Juniper Vale: You texted me something last night and then just went quiet — what was that about?
Hope Sterling: I know, I know — I fell down a rabbit hole and then it was midnight and I just, I couldn't explain it in a text. Okay so — not okay so — the thing is: Vitalik Buterin looked at Ethereum's state bloat problem and apparently put out this idea that if Ethereum used Bitcoin's UTXO model instead of its own account model, the bloat could drop by 99.8%. That's not a small fix. That's basically 'we built this wrong.'
Juniper Vale: And by UTXO you mean the way Bitcoin tracks value — the individual unspent outputs thing, not a running balance.
Hope Sterling: Exactly — like, Bitcoin doesn't store your balance anywhere. Your 'balance' is just the sum of all these individual chunks of crypto you've received and haven't spent yet. Satoshi Nakamoto designed it that way from day one. Physical cash, but digital. And Ethereum said, 'no thanks, we'll do the bank ledger version where everyone has an account with a number in it.'
Juniper Vale: Which is the choice that made Ethereum easier to build on — but also apparently a choice its own creator is now second-guessing.
Hope Sterling: And has not changed! Ten years later, Ethereum is still doing the bank ledger thing. That's — I mean, that's the thing I couldn't put down at midnight.
Juniper Vale: You know what's interesting though — I'm not sure it's actually a mistake waiting to be fixed. It might be more like, two genuinely different answers to a question nobody said out loud.
Hope Sterling: Wait — what's the question?
Juniper Vale: What should money feel like to the person actually using it — and who gets to decide that.
Hope Sterling: Okay that reframe is — yeah, I think that's why this gets complicated fast.
Juniper Vale: Because the accounting model is the answer, whether or not anyone asked the question.
Hope Sterling: But like — okay, what does that actually mean in your pocket, though? Because I think people hear 'Bitcoin doesn't store your balance' and they're like, wait, then where IS my money?
Juniper Vale: That's exactly the right question. Think of it like this — imagine you have a physical wallet. Not an app, a real wallet. And inside are actual bills: a twenty, a ten, two fives. You don't have 'forty dollars.' You have those specific bills.
Hope Sterling: Okay, yes.
Juniper Vale: Bitcoin works exactly like that. Each of those bills is a UTXO — an unspent transaction output. A discrete chunk of value you received and haven't spent yet. Your 'balance' is just your wallet adding them up. Nothing is stored anywhere as a single number.
Hope Sterling: Stop — so my Bitcoin wallet is basically doing arithmetic on a pile of old receipts every time I open it?
Juniper Vale: Basically, yeah. And then Ethereum is — it's just a bank app. One number, one account, one persistent running balance that updates in place every time something moves. That's the account model. Vitalik built that in from the start because it's genuinely simpler to reason about.
Hope Sterling: Okay but wait — if Bitcoin is the physical bills version, what happens when the bill is too big? Like if I have a twenty but I owe someone seven dollars?
Juniper Vale: You get change back. Literally. The whole twenty gets consumed as an input, seven goes to whoever you're paying, and thirteen comes back to you as a brand new UTXO. A new bill, basically, sitting in your wallet.
Hope Sterling: That is — I mean, that's actually kind of beautiful? And also insane? Like Satoshi Nakamoto designed an entire global financial system around the logic of a coffee shop transaction.
Juniper Vale: And that change output — that's a real thing with a real name. It affects how your wallet builds transactions, how fees get calculated, whether you accidentally reveal you own multiple addresses. It's not just a metaphor, it's the actual mechanism.
Hope Sterling: Wait so the change thing has, like, privacy consequences? Because if someone can see which output is the 'change' coming back to you—
Juniper Vale: They can start to map the wallet. Yeah. But that's — I mean, we should probably sit with the basic model first before we get into what breaks it.
Hope Sterling: No you're right, you're right. Okay so: Bitcoin is physical bills, Ethereum is the bank app, and the change you get back from Bitcoin isn't just flavor — it's literally how the system is built. That's the thing.
Juniper Vale: The change thing is actually where the privacy story starts, and it's also where it starts to fall apart a little.
Hope Sterling: Wait — fall apart how? Because I thought UTXOs were supposed to be the private option.
Juniper Vale: Structurally, yes. Every transaction creates new outputs instead of updating a persistent address — so in theory the history is harder to follow. But Bitcoin's transaction graph is completely public. Every input, every output, every address, the timing — all of it is on-chain.
Hope Sterling: Like — say I'm getting paid in Bitcoin. Small freelance stuff, three separate clients, three separate payments, over one month. And then I'm like, great, I'll just move all of that to one address before I send it to an exchange. That consolidation—
Juniper Vale: Is visible. All three inputs, now in one transaction, linking back to three separate sources. Anyone doing linkage analysis can see that one person controlled all of those UTXOs.
Hope Sterling: Stop. So I literally just — by trying to tidy up my wallet — I accidentally told the entire blockchain that I own all three of those payment streams.
Juniper Vale: That's coin selection. Which UTXOs you combine as inputs matters enormously. Done carelessly, it reveals ownership patterns the protocol never forced you to reveal.
Hope Sterling: So the privacy is — it's like, it's real but it's totally conditional on the user being meticulous? Which, no regular person is going to be? That's actually kind of dark because Ethereum's account model just links everything to a persistent address upfront — at least you know that going in.
Juniper Vale: Mm-hm, and that's the honest tension. With the account model there's no behavioral mitigation available — it's traceable by design, full stop. With Bitcoin you have a structural advantage that evaporates the moment you're not disciplined about it.
