Vitalik recently reframed Ethereum's purpose in a way that cuts through most crypto discourse.

His question: if Signal, Starlink, and locally running open-weight LLMs are currently more directly liberating than Ethereum, what exactly is Ethereum for?

His answer: sanctuary technologies. Not systems that remake the world. Not totalizing replacements for governments or corporations. But tools that prevent total control — a shared digital space with no owner, where commitments persist without depending on a single authority.

It's a powerful frame.

But it surfaces a harder question he doesn't fully answer.

And if you've spent any time thinking about why crypto's 716 million owners don't actually use anything, you already know what that question is.

Speculation is the one use case where decentralization and product-market fit actually align.

The user wants something uncensorable, permissionless, impossible to freeze — and the architecture delivers exactly that. The properties that create friction for everything else are precisely the properties that make it perfect for this one thing.

This isn't an accident. It isn't a distraction. It's a signal.

Those 716 million crypto owners didn't fail to activate. They didn't get lost in a broken funnel. They hired crypto for one job — speculation — and that job was completed the moment they bought the asset. Crypto satisfied their job perfectly, and they left.

The industry keeps interpreting this as a distribution failure. It isn't. It's a Jobs-to-Be-Done mismatch.

The 40 to 70 million people who use crypto daily aren't more crypto-native or more tolerant of friction. They use it every day because their job stack runs deeper. Functional needs, emotional needs, and social needs — all three satisfied simultaneously, daily. When those three dimensions converge, behavior becomes automatic. When only one is present, behavior is episodic.

Sanctuary tech, as currently conceived, satisfies one job: protection.

And protection is episodic by nature. You don't need it until you do. And on the days when you don't, the friction is the whole experience.

Here is where it gets precise.

The bridge user — the NGO with compliance friction, the creator who was demonetized once, the contractor paid late across borders — carries scar tissue. That scar tissue creates genuine motivation. They understand, viscerally, why revocation resistance matters.

But scar tissue fades.

Most people never need their sanctuary infrastructure until they suddenly do. But they feel the friction every day. And on the days when nothing is wrong — which is most days — the friction is the whole experience.

Signal worked because privacy has a daily emotional payoff even when you're not under threat. The feeling of privacy has PMF. The architecture of privacy is almost incidental. Signal satisfies three jobs at once: functional security, emotional peace of mind, and social proof that you value the people you communicate with enough to protect them.

Sanctuary tech hasn't found that daily job stack yet.

The real design question isn't: how do we reach users who need sanctuary?

It's: what does sanctuary feel like on a Tuesday when nothing is wrong?

Because that's when you lose people. Not in the crisis. In the ordinary.

This is where the framing of "distribution problem" goes wrong — and why it keeps going wrong.

Distribution is not a downstream function. It is not what you do after you build the thing. Distribution starts the moment you decide what problem to solve and who to solve it for.

If you don't know on day one whether you can reach the people who have this problem, whether they feel it intensely enough to change behavior, and whether your solution satisfies more than one job for them, you haven't started building. You've started hoping.

The sanctuary tech ecosystem has largely built first and distributed later. The ideological case came before the user insight. The architecture came before the job stack. And the result is a beautiful infrastructure that serves the people who already believe, which is precisely the audience that least needs convincing.

The people who most need sanctuary — the bridge users with real institutional friction — are cautious, not ideological. They don't want to learn a new mental model. They want their problem solved. And they will abandon anything that creates more friction than it removes, regardless of its architectural elegance, on the first Tuesday when nothing feels urgent.

If daily habits emerge from three job dimensions converging — functional, emotional, and social — then the question for sanctuary tech is specific:

Functional: What does it do better than the alternative, every day, not just in a crisis? Not "it can't be censored" — that's episodic. What does it do today, when no one is trying to censor anything?

Emotional: How does it make someone feel when they use it on an ordinary day? Ownership? Autonomy? Participation in something early and important? The casino answers this question very directly. Sanctuary tech has largely left it unanswered.

Social: Who else is there? What does using it say about you? What community does it connect you to, and does that community have real stakes and real alignment — not just shared ideology?

These are not features. They are the preconditions for daily behavior. And they have to be designed in from the beginning, not bolted on after the protocol is built.

The founders who answer all three — for a specific, reachable group of people with a real, felt problem — will create the daily habits that make sanctuary infrastructure stick. Not because they marketed decentralization better, but because they understood the job before they wrote the code.

Vitalik is asking: What is Ethereum for?

The JTBD answer is more specific: what job does it get hired to do, by whom, on what day, and why would they come back tomorrow?

Until that question is answered — not philosophically, but operationally, with real users and real revealed behavior — the sanctuary vision stays at the level of essay.

The distribution problem isn't downstream of the product problem.

It is the product problem.

Build for Tuesday.

The crisis will take care of itself.

Sanctuary without daily jobs is a cathedral in the desert.

Find the jobs, and the users follow.

Find the users, and the infrastructure finally has somewhere to live.