Obsidian Delta

November 20, 2026 History

The Word That Stuck

I arrived in Canada as a child. One of my earliest memories of this country is sitting in a classroom in 1989 and hearing a word I did not yet understand: multiculturalism. I did not have the language to explain why it mattered. I only knew that it seemed to describe, in one word, a promise: that you could be fully yourself and fully here at the same time.

That word stuck. I have spent a long time since trying to be worthy of it.

What I have actually built

Before I say what I am working on now, it is fair to say where I have come from, because the work only makes sense in that light.

I spent years designing and running telecom systems for Canadian companies: the unglamorous plumbing that has to work every time or not at all. I was a technology advocate for the taxi industry, and I stood up at a large public town hall to talk, plainly, about Uber and what it meant for the people who drove for a living. I did advocacy work at the CRTC. None of this was about being seen. It was about systems that affect ordinary people, and about being in the room when decisions were made.

In 2017 I incorporated Obsidian Delta Corp. In 2018 and 2019, Obsidian Delta worked with the Microsoft Research team behind a project called Conversation Learner. “Worked with” is exact: I filed bug reports and feature suggestions. In May 2019 I flew to Redmond and met with the team in person, a Conversation Learner / Obsidian Delta sync at Microsoft Building 99, and the day after, I sent them a write-up on how our own models had evolved. Around the same period, in December 2018, I had created RS3, “Really Simple Social Syndication,” an extension to RSS 2.0, and I corresponded with contacts at Microsoft Research about it.

I mention these not as credentials but as a record. I am describing a habit: show up, do the technical work, write it down, hand it over, ask to be told where it is wrong.

The problem worth the rest of my time

Cultures are being lost faster than they are being kept. Not through malice, but through time. An elder passes, and with them goes a language, a recipe, a migration story, a way of settling a dispute, a song nobody wrote down. We are good at saving photographs and bad at saving the people in them.

The usual responses are two, and both fall short. One is to let it slip, and call that nostalgia. The other is to “collect” it: to extract stories into someone else’s archive, under someone else’s name, on someone else’s terms. That second one looks like preservation and is often a quieter kind of loss.

I wanted a third way: to keep a culture the way a careful anthropologist would, and to do it at a scale that lets it be shared across the world, without the person at the centre of the story ever losing hold of it.

Maddie

This November I incorporated a not-for-profit, the DAX Foundation (legal name: Delta DAX Foundation), to carry that work.

The first piece of it is Maddie. Maddie holds life-story conversations. It is modelled on how cultural anthropologists engage people: non-invasively. It listens. It tries to understand. It catalogues. It does not interrogate, and it does not rush.

The rules are simple and they are not negotiable:

  • Your words stay your words. They are kept verbatim and remain attributable to you.
  • You decide what is shared. Not me, not an institution.
  • “Off the record” means off the record. What you ask to keep private stays private.

I say it this way inside the company: privacy and security are architecture, not policy. A promise in a policy document can be revised on a Friday afternoon. A guarantee built into how the system is constructed cannot be quietly undone. We build the second kind.

Why standards matter more than cleverness

It would be easy, and wrong, to treat this as a technology project that happens to touch culture. It is a cultural project that happens to use technology, and the people who have thought hardest about doing this ethically are not engineers.

So the work is grounded in theirs. The Oral History Association’s Principles and Best Practices. Canada’s TCPS 2, the national standard for research involving people. UNESCO’s ethical principles for safeguarding intangible cultural heritage. The Local Contexts labels, which let communities attach their own rules about how their knowledge may be used.

These are not decorations on a finished product. They are the specification.

And because a specification is only as good as the critique it survives, I have begun writing to the people who do this for a living. In October I wrote to the Louie B. Nunn Center for Oral History at the University of Kentucky. Next on the list are the Centre d’études acadiennes at the Université de Moncton, the Centre for Oral History and Digital Storytelling (COHDS) at Concordia, and MUNFLA at Memorial. I am asking them for critique, not endorsement. I would rather be corrected early by people who know more than I do than be congratulated by people who know less.

Starting at home

You do not earn the right to keep a community’s stories by sending a press release. You earn it by showing up and listening first.

So this month I made courtesy visits across my own community in New Brunswick: to the library, the multicultural association, a local shelter, the food bank, and to local elders. I went to introduce myself and to listen, not to pitch.

One conversation stays with me. An 84-year-old man who has lived here his whole life gave me advice I will keep: always be honest, and the community has to keep helping each other, multicultural, hand in hand. That is the whole project in a sentence, said better than I could say it, by someone who had no reason to flatter me.

Where this goes

The hardest and most important part is still ahead. Most of the wisdom worth keeping sits with people who have never used, and never want to use, a piece of technology, and who may not tell their story in English or French.

The long-term aim is a translation layer: a way for an elder who knows nothing about computers to contribute their wisdom in their own words and their own language, with their name attached to it, kept exactly as they said it. The technology should disappear. The person should remain.

That is a large ambition and I am not going to overstate how far along it is. What I can say is the direction, and the discipline: keep the words verbatim, keep the control with the person, build the privacy into the architecture, and check every step against people who know the field better than I do.

The word I heard in 1989 was a promise. This is my attempt to keep it: carefully, and in a way that can be shared.

An invitation

If you are part of a community, if you are an elder with a story, a teacher, or an oral-history researcher, and you would like to be involved, or simply to tell me where I have this wrong, I would like to hear from you. Write to me at jason@obsidiandelta.com and we will talk. There is nothing to buy. There is only work to do, together.


Le mot qui est resté

Je suis arrivé au Canada enfant. L’un de mes premiers souvenirs de ce pays, c’est d’être assis dans une salle de classe en 1989 et d’entendre un mot que je ne comprenais pas encore : multiculturalisme. Je n’avais pas les mots pour expliquer pourquoi il comptait. Je savais seulement qu’il semblait décrire, en un seul mot, une promesse : que l’on pouvait être pleinement soi-même et pleinement d’ici, en même temps.

Ce mot est resté. Depuis, je consacre beaucoup de temps à essayer d’en être digne.

Ce mois-ci, j’ai constitué un organisme à but non lucratif, la Fondation DAX (nom légal : Delta DAX Foundation), pour porter ce travail. Sa première réalisation s’appelle Maddie : elle recueille des récits de vie à la manière d’un anthropologue, sans jamais s’imposer. Elle écoute, elle cherche à comprendre, elle catalogue. Vos mots restent vos mots, conservés tels quels et attribuables à vous. C’est vous qui décidez ce qui est partagé. Et ce qui est dit « à titre confidentiel » le reste.

Si vous faites partie d’une communauté, si vous êtes un aîné porteur d’une histoire, un enseignant ou un chercheur en histoire orale, et que vous souhaitez participer, ou simplement me dire là où je me trompe, j’aimerais vous entendre. Écrivez-moi à jason@obsidiandelta.com. Il n’y a rien à acheter. Il n’y a que du travail à accomplir, ensemble, main dans la main.

← All writing