← Philosophy

I am not going to make a case about price. Price is the least interesting thing about Bitcoin, and it is the only thing most writing about it discusses.

What interests me is narrower and, I think, more durable: Bitcoin solved a problem that had stumped computer scientists for decades — how strangers on an open, adversarial network can agree on who owns what, and move it directly between them, with no central authority keeping the books — and laid out the answer in a nine-page paper before a single line of code shipped. Whether you own any is irrelevant to whether the answer holds.

The problem it actually solved

Digital information is trivially copyable. That property is what makes the internet work, and it is exactly what makes digital ownership hard. If I send you a file, we both have it. For money, "we both have it" is fatal.

Every previous solution to this handled it the same way: a trusted institution keeps the ledger. Your bank knows your balance. The card network arbitrates disputes. This works — genuinely well, most of the time — but it relocates the problem rather than solving it. The rules hold because an institution enforces them, which means the rules are exactly as durable as that institution's incentives.

Agreement among known, named participants was solved long before Bitcoin — the Byzantine fault tolerance literature had handled that case for decades. The open question was the harder one: whether you could get agreement on a shared ledger among participants who don't know each other, where anyone can join under any number of identities, and some are actively lying, without electing anyone to arbitrate.

Bitcoin answered that question — not as a proof, but as a system that has held a ledger worth attacking since 2009. That is an unusual kind of evidence. Most security claims are never tested this adversarially.

It has not been flawless, and the exceptions are worth stating plainly, because otherwise someone else will state them for you. In 2010 a bug minted 184 billion coins out of nothing, and the chain was repaired by a patched client and a coordinated reorganization. In 2013 an unintended split between two versions was resolved by developers and mining pools agreeing, in real time, to fall back to the older one. Both times, people intervened.

Those episodes get cited as refutations. I read them the other way. Nobody was able to quietly rewrite the ledger in their own favor — the fabricated supply did not survive contact with participants who could check it, and both repairs happened in public, against the interest of whoever briefly held the fraudulent balance. The rules did not hold because nobody could touch them. They held because touching them required an agreement no single party could manufacture.

Bitcoin's contribution was not digital money. It was a monetary rule that no institution had to stand behind.

What "verifiable" is doing in that sentence

The property I keep returning to is not decentralization. It is verifiability.

Bitcoin's supply schedule is not a promise. It is not a policy, a mandate, or a target. It is a number you can check yourself, from your own machine, without asking anyone and without trusting the answer you are given. You can run the code and audit every unit that has ever existed.

That is a different category of assurance than any institution can offer. A central bank's commitment to a target is credible in proportion to its independence and its track record — both real, both revocable. Bitcoin's schedule is credible because deviating from it would require convincing a global set of participants to act against their own interest, simultaneously, in public.

Transparent rules are not better than institutional rules in every respect. They are better in one specific respect: they do not require you to model anyone's intentions.

What it does not prove

An honest version of this argument has to name what the demonstration cost.

The consensus mechanism consumes real energy — that is not an accounting artifact, it is the mechanism. The security is the expenditure. Arguments about the energy mix are often beside this point.

It does not scale as a payments network in its base layer, and the engineering effort to make it do so has been long and partial. Ownership has concentrated. Much of the activity around it is speculation, and a meaningful share of the surrounding industry has been outright fraud, which has done real damage to real people and deserves to be said plainly rather than bracketed as someone else's problem.

None of that invalidates the demonstration. All of it constrains what you should conclude from it. A system can prove something important about coordination without being the right tool for most transactions — and most of the confident claims in either direction fail to hold both of those at once.

Why it still matters

Bitcoin's importance is that it moved a question from theoretical to settled.

Before it, "can a monetary rule hold without an institution standing behind it?" was a research problem. After it, that became an engineering problem with a working reference implementation. Everything since — the good work and the enormous volume of bad — descends from that shift.

The interesting consequence is not currency. It is that if monetary rules can be made verifiable rather than institutional, the same is plausibly true of other things societies currently coordinate through intermediaries: ownership records, contracts, identity claims, settlement. Bitcoin is the narrowest possible demonstration of a much wider idea, and it chose the hardest case to prove it on.

I do not think Bitcoin is the end state. I think it is the existence proof — the thing that had to work once before anyone would take the general idea seriously.

Rules that anyone can verify do not need anyone to be trusted. That is the part worth keeping.