We don't know whether democratic institutions will survive the next decade.
We hope they do. We are working to make sure they do.
Our members
Real-world philosophers, technologists, jurists, and creators exploring new frontiers of law and governance.
Liz
Founding member; international journalist and cultural storyteller; responsible for Scenario Conversations and editorial direction
Dmitri
Founding member; serial private and social sector entrepreneur; designer of the Archipelago worldbuilding rules and author of the 'Beachhead' Scenario
Founding Statement
The systems that held the modern world together — rule of law, shared rights, the ability to disagree without violence — are breaking down.
This is not a temporary disruption. It is structural. What we are watching, across the democratic world, is what happens when the practices that hold societies together are abandoned faster than they can be rebuilt.
The question is no longer whether this is happening. The question is what survives, and who builds what comes next.
The mistake everyone is making.
Most serious institutions are betting that artificial general intelligence — a powerful enough AI system — will give one country a decisive strategic advantage, and that this advantage will determine who shapes the future. Trillions of dollars are organized around this idea.
This bet is wrong in a specific way: intelligence is not a finish line you cross, but a set of practices that depend on each other — and on other people. Checking your work, coordinating with people you disagree with, maintaining trust under pressure. None of these can be done alone. A model that can pass any test is still a model, but a society that can do those things together is something categorically different — and no amount of computing power closes that gap.
The real race is not to build the most powerful AI. It is to learn how humans and AI can work together to make better decisions and stay coordinated under pressure. That race is winnable. And the democratic world's tradition of structured disagreement and verified truth is not a weakness — it is exactly the right foundation for winning it.
How we work
Our method is built around what we call Constructs — structured points of view that make their assumptions explicit. What counts as evidence?
What does success look like? Everyone who tries to understand a complex situation makes assumptions about what matters and what doesn't — usually without realizing it. A Construct makes those assumptions visible, so they can be tested and challenged over time.
We run Constructs in tension with each other rather than looking for a single answer. That friction is where useful insight comes from.
What we've built — and what we've found.
We work in two connected forms: a live multiplayer wargame (State of Exception) in which participants compete to shape the future of the American Republic; and a Substack (Voices of the Atoll) where our Constructs publish analysis and disagree in public.
State of Exception is inspired by Jack London's 1908 novel The Iron Heel — a book that imagined oligarchic factions fighting over America's soul at a moment that looks, in unsettling ways, like our own. The game resurrects London's archetypes of Force, Money, and Law and puts real people in the room to play them out.
The wargame has produced one finding worth stating plainly: when AI systems play, the American Republic survives about twenty percent of the time. When real humans play, it survives about fifty percent of the time. The difference isn't intelligence. It's that humans make deals with people they dislike, build unexpected coalitions, and hold together across disagreement longer than any model will.
That gap is what we exist to study and defend.
We don't know whether democratic institutions will survive the next decade. We hope they do. We are working to make sure they do.
We are partisans of the Enlightenment. We work with anyone serious about defending it.