Hope Sterling: Which means — okay, this is the thing that's messing with me — a privacy-focused project called Midnight, which is built on the UTXO model alongside Cardano, is treating that structural property as foundational enough to build an entire privacy-first chain on top of it. Like, the structure is real, they're just betting you can engineer around the behavioral gap.
Juniper Vale: And researchers are having to grapple with the same structural gap — Chartalist, which came out of NeurIPS 2022, is the first platform that does machine-learning analysis across both UTXO and account-based datasets. Ransomware tracking, money laundering detection — they literally needed different analytical tools for each model because the data shapes are that distinct.
Hope Sterling: Wait — so even the people trying to catch criminals are like, 'these are fundamentally different problems to solve.' That's — yeah, that's not nothing.
Juniper Vale: And honestly, the privacy question is almost the smaller consequence of the architectural choice — because what it cascades into, on both sides, is a story about how each model tried to fix its own weaknesses and ended up building entire ecosystems around the workarounds. That part gets genuinely strange.
Hope Sterling: Okay the workarounds thing — that's where I want to go, because the UTXO workaround is genuinely wild to me. Like, independent UTXOs don't share any state, right? So you can validate a bunch of them at the same time. Parallel. No waiting in line.
Juniper Vale: Transaction parallelism. Yeah.
Hope Sterling: But Ethereum — every full node is maintaining this one giant unified record. Every account balance, every contract, every bit of storage. And if two transactions touch the same contract at the same time, they have to go sequential. One at a time. They can't overlap.
Juniper Vale: Picture someone buying concert tickets at 11 p.m. Saturday — big drop, thousands of people hitting the same smart contract at the same second. Ethereum's global state means those transactions are queueing.
Hope Sterling: Oh my gosh, that's — that's not a hypothetical, that's just every NFT mint that's ever happened.
Juniper Vale: And the state that every single one of those interactions generates — it never goes away. Every account, every contract, permanently in the global record. That's state bloat. It doesn't shrink.
Hope Sterling: Which is why Vitalik's 99.8% number is so — I mean, that's not a rounding error. He's saying switching to something UTXO-like would cut the state by almost all of it. And Ethereum just… didn't.
Juniper Vale: Didn't change the base layer, no. What it did instead was Plasma — a Layer 2 built on top. Move the traffic off the main chain rather than fix what's creating the congestion.
Hope Sterling: Wait — so the answer to 'our foundation has a structural problem' was 'let's build a second floor and route everyone up there'?
Juniper Vale: I mean — it's not nothing, Layer 2 is real and functional. But yeah, it treats the tradeoff as acceptable rather than fixable at Layer 1. And the UTXO side tried to go the other direction — fix their weakness instead of working around it.
Hope Sterling: The Extended UTXO model. Cardano. Okay so — Chakravarty et al., 2020, Financial Cryptography Workshops, 89 citations — they basically formalized how you add datum and validator scripts directly onto UTXOs so they can run expressive smart contracts. UTXO chains were too stateless to do multi-stage contracts before that.
Juniper Vale: Cardano built their whole smart contract foundation on it. Which is — on paper that's elegant. You get UTXO's parallelism, you get smart contract expressiveness. So why isn't everyone using Cardano?
Hope Sterling: That's — yeah, that's the thing I cannot figure out. Like, if eUTXO genuinely solves the problem Ethereum hasn't solved at Layer 1, the market should be screaming about it. And it's just… not.
Juniper Vale: Elegant architecture and ecosystem gravity are two completely different things. And I think that gap — between technically better on paper and actually adopted — that's what the whole UTXO versus account story is actually about.
Hope Sterling: And that gap is — like, that's the thing that settled for me when I couldn't sleep. It's not that one model is broken. Bitcoin bet on explicit ownership, parallelism, the physical-cash logic. Ethereum bet on expressiveness and 'just feels like a bank.' And both of those bets had real costs that just — rippled out into everything downstream. Fee design, wallet architecture, whether you need a Layer 2 at all.
Juniper Vale: The coin selection complexity, the change output management, fee spikes when you've accumulated a bunch of tiny UTXOs — that's not a power-user problem. Regular people hit that.
Hope Sterling: Right — but also the account model's answer to that is just, you don't think about any of it. Sending tokens feels like a bank transfer. And that simplicity is a genuine advantage, not a consolation prize. It's why Ethereum's ecosystem got so large so fast.
Juniper Vale: Which is the uncomfortable part of the hybrid idea — a protocol that combines UTXO parallelism with account-model expressiveness sounds elegant on a whiteboard, but it requires fundamental changes at Layer 1 and introduces its own design problems. And by the time anyone could actually build that, Ethereum's ecosystem is already enormous.
Hope Sterling: So Plasma — and all the Layer 2s that came after — they might look less like innovation and more like, I don't know, a consequence? Like if a hybrid architecture had existed early enough, maybe you don't need any of it. But now the ecosystem sitting on top of that 'less elegant' choice is worth trillions, and switching is basically inconceivable.
Juniper Vale: Architectural elegance matters a lot less once there's a trillion dollars on top of the less elegant choice.
Hope Sterling: Which is — yeah. That's where I landed. And it's kind of funny because you started this whole thing asking what money should feel like to the person using it. And I think the answer is: whichever model got to that person first.
Juniper Vale: Vitalik texted into the void at midnight and nobody changed a line of code. I find that strangely comforting.
