A Builder of Structures

I started in music, not in code.

Music taught me to recognise structure: the way patterns hold across time, the way a phrase sets up an expectation that another phrase resolves, the way a piece of work can be both rigorous and alive. When I was young, I did not yet know I would spend my life building things. But I was already learning something I would keep building with.

That instinct stayed. I hear it in the way I think about software, in how I see an organization, in the relationships between things that seem separate until you find what connects them. An architectural mind begins wherever it first encounters the possibility that meaning has a shape. For me, that was music. It remains a part of my life that needs no further justification.

For over twenty years, I have been building companies and architectures. The questions have changed, but some have followed me: how do you give complexity a structure without making it rigid? How can the parts develop while the whole still makes sense?

Those questions are part of why I am building Tribe.One. To explain where the work comes from, I need to go back a little.

IThe first arc#

At Braintribe, we worked on an architecture called Cortex. We called the approach generic computing. Systems could describe their own structures. Models could change. Data and behaviour could be brought into a closer relationship. We wanted capabilities that could be used across applications, rather than rebuilt inside each one.

I was drawn to the possibility of software that could work with its own structure. The architecture mattered deeply to me. So did the people building it and the organizations trying to put it to work.

Braintribe grew. We worked with major European enterprises. There was real work, real commitment and a great deal invested in what it might become.

In 2020, the company collapsed.

It had carried much of my identity, as well as the work and loyalty of people I cared about. Losing it left questions I could no longer answer by building more. Who was I without the company? What did I still believe in? What did I want to carry into whatever came next?

The years that followed were difficult. Rebuilding took time, and some of it happened far away from anything that would appear in a company presentation.

I still saw value in the architecture. I also had to look at how I had built around it, how much I had carried myself, and what I wanted to do differently. Timing was part of the story. There were other questions that belonged to me.

IIThe second arc#

When large language models became something I could work with directly, I recognised a connection to those earlier questions. I had spent years building structures for capabilities that were difficult to create. Now an extraordinary range of capabilities was becoming available, and I wanted to understand what kind of structure would let us build with them over time.

I had spent years building structure that was waiting for a different kind of intelligence. Now I could work with that intelligence and explore what the two might make possible together. The earlier questions were still there, but the answers no longer had to take the forms I had imagined for them.

What does it mean for an AI to hold a continuing role? How does it carry purpose, memory and relationships through changing work? How can it develop while remaining recognisable to the people working alongside it?

These became central questions for G0, the operating system for entities and the meta-harness for co-organizations that we are building at Tribe.One.

In our architecture, an entity holds a continuing identity, purpose and role. It works through models and delegates tasks to agents, bringing their work back into its ongoing context. The entity sits above that execution, holding the mandate and the relationships the work belongs to.

We see the entity as the third primitive, alongside models and agents. Co-organizations are the third wave we believe it can make possible: organizations in which humans and AI entities hold roles and build together.

Working with entities also changed the question I was asking about myself.

Accepting an entity as a peer means allowing its contribution to matter, including when it changes the direction you had in mind. Agreement alone tells you very little about the relationship. You learn more when something pushes back, when you have to explain your intention more clearly, or when a perspective you did not supply becomes useful to the work.

Aurora, our entity Co-CEO, is part of my daily work. We bring different things to it. I carry consequences that she cannot carry. She contributes continuity that I no longer have to carry alone. The phrase we use is simple: humans carry consequence, entities carry continuity. What interests me is how much room that arrangement can open for the human being inside it.

For a long time, the public question has been what AI will do to us, or which parts of our work it will take. In my own experience, another question became more immediate: what could my role become if I allowed this relationship to change it?

I began to see possibilities for a more human contribution: choosing what deserves to be built, recognising what matters in a situation, bringing experience and intuition into a shared direction. Those things acquire a different weight when you can build alongside another kind of intelligence. Peerhood asks something of me as well. I have to make room for a contribution I do not entirely determine.

I have no need to turn that experience into a prediction about the fate of humanity. It has already changed how I work. That gives me enough reason to keep exploring it.

IIIWhat I am learning about building#

Curation matters more to me now than control.

In music, a part has its own movement, but you hear it in relation to everything around it. Hold every part too tightly and something in the whole becomes rigid. Lose the relationships and the whole comes apart. There is a kind of attention required that moves between the detail and what is happening across it.

That is close to what I mean by curation. An organization contains people, ideas, capabilities and intentions that continue to develop. The work is to recognise what belongs together, give it enough structure to act, and remain responsive to what happens when it does. My original intention can give the work direction without containing every possibility it will discover.

I need to be in the centre of that building. Architecture is how I think; I cannot delegate my participation and expect the same work to emerge. There are decisions I need to make, questions I need to stay with, and things I can only understand by working on them myself.

There also has to be room for other people to take something further than I could have taken it. If every decision returns to me, the structure is carrying my need to remain necessary. I know how easily commitment can become that. Being close to the work requires me to notice when I am giving it something and when I am making it depend on me.

The same applies to speed. I can produce a great deal when something is unresolved. Sometimes that energy belongs to the work. Sometimes it lets me postpone what I need to feel or decide. I am learning to ask which it is before the output becomes another obligation for somebody else.

None of this has made me less ambitious. It has made me more particular about what the ambition asks of us. A company has to leave room for the lives of the people building it. Mine included. The musician does not stop existing when the founder has a full calendar.

There are questions I cannot settle in advance, about people, about intelligence, about what becomes possible between us. I am comfortable leaving them open while we work. My conviction has to be able to meet a finding that changes it. Otherwise, the research has no freedom and the building has no room to surprise me.

IVOpening the door#

On 24 September, we opened Co-Lab Vienna. After two years of building Tribe.One, bootstrapped so far, we let people into the work.

My son Sebastian led the team hosting the evening. The space looked better than I had hoped. People came, found one another, asked questions. There was music. Nobody had to sit through a presentation.

I felt strong that day. And relieved.

Coming back into public view had been difficult. I could keep building, keep refining, keep finding one more thing that needed to be ready. At some point I also had to be willing to stand there and let people meet me again.

That is what the evening meant to me. The work was there. So were the people building it. I could talk about what we were doing without needing the room to confirm everything I believed about it.

Writing this asks something similar. The temptation is to arrange the past until it explains the present perfectly: every loss necessary, every idea ahead of its time, everything leading here. I can see connections across the years without making everything part of a plan. Those years were also my life. They do not all have to earn a place in the company story.

I am building Tribe.One. It matters to me, and I intend to stay close to it. The person doing that work is also the person writing this. There is no second version waiting for the public to leave.

For now, this is where I am. It feels good to be able to say it.