S15 E16: Paul Sztorc Talks Bitcoin Scaling, But Can’t Say Drivechains

Paul Sztorc has been on Bitcoin Takeover so many times in three months that my girlfriend has a legitimate grievance. We had already burned more than twelve hours on the topic that shall not be named. Season 15 Episode 16 is what happens when I invite him back anyway, pour Blockstream-branded Liquid into a glass, and invent a drinking game designed to make him talk about everything else.

The rule is simple. Paul may discuss sidechains, rollups, BitVM, Lightning, covenants, Rootstock, Liquid, client-side validation, mining politics, Michael Saylor, conference theater, and the general mess of Bitcoin scaling. He may not say “Drivechain” or “BIP 300.” Every slip costs a Lightning payment to me. If Lightning fails, we both take a shot. He hates Lightning. I have Phoenix open. The vodka is ready. Welcome to the drunkast.

If you want the version where he is allowed to name the thing, start with my earlier write-up on Baltic Honeybadger and BIP 300, or the later defense of Roger Ver. This episode is the comedy special about the rest of the menu.

As Good as Cash: What Even Counts as Layer Two?

Paul opens with Leonardo DiCaprio in Catch Me If You Can, waving a cashier’s check and insisting it is as good as cash. That is his Layer Two definition in one scene: you do not hold an L1 UTXO. You hold something else that people treat as redeemable one-to-one. You gain convenience. You lose certainty.

He walks through Arc-style VTXOs, Lightning balances, federated pegs, and anything that markets itself as “Bitcoin but elsewhere.” The important part is the peg psychology. Eleven BTC of claims that trade at par with eleven BTC of mainchain coins are Layer Two in his framing, even when the security story is a third party, a federation, or a hashpower process. Monetary belief first. Cryptography second. Marketing third.

I push him to separate the monetary layer from the technical layer. He refuses to pretend those are cleanly separable. If users redeem at par and transfer among believers, the marketing already won half the argument.

Rootstock Is Trying. Liquid Is Comfortable.

We use Rootstock and Liquid as the morality play. I had just interviewed Sergio Demian Lerner; that conversation is S15 E7. My take on air: Liquid looks static. It accepts the federation. It does not pretend to evolve into something trust-minimized, and (on their own published story) fees do not flow to Bitcoin miners. Rootstock still uses HSMs today, but Sergio keeps chasing trust minimization, merge mining, and even BitVM experiments. It also pays miners. Paul agrees that fee pass-through is not a branding choice. In his view it is the only long-term category that survives.

That split becomes the episode’s quiet thesis. Some L2s are hotels: nice lobby, fixed staff, no ambition. Others try to become neighborhoods attached to Bitcoin’s security budget. Paul works at LayerTwo Labs. You can guess which neighborhood he prefers. You just cannot hear him name the street during this recording without a Lightning invoice firing.

Covenants: Fashion Week for Soft Forks

Wasabi gets its sponsor beat, then we slam into covenants, the soft-fork costume everyone was wearing in early 2024. Paul’s short version: normally you lock coins to a key. With a covenant you lock them to future behavior. The spend path itself is constrained. Old joke name: “the check is in the mail.” Jeremy Rubin’s CTV is the limited, trailblazed version. Paul says Jeremy deliberately kept the scope small so critics would run out of imaginary disasters. Then “everyone” (by which he means roughly three people) decided they wanted more expressive toys like TXHASH anyway.

He names the usual suspects for the expressive camp: Steven Roose, Robin Linus, Burak from Ark. He also says most Twitter covenant discourse is people repeating a trending topic without understanding the mechanism. I remind him that Taproot activation sold “Bitcoin finally has smart contracts” to an audience that still thought Schnorr was a furniture brand. He agrees. Fashion is not a security model.

Where covenants do matter for him is Arc and similar designs that are awkward or impossible without something CTV-shaped. Scaling narratives that treat covenants as magic Lightning steroids get less sympathy. Mechanism design first. Hashtag later.

Soft Forks, Miners, and the Expert Shortage

This is where the drunkast gets serious. Paul argues that soft-fork activation is not a theology problem for Twitter theologians. Miners can coordinate signaling. Users who break the heaviest-valid-chain rule are the ones hard-forking onto a new network, not the miners who stay on the old validity rules. He says Luke Dashjr and Adam Back are basically the named Bitcoin experts who disagree with that framing, and he wants more people to show up in person if they claim otherwise. At TABConf, Buck Perley told him going to miners for activation was the correct move.

