Forming Islands of Coherence

Every Thursday morning, we show unfinished work.

Not polished presentations. Not demos of things that shipped last week. Work in progress. Ideas that might become something, problems that aren’t solved yet, experiments with no conclusion. The only rule for showing your work is that you should end the same way every time: “If you think this is cool and you want to get involved, let me know.”

No assignment follows. No mandate. No steering committee decides who should care about what. The idea goes in front of the room. The people who find it worth their attention show up.

A few days later, what started as one person’s idea has five people mixed up in it. Not because anyone told them to be there. Because the idea had gravity.

That’s what an island of coherence looks like when it’s forming.

Coherence Is Not Order

Most executives want coherent organizations. They just don’t know what coherence actually is, so they build something else and call it the same thing.

The tools they reach for: alignment meetings, culture initiatives, OKR frameworks, off-sites where leadership presents the strategy and everyone nods. These things produce a different outcome, the appearance of coherence. Appearances require constant maintenance. The off-site ends. The culture deck stops circulating. The OKR dashboard goes stale. Every one of them dies the moment you look away.

That’s what happens when you spend your time trying to build order instead of conditions for coherence.

Order is different from coherence. Order is rigid, imposed, and expensive to maintain in an environment defined by chaos. Coherence is something humans build for themselves, in the presence of an idea worth following.

Ilya Prigogine, studying thermodynamic systems far from equilibrium, described it this way: “Small islands of coherence in a sea of chaos have the capacity to elevate the entire system to a higher order.” He was describing physics. He was also describing exactly what happens inside organizations when a good idea finds the right audience.

Chaos is the medium. You can’t overcome it. Build inside it. Create something durable: a pocket of coherence where the work feels like it matters, where the right people found each other, where something real is happening. Not because anyone ordered it into existence, but because the conditions were right and the gravity was there.

Gravity, Not Assignment

Projects get populated by assignment. Someone in authority decides who works on what, produces a staffing model, and allocates headcount. This works for known work: maintenance, repeatable delivery, things where the outcome is defined before anyone starts. Projects are a reasonable tool for that kind of work. But how much knowledge work truly exists in a place where conditions and outcomes are predictable ahead of time?

Islands don’t work that way. You can’t staff an island. The moment you try, it stops being an island and becomes a project. Projects have a fundamentally different relationship to the work than islands do. Assigned participation produces compliance and apathy. Voluntary participation produces interest and passion, something far closer to ownership.

Islands form around ideas with gravity. Gravity here is specific: the pull a genuinely interesting idea exerts on the people who recognize its worth. Not everyone will recognize it. Not everyone should. The right people, the ones whose involvement is pivotal for the idea to develop, arrive because they wanted to be there, not because they were told to.

The Thursday showcase is built around this mechanism deliberately. The work-in-progress format matters: polished presentations invite evaluation, not participation. Showing something unfinished signals that there’s still room to contribute. The “if you think this is cool” close is not a formality. It’s how gravity is made operational, converting recognition into action without requiring anyone in authority to coordinate the transaction.

We have a checkpoint built into how we work: “Did you kick ass today? Who agrees with you?” The second question is the one that matters. Not because consensus validates the work, but because coherence requires more than one person. An idea one person finds exciting is just an idea. Two people who find the same thing worth their attention, and can say why, is the beginning of an island.

If no one agrees with you, that’s information too. Not every idea has gravity.

The Lifecycle Most Leaders Won’t Respect

Every successful island eventually reaches the same moment. The idea formed, gathered gravity, grew into something real, and did what it set out to do. The problem got solved. The customer need got met. The product impressed.

At that point, the right relationship to the island is to walk away from it.

Most people can’t do it. The island became a place they knew how to navigate. It generated momentum, maybe revenue, certainly recognition. Walking away from something successful feels like abandoning a part of themselves. So instead of releasing the island to be what it became, they keep investing, wondering what else it could be, what new direction it could take, how much further it could go.

This is where the crystal castles get built. Elaborate, unrealistic visions assembled by people who prefer the comfort of a familiar island to the discomfort of starting fresh on a new one. The hallmark is that everyone, including the people building castles in the sky, knows they’re unrealistic. But even imaginary castles are more comfortable than the alternative: leaving to start over.

Islands don’t disappear when people walk away. They remain, still doing the work they were built to do, available to be rejoined when conditions warrant. What disappears is the energy and attention that could be forming new islands somewhere else, around ideas that haven’t found their people yet.

Drift is not failure. Drift is the lifecycle playing out correctly. Mature islands shed their teams as the work shifts from creation to maintenance, and the people who built them are needed where the next thing is forming. The organizations that build the muscle of walking away, and coming back when new gravity calls for it, keep forming new islands. The ones that won’t walk away collect nothing more than tall stacks of blueprints for crystal castles.

What You’re Actually Building

You can’t install coherence. You can’t mandate it, program it, or align people into it. What you can do is build the conditions where it forms on its own, recognize it when it does, and protect it while it’s fragile.

The Thursday showcase is one set of conditions. The “who agrees with you?” checkpoint is another. The discipline of walking away from a mature island is another still. None of these are culture initiatives. They’re operating decisions that shape what the environment rewards and what it ignores.

Prigogine’s insight was that the chaos isn’t the obstacle. It’s the medium. The islands form inside it, not instead of it. Your job isn’t to eliminate the chaos. It’s to build inside it, one island at a time.

That’s the whole job.