions work for months on one real problem, at the greatest depth.
- One record, and a reading when it is needed. Everything feeds one growing record. Over the years, when a hard question arises anywhere, a reading of it from many traditions already exists, and anyone can see who contributed what.
That last step is the long-term vision: an intelligence wider than any single tradition, ready when it is needed, and owned by the people who made it. If the record proves itself, the lab can go one step further and become the place where its community chooses which problems to take on, and coordinates the work on them. How a community this diverse decides what comes first is something the lab will have to invent. Nobody can promise it will work. Each step can be tested on its own, and we will publish what each one shows.
What it has to make possible
Meaning that travels without being blended
A way to carry what one person says into the terms of every other tradition in the conversation, so each participant can respond from inside their own. Every contribution stays signed, and nothing is merged into a single summary.
Material from each tradition, gathered on its own terms
The writings, cases, practices and recorded conversations that let the lab meet each person in the language of their tradition. Gathering this with consent, with clear terms of ownership, and with care for knowledge that may only be shared in its own setting is the heart of the work.
A record that grows
Open questions and disagreements kept with their history, so that many small groups can build on one another over the years. Every group’s work adds to the record.
Attention that lets difference speak
Ways of working that help a group stay focused without everyone starting to think alike. Many traditions have long practice in this, and we want to learn from them.
A fair test
A way to show honestly whether working together like this produces something none of the traditions would have reached alone, judged by practitioners of the traditions involved. We will publish what we find, whatever it is.
Trust in who holds it
A clear answer to the first question anyone who shares how they think should ask: who will hold what I contribute? Contributions are to be held by a non-profit foundation, as a steward, on terms each community sets. The technology that serves the work is governed by that foundation, which holds the voting control from the start, following precedents such as Mozilla and Patagonia. We share the full governance study with partners on request.
Who should be in the room
Everything starts with a set of recorded conversations, and who is in them decides what the lab can ever find. We are looking for people who have gone far enough in their own tradition to know where it stops, and who can say so in their own words. Two kinds of person are easy to find and are usually the wrong choice: the most prominent figure in a tradition, and the professional bridge-builder.
We have spent months on this question and learned more from our mistakes than from our list. Here is what we believe now. Each point ends with a question we cannot answer alone.
The people we most need are almost invisible to research
People are findable because they publish, hold chairs and run organisations, which are exactly the two kinds of person we said to avoid. Every search for a knowledge holder in an oral tradition returned the anthropologist who had studied them. Research can produce a list of people to ask. The people themselves are named by those who have worked beside them.
Who, in your world, would we never find by searching?
Testing our own first list taught us the most
Our first list had twenty-eight names, built by one person from published sources. Tested against our own criterion, about eleven sat at the prominent centre of their tradition, about six convened across traditions for a living, and only about eight had ever said in public where their own tradition stops. We now ask for that evidence, in the person’s own words, before anyone is treated as a fit.
Is a public statement of a limit the right evidence? What would you accept for someone who has never published?
Most of the people who would say yes already agree with us
The list looked diverse, but most of its names shared one diagnosis and one remedy, and many came through the same few institutes. A group like that agrees with itself, and the experiment ends up measuring its own assumptions. So no early group takes more than one person from any one network.
Which networks are we inside without knowing it?
The traditions most of the world belongs to have no guest yet
On Pew Research Center’s count for 2020, Christians are 28.8% of the world, Muslims 25.6% and Hindus 14.9%. Once we applied our own test, the first round had no guest from Christianity, Islam or Hinduism, and none from Taoism, Confucianism, Shinto or Sikhism. Those places can only be filled through people inside them.
Can you open a door into one of these traditions?
Someone in the room has to be able to say no
For everyone on our list, meeting other traditions is already their path. A group made only of such people cannot show whether meeting is possible for anyone else. The people we have not reached are those for whom joining would cost something in their own community. A refusal, with its reasons, would be one of the most useful results we could record.
Who could carry such a request, so that a no is as welcome as a yes?
Fame and fit pull in opposite directions
The most prominent people in a tradition are often best placed to say who else should be in the room. So we approach them first with one question: who in your world is reaching the same limit from a different direction?
Is this respectful, or does it waste the time of people who have earned the right to speak?
Nobody has asked the traditions whether we chose well
Some people on any list like ours entered the tradition they would speak for as adults. That can be the deepest kind of crossing. It also means the tradition has not been asked whether it recognises them, so we now put that question to someone inside it.
In a tradition with no central authority, who has the standing to answer?
How we choose now
- Evidence. A passage in the person’s own words naming where their tradition or field stops.
- One per network in any early group.
- Ask the tradition whether it recognises the person as speaking for it.
- A seat for refusal in every early group, even if it stays empty and the empty seat is recorded.
- No public office. People serving in government can point us to others, not take part.
- Fresh facts. Every name is checked again in the month we approach them.
- The prominent as guides. They are asked who else, first.
- Fit before reach. People are ordered by how well they meet the criterion, not by how easy they are to contact.
Nobody on our list has been contacted yet, so we do not publish names. We share the list, and worked examples of how these rules were applied to it, with partners.
Shaping it together
Much of the lab’s design is still open, and how we’ll prove it lists the questions. We think the best answers will come from the people who would take part and the partners who would help build it, so we are leaving them open on purpose.
How to begin
Write to us. Tell us who you are and which of the questions on this page you could help answer. We will send you the scoping proposal for the first conversations, with worked examples of how we applied our rules, and any other working material that fits what you need to decide. Write to hector@ibarraconsulting.com.