S16 E27: David Seroy on Alpen Labs & ZK Rollups on Bitcoin

David Seroy runs the non-technical side of Alpen Labs, a company founded by four MIT alumni who all come from Nepal, and the name is a hint: "if you rearrange the letters, it spells Nepal." We recorded with Bitcoin at $109,000, the stream ran 2 hours and 29 minutes, and David moved a business development meeting back by an hour so that I could keep asking questions. Ribbit led the round, with Wences Casares, Uma Roy of Succinct and Jeremy Rubin among the other backers.

I wanted him on because Alpen is one of the few teams trying to put ZK rollups on Bitcoin without a federation holding the coins, and because David is willing to argue the unpopular side of most things that come up on this show. The conversation started with origin stories and the name, wandered through Knots, Lightning, stablecoins and a fight on a Las Vegas stage that never happened, and ended on a disagreement I did not manage to close. The pickleball tournament Alpen won also came up, though I had no strategy for that one.

Four Nepalis, a Grant and an Anagram

Alpen's founders started out on research, not on a product. John Light had a grant from StarkWare and the Human Rights Foundation to study sovereign rollups, and the Alpen founders took that work and widened it. Then Robin Linus developed BitVM, and inscriptions made it possible to put ZK verifiers on Bitcoin in practice. David described the position of anyone using a rollup bridge before that as stark: "you would essentially have to give up your Bitcoin to someone else."

The name came out of a long naming process, and David has no illusions about how that goes: "naming is not easy. It's easier to build the tech than, you know, to name something." I pointed out that Citrea is a similar story, since it is the Turkish founders' word for orange. I had Orkun Kilic on in S15 E18 about the first ZK rollups on Bitcoin. David calls the two projects "conceptually quite similar" and expects the teams to drift apart as they move into different execution domains, SVM on one side and shielded CSV on the other. He added that he counts Orkun among "one of my favorite people," and passed along Talip's line that Bitcoin blocks have room for three or four ZK rollups.

Spam, Scaling and the Knots Argument

Rollups post data to Bitcoin, and the people who dislike that data have names. I asked about Knots and OCEAN, and David was dismissive of the technical case: "I mostly see this as like a guerrilla marketing tactic rather than a competent technical argument." On the supposed damage to the base layer he went further, calling it "pretty much an absolute nothing burger," and on the data in question: "this is not spam. This is scaling Bitcoin as money."

I added the old proof-of-work argument: it was designed to stop spam by making you pay for it, so anything with a fee attached is hard to call spam. David also took a swipe at the personalities, saying "you have people like Luke Dashjr, who is like extremely technically competent, but frankly, like at times, maybe a little bit like deranged." He said his own motive is plain: "I'm in this for money and not for, you know, NFTs or, you know, Dick Butts or, you know, like random tokens." On the security budget, rollups compress data "up to like a 30 to 50x compression effect," and he thinks valuable transactions stay on layer one, which he calls "a pretty significant criticism of drivechains." Near the end of the show he returned to the same fight: "the leverage in the debate has kind of switched," and "I don't need their permission. I don't care."

The Bridge and Its 1-of-N Promise

Alpen's bridge is built on BitVM and uses a 1-of-N trust assumption, which David summed up as working "as long as like one person or one party is honest." That was the point where I brought up drivechains. David prefers bridges that are "more closer to what we call cryptographically secured," as opposed to ones secured economically by miners. He talked about drivechains with respect, and I covered the same ground with David Burkett on MWEB and why drivechains need a use case.

On soft forks, the current plan needs none. In his words: "we don't need any sort of soft fork whatsoever, but it could make what we're building more efficient or cleaner." His wish list has four items: CTV, which he called elemental, CheckSigFromStack for efficiency, OP_CAT to remove the BitVM machinery and its trust assumptions, and multiplication for verifying ZK proofs. The multiplication gripe was vivid: "right now in Bitcoin script, like to try to do multiplication. "It's like, you have to like hack together, like addition, OP_ADD, OP_ADD, OP_ADD, you know, a bunch of crazy stuff." I had James O'Beirne in S16 E23 on CTV, CSFS and big blocks.

David also rejected the MEV worry for Bitcoin layer one outright: "I don't think that there's any credibility to the MEV argument." Alpen's research, he said, could yield a 100x to 1000x improvement to BitVM. A private devnet is running, a public testnet is expected in weeks or months, and mainnet is targeted for Q4.

