Jack Dorsey on Nation-State Bitcoin Reserves, Mining Centralization, and Why “Digital Gold” Is a Failure Case

In an exclusive Hosting Bitcoin interview, Jack Dorsey argued that a Bitcoin used only as a reserve asset is a failed project, that ASIC manufacturing is the network’s most dangerous single point of failure, and that the only durable mining strategy pairs efficient power with data centers flexible enough to serve other kinds of compute.

Jack Dorsey is the co-founder and two-time former CEO of Twitter and the founder of Block, the parent company of Square, Cash App, Bitkey and Proto. Proto is Block’s open-source Bitcoin mining chip and rig program, which puts Dorsey in the unusual position of commenting on mining economics as a manufacturer rather than as an observer. Episode published 16 January 2025. Runtime 45 minutes.

The claim that should reframe how operators talk about Bitcoin

Most interviews with a figure of Dorsey’s profile produce agreeable generalities. This one did not. Asked directly about nation-state Bitcoin reserves, the answer cut against the position most of the industry was celebrating at the time.

“I think it’s good for the nation state. I don’t know if it’s necessarily good for Bitcoin. It’s probably good for Bitcoin holders in the moment that the Bitcoin is bought. But if this ends up just being digital gold, I think it’s failed.”

Jack Dorsey, Hosting Bitcoin

He anchored it to the original document rather than to sentiment.

“You have to go back to the white paper. This is peer to peer electronic cash, and we have not realized that yet. That’s why I’m in it. I’m not in it to create digital gold and just watch a number go up.”

Jack Dorsey, Hosting Bitcoin

For anyone who owns hashrate, this is not a philosophical aside. It is a statement about where the network’s long-run security budget comes from. A Bitcoin that settles into a purely custodial reserve asset generates thin fee revenue and leaves miners dependent on the subsidy schedule. A Bitcoin that carries real payment volume does not. Dorsey is arguing, in effect, that the payments case and the mining case are the same case.

The single point of failure most miners never price

Block’s move into mining hardware is usually read as a product decision. Dorsey framed it as a network-risk decision.

“If we want Bitcoin to succeed, we need to remove single points of failure and centralizing efforts. The mining hardware is one of those. You have effectively one company who’s building all this hardware, they’re out of China, and that is a single point of failure that can really damage the network.”

Jack Dorsey, Hosting Bitcoin

He then described the commercial opening underneath the risk, which is more interesting than the risk itself: the incumbent’s customers are not satisfied. Dorsey said the miners Block talked to had “significant reliability issues,” could not customize for their applications, and were not happy with performance. Proto’s stated design goals follow directly from that list, and Dorsey described the target as a rig anyone can take and design their own system around, whether the application is energy recapture, a standard mining deployment, or a consumer heat product.

The practical read for an owner is that hardware supply is becoming a variable rather than a constant. Machine selection has been a one-vendor decision for most of the industry’s history. It is starting not to be.

What Dorsey thinks separates the miners who survive

Asked what distinguishes operators who last from those who do not, over the next decade rather than the last one, Dorsey gave the most directly operational answer in the interview.

“It feels like it’s a strategy between finding the right and efficient power sources and then having flexibility in your data centers to maybe do things inclusive of AI computation. So having that sort of flexibility, knowing that you have these giant warehouses and you have access to significant power, having strategic sense of how to make that better and also the flexibility feels like the only winning combination.”

Jack Dorsey, Hosting Bitcoin

Two things are worth separating out of that sentence, because they are often collapsed together.

The first is power. Not cheap power in the marketing sense, but the right power source: a supply whose cost structure holds through a full cycle rather than one priced off a promotional window. The second is optionality at the facility level. Dorsey is describing sites that can redirect capacity toward other compute demand when that is the better use of the same electrons, which is a real constraint on how a building should be designed rather than a slogan.

This is also the point in the conversation where an operator’s own position becomes testable. Abundant Mines runs on Oregon hydroelectric power and builds for redirectable capacity, which is the combination Dorsey names. That is a claim you can check against our published Oregon hydro cost structure and our hosting cost breakdown rather than take on faith.

Nation-states mining: neither the threat nor the win it gets sold as

“I think it can go either way. Bitcoin works because it’s both a consensus and there’s adversarial parties who are trying to remove and de-risk it. So a government or a corporation, there can be centralizing forces, and the counters against that are the developers and the node runners. As long as the network is hyper aware of what can be centralized and what is being centralized, and understanding that risk, I think that’s pretty critical.”

Jack Dorsey, Hosting Bitcoin

The framing here is worth keeping. Dorsey does not treat state participation as a binary. He treats it as a force that has to be continuously counterweighted, and he names the counterweight: independent developers and independent node operators. Distributed private hashrate belongs on that list too. Every individually owned machine on a network is a vote against the concentration he is describing.

Where this leaves an owner

Three durable takeaways, stated as Abundant Mines reads them rather than as Dorsey said them.

Hardware concentration is a real risk with a real remedy in progress, which means machine-selection strategy is about to matter more than it has. Power quality and facility flexibility are the two variables that survive a cycle, and both are decided at the site level before a machine is ever plugged in. And the payments thesis is not separate from the mining thesis: a network that carries transactions is a network that pays its miners from fees rather than from a schedule.

If you are weighing whether to own hashrate at all, our guide to hosted Bitcoin mining covers the structural decisions in order. If you already own machines and are evaluating where they should sit, the hosting comparison is the more useful starting point.

Browse every episode of the Hosting Bitcoin podcast

Quotations are drawn from the episode audio and lightly punctuated for print. Wording is unchanged.