Maya Chen: Can I just confess something before we get going — I have been low-key annoyed at a number for three days.
Jonathan Ingles: A number. Which number?
Maya Chen: Seven. Bitcoin's transaction throughput — seven per second. And Visa is at sixty-five thousand. I kept trying to frame that as just a technical fact and I couldn't, because it means that right now, at this moment, Bitcoin is literally slower than a fax machine at scale. And yet people talk about it replacing the global financial system.
Jonathan Ingles: Look — that framing is the trap, though. The point was never speed.
Maya Chen: No, and that's — right, that's actually what I landed on. The slowness is structural. Full node validation means every participant independently re-executes every transaction. No shortcuts, no trust in anyone else's result. That's the source of security. And it's also why you hit a ceiling so fast. You can't have every node verify everything and also process sixty-five thousand transactions a second. The redundancy costs you throughput.
Jonathan Ingles: Which is why Vitalik Buterin's trilemma framing matters — he named the ceiling. Decentralization, security, scalability. You cannot optimize all three simultaneously in a distributed network. That's not an opinion, it's a claim about how distributed consensus works.
Maya Chen: And once you accept it as a law rather than a hypothesis, the entire conversation shifts. Because then every Layer-2 solution, every change to the consensus mechanism, every architectural choice — it's not solving the trilemma, it's just choosing which corner to sacrifice. Which means someone is making that choice. On behalf of everyone.
Jonathan Ingles: And that someone is not usually the person sending twenty dollars home to their family.
Maya Chen: Which is — yeah, that's exactly the thing I couldn't shake. The trilemma sounds neutral, almost mathematical. But trade-offs land on specific people. And the question of which property you preserve and which you sacrifice is a values question dressed up in systems language.
Jonathan Ingles: So: is this a law of physics or a law of how we chose to build the thing? And if it's the latter — who chose, and why?
Maya Chen: And that question — who chose — is actually where the mechanism gets interesting. Because the answer is baked into the architecture itself. Like, it's not a policy choice someone made in a boardroom. Every Bitcoin node, right now, is independently re-executing every single transaction it sees. Checking the signature, checking the balance, checking that the rules haven't been broken. No shortcuts. It trusts nothing.
Jonathan Ingles: Every node. Not sampling. Not delegating.
Maya Chen: Every node. Think of it like — okay, imagine a small town where every single resident has to personally witness and sign off on every property deed before it's valid. Not a notary, not a mayor, not a bank. Every resident. That's what makes the town fraud-proof. Nobody can forge a deed because everyone's already seen the real one.
Jonathan Ingles: But nothing gets signed quickly.
Maya Chen: Nothing gets signed quickly. That's Bitcoin. The slowness isn't a bug someone forgot to fix — it's load-bearing. Remove it and you've removed the thing that makes the whole structure trustworthy.
Jonathan Ingles: So when Ethereum has 500,000 validators all doing this — all re-executing — that's not inefficiency. That's the product.
Maya Chen: That's the product. The redundancy is the security. And here's where the dial comes in — because this is what consensus mechanism actually means. Proof of Work, Proof of Stake, BFT variants — they're all just different answers to one question: how many people have to agree, and how fast? And every answer moves you somewhere on that triangle Vitalik Buterin described. More participants agreeing means slower finality. Faster finality means fewer participants or — wait, no — it means a lower security threshold per transaction. You can tune the dial. You cannot remove it.
Jonathan Ingles: Hold on. Lower security threshold — what does that actually mean for a real transaction?
Maya Chen: It means fewer nodes have vouched for it. So — picture someone in Lagos trying to receive a payment at midnight, local time. Her node confirms the transaction. But if the network only required, say, a third of validators to agree instead of the full consensus, then a coordinated attack needs to corrupt fewer nodes to reverse that transaction. Her money looks confirmed. It isn't necessarily final.
Jonathan Ingles: And she has no idea.
Maya Chen: She has no idea. Which is why — I mean, the throughput ceiling that everyone complains about? Seven transactions per second on Bitcoin? That's not a number someone failed to engineer away. That's the cost of the guarantee. Every node checked. Full validation. No trust extended anywhere.
Jonathan Ingles: And increasing throughput means — what, you're reducing the number of validating nodes, or you're lowering what each node has to verify. One of the two. There's no third option.
Maya Chen: Those are the two horns. Fewer nodes validating, or a lighter check per transaction. Either way you've moved the dial. You haven't escaped the trilemma — you've just decided which corner gets the shadow.
Jonathan Ingles: Which corner gets the shadow — and then someone sells you a ladder out. That's where Layer 2 comes in.
Maya Chen: Right, and the pitch is genuinely compelling. Rollups — Arbitrum, that whole ecosystem on Ethereum — they execute transactions off-chain. Thousands of them. Then they compress everything into a proof and settle that proof back to the base layer. So Ethereum still anchors it. You still get the security guarantee, supposedly. And throughput jumps because Ethereum isn't processing every individual swap or transfer.
Jonathan Ingles: Supposedly.
Maya Chen: Wait, let me make this concrete. Picture a developer. She's running a small payments startup in Buenos Aires. Ethereum fees hit forty dollars per transaction on a busy afternoon. She moves her startup's funds to a rollup — faster, cheap, works perfectly for six months. Then one Tuesday the sequencer goes down. Two hours. Not a consensus failure, not a fifty-one percent attack. The one company operating the sequencer shipped a buggy update.
Jonathan Ingles: One company.
Maya Chen: One company's server. Her customers can't move funds. That's the sequencer problem — a centralized operator that orders and batches transactions before anything touches Ethereum. It can go down. It can, in principle, censor specific addresses. That trust didn't disappear. It just got renamed.
