23. The Way Forward

Reasons it hasn't happened yet. That was the last chapter's verdict, delivered after seven obstacles, and I want to take it at its word before I build anything on top of it. Fragmentation, chicken-and-egg, the punishing user experience, the privacy and regulatory fights, the walled-garden incentives, the centralization trap, the deep tension between pseudonymity and Sybil-resistance — every one of them turned out to be a coordination problem, an adoption problem, or a design problem, and not one of them a law of mathematics. That distinction is not a rhetorical flourish. It is the foundation this chapter stands on. Coordination problems, adoption problems, and design problems are the kind that get solved, slowly and then all at once, the way every infrastructure problem in the history of the network has eventually been solved. The pieces of the answer exist. They are scattered. And scattered is a verdict you can reverse.

So this chapter turns the corner. It tells you what the assembled thing must look like, what an operator can do right now without waiting for the world to converge, and then it delivers the claim this entire book has been walking toward at full volume. I have spent twenty-two chapters being careful. I am going to spend most of this one being careful too, because a triumphant claim that cannot survive scrutiny is just another piece of vendor noise, and you have read enough of those. But there is a payoff at the end, and it is real, and I am going to state it plainly — once I have attached to it the one condition the whole book has been honest about, which is that it describes where the gradient points if the four families compose and get adopted, not a result anyone is entitled to assume.