For years, I assumed that collaboration had a relatively simple economic logic. Someone owns the company. People work for it. Their roles are defined. Salaries, equity, or contracts determine how value is distributed.
But what happens when none of that exists yet?
What happens when a few people come together around an idea, without knowing whether it will become a company, a product, a community, or something else entirely? Who owns what? And, more importantly: how do we recognize the value each person helped create?
This question has followed me through several projects. I still don’t have a final answer. What I have is a series of experiments that have gradually changed the way I think about contribution, ownership, and reward.
The problem with hours
The obvious starting point is time. If three people build something together, we can simply track how many hours each person contributes. It is measurable. It feels objective. And it is certainly better than relying on memory months later.
But it quickly becomes clear that an hour is not a unit of value. One person may spend ten hours implementing something that another person conceived in a ten-minute conversation. Someone may introduce the person who changes the direction of the entire project. Someone else may ask one uncomfortable question that prevents months of work on the wrong problem.
There are also contributions that are difficult to see: holding the group together, creating trust, bringing energy at the right moment, challenging an assumption, opening a door, or simply helping an idea survive long enough to mature.
Time matters. But time and value are not the same thing.
Discovering contribution as a network
Through my work with the Hamama community and later with VAST — the Value Attribution and Sharing Toolkit — I started thinking differently. Instead of asking “Who owns how much?”, I became more interested in asking “What actually contributed to the value that emerged?”
That subtle change opened a much larger space. A project is not simply a collection of tasks performed by individuals. It is a network of contributions. Ideas influence decisions. Decisions lead to experiments. Experiments create learning. Introductions create relationships. Relationships create new opportunities.
Value often emerges from the connections between contributions, not from any single contribution in isolation. This suggested that perhaps attribution should not be decided once, at the beginning of a project. Maybe it should emerge gradually from the history of the project itself.
Recognition before valuation
Another insight came from experimenting with community currencies. Instead of immediately trying to calculate the monetary value of every contribution, what if participants could first recognize value in each other?
Imagine that every member of a project periodically receives a limited amount of recognition currency. You cannot keep it forever. Its purpose is not accumulation. Its purpose is circulation. When someone helps you, moves the project forward, brings an important insight, creates something useful, or contributes in a way you believe matters, you can transfer some of that currency to them.
This creates something interesting. Instead of asking an algorithm to determine value, we begin by capturing human perception of value. Over time, a graph appears: who contributed, who recognized whom, which contributions were repeatedly considered meaningful, where value seemed to emerge.
The system does not need to declare that these signals are value. They are evidence about value. That distinction became increasingly important to me.
Gamification as an engagement layer
A new idea entered our experiment recently, suggested by Victor: gamification. At first, this might sound like a small product feature — points, badges, streaks, leaderboards. But I think there is a more interesting possibility.
One of the challenges of any contribution model is participation. A system can only learn from contributions that become visible. Yet documenting work, recognizing others, reflecting on progress, and keeping commitments up to date can easily feel like administrative overhead. Gamification could help turn these behaviors into part of the experience itself: recognizing documenting a contribution, acknowledging someone else’s help, completing a commitment, connecting two ideas, or helping the team move toward its shared North Star.
But there is an important danger here. What we reward will influence how people behave. If we award points for hours, people may optimize for hours. If we reward the number of contributions, we may encourage quantity rather than significance. And if we create a leaderboard, collaboration can quietly become competition.
So I am beginning to see gamification not as another measure of value, but as an engagement layer around the value-attribution process. Its purpose should not be to tell us who is winning. Its purpose should be to encourage the behaviors that make collective value more visible, more recognized, and easier to understand. This creates an important design principle for our experiment:
Gamify participation, not human worth.
Then came EMET
EMET gave us an opportunity to turn these ideas into a real experiment. We are a small group creating something together while simultaneously trying to understand what that “something” should become. There is no mature company structure. There are no neatly separated job descriptions. Contributions include software development, ideas, research, conversations, writing, design, coordination, money, relationships, and emotional energy. In other words, exactly the kind of environment in which conventional attribution becomes difficult.