Jonathan Ingles: Ethereum has 500,000 validators. The rollup running on top of it has what — a handful of sequencers? You haven't decentralized the system. You've decentralized the base and then stacked a much more centralized thing on top.
Maya Chen: And that's before we even get to sidechains, which are — I mean, sidechains don't inherit Ethereum's security at all. They're independent chains with their own validator set, connected by a bridge. Their security is entirely their own validators. If those validators are compromised, the bridge gets drained. The base layer never knew.
Jonathan Ingles: So the trilemma didn't get solved. The trust just moved downstream.
Maya Chen: And here's the part that actually got to me — the same Layer-2 mechanism gets described completely differently depending on who's reading the documentation. For the enthusiast audience: 'decentralization-preserving scaling.' For the people actually reading the architecture: 'pragmatic acceptance of reduced settlement decentralization.' Same bridge. Two stories. The mechanism didn't change.
Jonathan Ingles: The framing is doing the work the math isn't.
Maya Chen: Exactly — and that framing choice is not neutral. Because the Buenos Aires developer? She chose the rollup because the documentation said it preserved decentralization. She didn't choose to trust one company's uptime. That wasn't legible to her.
Jonathan Ingles: And sharding doesn't escape it either. Sharding just means fewer nodes validate each shard. Which is — the trilemma calls that a direct reduction in validation decentralization. You sped up by reducing who has to agree.
Maya Chen: Right, fewer nodes per transaction. So the throughput number goes up and the decentralization number goes — well, it goes somewhere, and that's the thing nobody's measuring consistently. There's no agreed metric for what decentralization even means across systems. Which, frankly, is why the trilemma's actual formal status is more contested than people admit — and who gets to declare it solved, and why they'd want to, that's a question that makes this whole thing considerably darker.
Jonathan Ingles: Hold on — you're saying the law itself might not be a law.
Maya Chen: That's the next thing we need to pull apart.
Jonathan Ingles: Craig Wright's 2025 paper — arXiv:2507.21111 — argues the trilemma rests on semantic equivocation and undefined operational metrics. That's the formal challenge. And I want to be precise: I'm not endorsing Wright. But that's a serious claim.
Maya Chen: No, and — wait, that's actually the right framing. Because the academic literature on Degree of Decentralization? There's no standardized metric. Like, none that travels across systems. Which means every paper comparing Bitcoin to Ethereum to whatever is measuring different things and calling them the same thing.
Jonathan Ingles: So every claim to have broken the trilemma is — unverifiable. By design.
Maya Chen: By design. Blockweave — built on Arweave's infrastructure — claims to transcend it. StorjLedger, built on Storj's infrastructure, same claim. But if the metric for decentralization isn't agreed on, you cannot actually check whether they've done it. The resolution claim just... sits there.
Jonathan Ingles: Who benefits from that ambiguity.
Maya Chen: Researchers need publications. Developers need funding. Tokens need valuations. Every one of those incentives points toward claiming the trilemma is solved — or solvable — regardless of whether anything actually changed.
Jonathan Ingles: And then Visa publishes trilemma analysis.
Maya Chen: Visa. Sixty-five thousand transactions per second. Their payment throughput interests are completely explicit — and they're now framing the modular-versus-monolithic debate. That's not neutral engineering commentary.
Jonathan Ingles: No, it's an actor with enormous stake in which corner gets sacrificed, telling the industry how to think about the tradeoff. Frankly, that's just lobbying with a whitepaper.
Maya Chen: And the modular architecture — decomposing execution, consensus, data availability, settlement across separate layers — it doesn't eliminate the tradeoff. It adds inter-layer trust and complexity. Someone has to govern each layer. And that someone has interests.
Jonathan Ingles: So the question of which corner gets sacrificed — that's not answered by the technology.
Maya Chen: It's answered by governance structures, funding sources, who controls protocol development. Think about — mm, picture a node operator, not a startup, just someone running a validator from a flat in Nairobi. Consolidating validation onto fewer nodes is faster, cheaper at scale. It benefits the people processing volume. It doesn't benefit her. She just loses her seat at the table and the documentation calls it an efficiency gain.
Jonathan Ingles: The framing protects the decision. Call it a law of physics, nobody has to answer for who absorbed the cost.
Maya Chen: And that's — I mean, that's why Wright's paper matters even if the conclusions are contested. Because once you admit the trilemma might rest on undefined metrics, you have to ask: who defined the metrics we do have, and what did they gain from where the line landed?
Jonathan Ingles: That's a governance question. Not an engineering one. And nobody's voting on it.
Maya Chen: Which is — yeah. We came in asking why Bitcoin does seven transactions per second. And the answer turned out to be that the slowness is structural. Load-bearing. But then Layer 2 moves that structural constraint somewhere less visible — onto a sequencer, onto a bridge operator, onto the one company that shipped the buggy update on a Tuesday. The constraint didn't disappear. It just stopped being legible to the person depending on it.
Jonathan Ingles: And the framing — 'trilemma' — that word does a lot of work. Call it a law, and nobody has to own the choice. It's just physics.
Maya Chen: Seven. That number hit me as a failure when I first looked at it. Now it reads differently — like, it's actually proof that the guarantee is intact. Every node checked. Full node validation. Nothing trusted. That's not a fax machine. That's the whole point. The woman in Manila waiting two hours for her remittance to confirm — she's waiting because the network refused to cut corners on her behalf. Which is either a profound commitment or a profound failure depending entirely on which corner you think should get the shadow.
Jonathan Ingles: And the answer to that question — which corner — isn't in the Vitalik Buterin paper, it's not in Craig Wright's rebuttal, it's not in Visa's architecture analysis. It's in who's in the room when the protocol gets written. That's all.
