Field notes · Stablecoins & money movement
Stablecoins are becoming invisible infrastructure: a dollar balance in an app that feels just like a bank. The thing that disappears along with the word "crypto" is the safety net the user still assumes is there.
What I've come to believe
Stablecoins are quietly becoming the best money-movement rail anyone has built: borderless, near-instant, near-free, and on the verge of vanishing into ordinary apps so completely that users will not know they are there. That invisibility is the goal, and it is a real achievement. It is also the problem, because an invisible rail still has every moment where money goes wrong, and the disappearing act takes the recourse with it while leaving the user feeling exactly as protected as they did at a bank.
Henri Stern of Privy has a good line: stablecoins are "Starlink for money." Once value is up in orbit it moves with almost no friction, and the hard part is the ground stations, the on-ramps and off-ramps where crypto meets banks and regulation. Stripe's acquisitions of Bridge, the orchestration layer, and Privy, the wallet layer, are a bet on owning those ground stations and turning janky blockchain primitives into clean, compliant, invisible APIs. The convergence thesis underneath it is that there will be no separate "crypto economy," just one rail where the technology underneath is irrelevant to the user. I think they are mostly right, and the use cases are real: remittances that skip the chain of SWIFT hops, dollar access for people whose own currency is failing them, corporate treasury moved home without predatory spreads. This is not a scam-coin story. It is an infrastructure story, and a good one. Which is exactly why the part it leaves out is worth naming.
The stated goal, in the notes and on every roadmap, is for stablecoins to recede into the fabric of reality the way cloud computing did, until the user never needs to know they are "using crypto." Here is the thing to hold onto: two different things are being made invisible at once, and they are not the same thing. The first is the crypto plumbing, the gas tokens, the private keys, the chain names. That should disappear, and good riddance. The second, quietly riding along, is the entire consumer-protection layer a regulated bank rail carries: reversibility, chargebacks, dispute resolution, fraud recovery, a hardship path, a human on the other end when life breaks.
On an invisible rail the user gets an interface that looks and feels precisely like a bank app, a dollar balance and a send button, sitting on top of a settlement system that is final, irreversible, and has no other end to the line. The screen promises a bank. The rail delivers cash in an envelope that, once sent, is simply gone.
The crypto disappears, which is good. The recourse disappears with it, and the user is never told.
The two things going invisible are not the same. One is the crypto plumbing, which should disappear. The other is every protection a bank rail carries, which the user still assumes is there.
My take
I read every financial system by asking what happens to the person on their worst day with it, and stablecoin infrastructure is the cleanest test of that question I have seen in a while, because the entire design goal is for that person to never know what they are standing on.
Here is the moment I keep picturing. Someone sends the wrong amount, or sends to an address a scammer handed them, or has the account drained, on a rail that feels exactly like their bank, because that was the whole point. At a bank, that day sets off a system: a dispute line, a chargeback, a fraud team, a provisional credit, a human who can sometimes claw it back, and underneath all of it the quiet assumption that the institution carries some of the risk with you. On an invisible stablecoin rail, the same day sets off nothing, because there is nothing to set off. The money is final. There is no other end of the line. And the part that gets me is that the user had every reason to believe otherwise, because a decade of design effort went into making it feel safe.
I want to be precise, because I am not anti-stablecoin, and the DeFi version of this argument is too easy. The old crypto pitch at least had the honesty of looking dangerous. What is new, and what worries me, is the polish. Stripe and the orchestration layer are very good at making this invisible, compliant, and frictionless, and those are real virtues. But frictionless and final is a specific combination, and it is the most dangerous one for the person who can least afford a mistake. Inclusion that reaches the vulnerable and deletes their recourse is not simply a gift. It is access and exposure handed over in the same envelope, with only the access part written on the label.
When the rails were ugly, you knew you were on your own. The whole achievement now is that you will not, right up until the day you need someone and find there is no one there.
The trajectory is real. By the late 2020s stablecoins probably do recede into the fabric, the invisible superpower the notes predict, and most of the time that will be wonderful, cheaper and faster and more global than anything banks ever shipped. But "most of the time" is carrying a lot of weight, because the back end is the business of the times it goes wrong, and that is the part the convergence thesis is quietest about. The version of this future worth wanting is not the one where crypto disappears. It is the one where, as it disappears, somebody rebuilds the protections it left behind: reversibility where it counts, a real dispute path, a human on the other end, a hardship route for the person whose stable dollars just vanished. Invisible rails do not remove the need for a back end. They only make it easier to ship without one, and easier not to notice who is standing there when it breaks.
Threads worth pulling: notes synthesized from stablecoin-infrastructure material, including Henri Stern (Privy) and Zach Abrams (Bridge), Stripe's acquisition strategy and the Tempo payments-chain thesis, the GENIUS Act and MiCA, and remittance and neobank use cases such as Felix Pago and DolarApp. Claims here are paraphrased from my notes and are my own synthesis, read through the consumer-credit back end I work in.