
They say that 80% of human communication is nonverbal. But it may be that 80%, perhaps 90%, perhaps 99% of all that can be known is inarticulable. This insight sits at the heart of the coordination problem.
The economist Friedrich Hayek won a Nobel Prize for demonstrating that a centrally planned economy cannot succeed because a centralized authority cannot know all the things required to run a society. Knowledge is dispersed and fragmented. He argued for the price system because prices encode that distributed knowledge into signals everyone can read. A high price communicates scarcity and demand without requiring anyone to survey every buyer and seller.
Thomas Sowell extended this work. He wrote that knowledge can be enormously costly and is often widely scattered in uneven fragments too small to be individually usable in decision making. The communication and coordination of these scattered fragments is one of the basic problems, perhaps the basic problem, of any society as well as its constituent institutions and relationships.
Beyond this dispersion, knowledge itself comes in different types. There is explicit knowledge, what you and I communicate now in words. And there is tacit knowledge: judgment, experience, craftsmanship. Aristotle wrote about this in Nicomachean Ethics. When you think about apprenticeship, you are not just learning instructions like carve this kind of wood like this. You learn the feel of the wood over years of practice. You know when it is done. You know how to give it shape. This cannot be transmitted in a manual.
To know certain things, we must become the kind of person able to receive that knowledge. To learn temperance, you must become temperate. You cannot actually communicate in a book how to be temperate. You have to live your way into it. This takes time. It happens with people, in community, across generations. This is the point of institutions and traditions and rituals.
Systems like our app, knowledge bases, and AI can only handle the articulable fraction, which is the smallest fragment of all possible knowledge. Even that cannot be known by everyone all at once. The three of us trying to run a company, transmitting information between our heads, is difficult. Scale that to twenty people, or a hundred, or ten thousand.
Most tools approach this by centralizing knowledge. Wikis, documentation, shared drives. But centralized repositories fail for the same reason central planning fails. They can never be complete. They are always out of date. They require everyone to maintain them, which nobody does. And they cannot surface context when you need it, only when you go looking.
Hayek's insight suggests a different approach: instead of centralizing knowledge, create signals that help distributed knowledge coordinate. That is the question Snapwork tries to answer. Not how do we store everything, but how do we surface the right context at the right moment. Knowledge does not need to be centralized. It needs to flow.
Written by a human.