SegWit nostalgia gets roasted. The 2017 sales pitch was Lightning enablement and malleability fixes. Later it got rewritten as a block-size increase story or a culture war trophy. Taproot got sold as privacy for multisig. Paul is allergic to the pattern: ship a soft fork under one narrative, then pretend the narrative was the consensus. He even toys with the idea that soft forks can be deactivated in principle, which is the kind of sentence that makes timeline people open a new Excel sheet titled “heresy.”

Ocean, Rollups, and Lightning as Custodian Cosplay

Ocean mining and inscription drama show up as proof that the system keeps working even when powerful people hate a use case. Then I ask the Argentina question: are sidechains obsolete now that sovereign ZK rollups are fashionable?

Paul’s answer is not diplomatic. Sovereign rollups, in the form he thinks people mean, are closer to inscriptions theater: Ethereum-style systems renting Bitcoin block space without the fee and security story he wants. He would rather see experiments that pass fees to miners and keep the peg honest.

Lightning gets the longest knife. Paul has been saying for years that a majority of hashpower can censor justice transactions and effectively custodian channel balances. I press him on whether “miners” are a fixed cast of villains. He insists mining is a process, not a permanent club, but the economic point stands: if justice depends on timely inclusion, censorship is custody with extra steps. That is why the drinking game uses Lightning payments as punishment. The bit writes itself.

Almost Said It: Zucco, CSV, and Pouring One Out

Around the two-hour mark he nearly says the forbidden words while talking about Giacomo Zucco’s Bitcoin Amsterdam debate. He catches himself. We drink anyway, because the joke is the point. His read of that debate is unkind and funny: more tone policing and RGB competition anxiety than a clean technical disagreement. Peter Todd almost triggers another fine when Paul frames Todd’s long sidechain history as wanting to invent something better than the proposal we cannot name. He starts the sentence, realizes, and reaches for the glass.

Client-side validation, node cost, and “many chains of different shapes” get a clearer stretch. Paul wants a universe where privacy chains can be more expensive and more paranoid, while decentralized-enough cheap chains handle mundane throughput. Geographic specialization. Not one sacred 1 MB morality play for every human activity.

We pour one out for Calvin Ayre’s Craig Wright spending. Hampus from Sparrow is in the chat under a nickname. Lobster ravioli is involved. It is 3 a.m. in Romania and somehow still going.

Saylor, Conferences, and the Subversive Edge

I read Michael Saylor’s older skepticism about Bitcoin as a payments network and ask what happens when the loudest institutional bull also builds soft self-regulation for North American miners. Paul is less interested in Saylor-as-Satoshi fan fiction and more interested in the cultural shift: privacy becomes taboo, ETF narratives dominate, and the subversive edge of Bitcoin gets sanded down into a treasury product. Conferences pay speakers to validate the room. Content follows the check.

That complaint is also self-aware. Paul is a public figure now. Someone in chat jokes that he compared himself to Taylor Swift while explaining why he cannot owe every stranger two minutes of attention. He means first-principles extremes. The clip will live forever.

BitVM, Try Everything, Cost-Benefit Theater

Near the end we get to BitVM. Robin Linus had joked on Twitter that BitVM is a way to get the forbidden BIP-shaped outcome without the soft fork. Paul likes the ambition and hates the opacity. Super Testnet warned him about brutal resource requirements. Sergio thinks BitVM can approximate the unnamed design. Paul’s institutional answer from LayerTwo Labs is the least tribal sentence in the episode: try all of it. Do not lie about rival designs to prop up your own. Wait for the concrete to dry before declaring winners.

He is also annoyed that soft-fork debates rarely do cost-benefit analysis. Any code change can introduce bugs. Soft forks are not uniquely radioactive; they are uniquely political. That is the dysfunction.

We close on ancient mining trivia: pool insurance, pay-per-share, Jet, the Bitcoin Mining Council, conflicts of interest, and whether listeners should optimize for quality over quantity. Someone still watching at 3:30 a.m. Bucharest time asks a question. Paul answers it. The vodka has done its job. The forbidden words mostly stayed forbidden. Lightning may or may not have.

Find Paul via Truthcoin, the original Drivechain essay he was not allowed to advertise on this episode, LayerTwo Labs, GitHub, and @Truthcoin. For the BIPs he kept dodging aloud, see BIP 300 and BIP 301. And if you only remember one thing from S15 E16: the best way to make Paul talk about every other scaling idea is to fine him every time he names his own.