A Grant, a Sacrificial Lamb and No Permission Slips

We spent some time on who gets to try these things. David pointed to the $1 million grant that was "put out" for this kind of work, and described the early adopters as "almost like a sacrificial lamb," since somebody has to go first and take the hit. On permission he was flat: "we don't need to ask for permission."

The backers matter to him too. Wences Casares, whom David described as "may be the single largest Bitcoin holder aside from Satoshi on the planet, who was like patient zero in Silicon Valley," made his first Bitcoin investment in over eight years by backing Alpen. David put it as "the guy that came up with the green Bitcoin logo" when the topic came up. His father texted him about some news before David had heard it himself: "I honestly didn't even know until my dad texted me."

There was a long detour around Nic Carter's Karate Combat fight and his recovery from an autoimmune condition, plus David Hoffman's broken ribs, and a story about the Alpen team taking "first place in the pickleball tournament." I also drifted into Phil Zimmermann and the 1990s PGP legal fight, when people printed the code on T-shirts as protected speech. Hoffman's ribs, I think, came out of an encounter with some gym equipment.

Dark Dollars and the Trojan Horse

David wants a lot to happen on top of a Bitcoin rollup. He described Bitcoin-backed lending protocols like Morpho as the alternative to what exists now: "you can just go to Morpho and borrow at 4.5% without any of the other hassle instantly." The current options are worse, because the collateral is "wrapped, so it's trusted, it's centralized, you're using WBTC." A rollup with real bridges would mean "no more excuses." He knows there are limits: "there is no amount of ZK cryptography, BitVM, that can ever fully eliminate the trust assumptions" between two consensus systems. I asked whether Alpen users would face a hidden catch, and he said "With Alpen? No."

On stablecoins we disagreed immediately. David said "I see stablecoins as actually kind of like a bridge to hyperbitcoinization," and I replied that the T in Tether "stands for Trojan horse." He countered that "every dollar that goes into Tether 100% supports the fiat game," and then, on my objection that stablecoins undermine Bitcoin, called it "that's a total red herring." His proposal is what he calls dark dollars: privacy-focused, Bitcoin-collateralized stablecoins.

Privacy rollups drew my skepticism. David's answer was that "this intersection of like stablecoins and privacy, it's non-existent," so there is room to build it. He noted that privacy is a "less talked about" use case and that "it's a very precarious thing to say that you support" it. Shielded CSV from Liam Eagen and Robin Linus offers scalability and privacy together, and Liam Eagen, Robin Linus and Jonas Nick covered it in S15 E58. David also corrected our vocabulary: "the technical term we should be using is like a validity proof rather than a zero knowledge proof." Alpen's proofs give validity, not privacy.

Lightning, Merchants and a Butter Robot

I asked about merchant adoption, since the show has run on the claim that Bitcoin should be spendable. David said the focus has moved, and that "I think we just oversold what Lightning can actually do." He sees Lightning as fixated on high-frequency payments, with a custodial user experience to match: "Lightning is more custodial, so is Fedimint." I answered with the butter robot meme ("what is my purpose?") and a more serious worry, that "you get Bitcoiners accustomed to using custodians."

David is comfortable with unpopular positions: "I like being the underdog." He added that "the puck is moving in a different direction," and that conference organizers keep seating him against Paul: "they always put me on panels with Paul." The stack he wants to build is "radically unopinionated," and the crowd that cares about BitVM and ZK rollups is small, since a lot of other people "don't care about BitVM and they don't care about ZK rollups."

One more idea from him: account abstraction on a rollup would let a phone's secure enclave act as a hardware wallet, because modern smartphones already have the hardware and only lack the programmability. I also joked that David was trying to get me "jumped in Vegas," after a bit about a panel that went sideways in my telling.

No Token, and a Fight I Did Not Win

Alpen has no token, and David said so before I finished asking: "No, we don't have a token." Gas is paid in Bitcoin, and the team is "bending over backwards to avoid one." His anger at the industry comes out here: "that's my biggest issue with the tokens is they've just like, they've radically set back our industry from like a development and engineering perspective." He also holds that "we can't, you know, we can't let Bitcoin development be dictated by people who are not developers and who literally have like no idea about how these protocols work."

Whether any of this reaches mainstream users is something he leaves open: "the jury is still out." He expects "insane amounts of BTC" to flow into the system once it works.

Where we still disagree is the stablecoin question. David believes Bitcoin-collateralized stablecoins on a rollup lead people toward hyperbitcoinization, and that the dark dollar idea can offer privacy as well. I still think a dollar on Bitcoin teaches people to hold dollars, and that Tether-style trust sits uneasily next to a project built on removing trust. We did not settle it, and both of us said so.