So we started with something deliberately simple. We created EMET Trace. The first layer records contributions. The second records time. Not because we believe time equals value, but because time is one useful signal and, importantly, one we can capture relatively objectively.
From there, we want to experiment with additional signals: peer recognition, financial contributions, decisions influenced, ideas developed, commitments fulfilled, and eventually the relationships between these contributions. The goal is not to create a perfect formula. The goal is to create a trustworthy history from which fairer conversations about value can emerge.
The agent changes the equation
There is another dimension to this experiment that I did not anticipate when I first started thinking about contribution models. AI agents can now participate in the process itself.
An agent can observe the project’s history, remember decisions, ask team members to record missing contributions, identify connections between ideas and outcomes, and help maintain the project’s collective memory. This could dramatically reduce one of the biggest problems with contribution tracking: nobody wants to spend their life filling out forms.
But it introduces a fascinating new question. If an agent helps create value, how should that value be attributed? To the person who prompted it? To the person who built it? To the team that shaped its context? To everyone whose previous contributions became part of its memory?
Perhaps even the question itself is becoming outdated. The emergence of agents makes the traditional model of individual authorship increasingly inadequate. Creation is becoming collective not only between humans, but between humans and machines.
From accounting to sensemaking
I used to think we were trying to build a better accounting system. I no longer think that is the main challenge.
We are trying to build a sensemaking system for value creation. Accounting asks: how much does each person get? The deeper question is: how did this value come into existence?
If we can build a rich enough history of contributions, relationships, recognition, decisions, and outcomes, then reward becomes a later layer. Money, equity, revenue sharing, reputation, or community currency can eventually be mapped onto that history. But if we jump directly to the reward formula, we risk optimizing what is easy to measure rather than what actually matters.
Direction matters too
There is one more lesson I have learned recently. Contribution cannot be evaluated independently of direction. A person can work extremely hard and still move the project away from where the group wants to go.
So before asking how much each contribution is worth, a team needs some shared understanding of what it is trying to create. A North Star. Not necessarily a rigid business plan. In early-stage collaboration, uncertainty is part of the process. But there must be enough shared direction to distinguish between effort and progress.
This connects contribution attribution to something much larger: collective purpose. And it creates another interesting role for an agent. Perhaps the agent should not only remember what we did. Perhaps it can also help us continuously connect our actions, commitments, and contributions back to the shared direction we agreed upon. Not as a manager. Not as a judge. But as a guardian of collective memory and purpose.
A layered model is beginning to emerge
Looking at these experiments together, I am beginning to see several distinct layers.
- Trace captures what happened: who contributed what, what decisions were made, how much time or money was invested, what commitments were created and completed.
- Recognition captures how people experienced those contributions: who helped whom, what felt meaningful, what contribution changed something for another participant or for the project.
- Gamification encourages participation: it helps make documenting, recognizing, reflecting, and contributing part of the ongoing experience rather than an administrative burden.
- Direction provides context: contribution only becomes meaningful relative to what the group is trying to create.
- And only then comes Value Attribution: not a magical formula that calculates human worth, but a process that uses all these signals to support a fairer conversation about how the value created together should eventually be shared.
This layered view is still a hypothesis. But it feels very different from where I started.
Where I am today
I don’t have a reward model to sell. I have a hypothesis, and an experiment.
I believe that fair attribution in collaborative creation may emerge from combining several imperfect signals: time and effort, documented contributions, peer recognition, financial contribution, relationships between contributions, outcomes, and alignment with a shared direction.
None of these signals should determine value alone. Together, however, they may tell a much richer story. And perhaps AI can help us maintain that story without becoming the authority that decides what people are worth. That distinction matters to me.
I don’t want to build an algorithm that tells a group “This is your value.” I want to create conditions in which a group can look at its shared history and say, “Now we can have a better conversation about what we created together, and how we should share it.”
Maybe that is the real shift. From measuring people to understanding contribution. From allocating ownership to tracing value creation. From designing the perfect reward formula to creating the conditions for fairness to emerge.
That is where my journey has brought me so far. And for now, I think the experiment is more interesting than the answer.
← All posts