Both models widen access to bitcoin. Only one of them lets the user keep the keys — and that difference is the whole argument.
Two stories crossed our desk this week that looked unrelated and were, in fact, the same argument seen from opposite ends. Put them side by side and you can watch the future of bitcoin adoption split into two roads.
In Nairobi and Juba, a working group of African developers wired mobile money to bitcoin using open-source software that no company or government controls. Each national gateway keeps custody of its own funds; the coordinating server, by design, cannot touch them; and value crosses the Kenya–South Sudan border at zero routing fees. The stated goal was not efficiency. It was sovereignty — rails that don’t depend on closed servers or custodians off the continent.
In Hanoi, a government announced it will permit bitcoin trading starting in September — through up to five licensed firms, only in the national currency, behind a capital wall of roughly $380 million. The stated goal there was not sovereignty. It was control — a supervised market the state can see, tax, and hold accountable.
It is tempting to file both under a single happy headline: bitcoin goes mainstream. But the two models differ on the one axis that has always mattered. The African rail is built so the user — or at least the local operator serving her — holds the keys. The Vietnamese market is built so a licensed intermediary holds them for you. One expands the number of people who can transact without asking permission. The other expands the number of people who can transact with permission.
None of this makes the top-down model foolish. Most people do not want to run a node or memorize a seed phrase; they want an app that works and a number to call when it doesn’t. Licensed exchanges bring consumer protection, tax clarity, insurance, and banks that are unlikely to vanish overnight. This very newsletter spent last week documenting what happens when self-custody goes wrong: a firmware bug that drained $116 million from people who did everything right. A regulated venue would have carried different risks, but at least a phone number to call.
And the bottom-up model has its own fragilities. Open-source rails run on the goodwill of small teams and the uptime of coordinators; they must still answer to the same regulators — Kenya is moving to report crypto users individually — that the permissioned model courts. Sovereignty at the protocol layer does not guarantee it at the legal one.
Here is the asymmetry that decides it. A licensed exchange is a switch, and switches can be flipped — by a regulator, a sanctions list, a change of government, a bank pulling service. The Vietnamese market that opens in September can be narrowed or closed in the same document that opened it. The African rail, if it works as designed, cannot: there is no company to lean on, no custodian to freeze, because the funds never leave the operator’s hands or the continent.
That is why the Nairobi hack week matters more than its modest scale suggests. Anyone can build an app that moves bitcoin. Building one that no one can switch off is the harder thing — and the only version that keeps bitcoin’s original promise intact. Permission is convenient. It is also, by definition, revocable. The rail you control is the only one that is still there on the day someone wishes it weren’t.
Both roads carry more people to bitcoin. Ask which one still runs when the people in charge change their minds.
Opinion. The Take is the newsroom’s argument, not a news report, and airs the strongest case against its own thesis on purpose. Nothing here is financial advice.
Free. Five minutes. No hype.
Subscribe free