On The Brink with Castle Island - Alexei Zamyatin (Build on Bitcoin) on True Rollups on Bitcoin (EP.608)

Episode Date: April 2, 2025

Alexei Zamyatin, the cofounder of Build on Bitcoin (BOB), joins the show to talk about the state of the art in Bitcoin Rollups. In this episode:  Can Bitcoin achieve parity with Ethereum rollups? How BitVM and BitVMv2 changed the game for Bitcoin rollups When will Bitcoin rollups be considered "Stage 1" Ethereum's existential crisis Bitcoin fees in a rollup world The risk of L2s cannibalizing Bitcoin Will Bitcoin DeFi look different from Ethereum DeFi? Why Bitcoin deserves Rollups What soft forks is Alexei looking forward to? Does institutionalization ossify Bitcoin? BOB's progress and roadmap for 2025 Learn more about Build on Bitcoin Try BOB: app.gobob.xyz

Hello and welcome to On the Brink. I'm Nick Carter. Today we're sitting down with Alexei Zemiatin, who is the founder of Build on Bitcoin or Bob for short. They are hybrid all to utilizing BitVMV2. And they will be producing a roll-up on Bitcoin. They are already live on Mainnet, actually. Today we're talking about whether we are going to see these true roll-ups on Bitcoin. What are they going to look like? Are they going to look similar to what we have? on Ethereum, will the security guarantees be the same? And will Defi make its way to Bitcoin? And what does that mean for Ethereum? We dive into the deep technical aspects of this. It's a very important topic. And I think we are at an inflection point in Bitcoin Season 2 or the Bitcoin Renaissance, whatever you want to call it. And Bob is right at the forefront of that. To be clear, Castle Island is an investor in Bob. I just want to be extremely clear about that. But I think it's a great conversation. So let's dive right in. All right. I'm sitting down with Alexi, the founder of Bob Build on Bitcoin.

Alexi, welcome to the show. Hey, thanks for having me. So it's an exciting time in Bitcoin Season 2, I guess we're calling it, Bitcoin Renaissance. Yeah, tell me, what's, give us kind of the lay of the land, you know? What's been going on this year? What does 2025 mean for Bitcoin Season 2? Yeah, I mean, I think what we've, what we've seen? seen is the hype has kind of come down a bit in terms of the number of new Piccon Layer 2 is being launched.

I think we see a lot of attention shift towards application level and the Defi side of things. We're experiencing, I feel, a DeFi ecosystem renaissance across different LB1s at the same time. So you see more top tier, tier one, DeFi operators that, you know, have been very successful previously come back and look at Bitcoin. I mean, you have Louis Louie Lou coming back with Optinex. You have micro from curve coming back with yield bases. And there's more actually coming up that we're aware of. And that just indicates, you know, this general interest in the Bitcoin ecosystem and the ecosystem matured. So the infrastructure is there. I feel DFI builders and application builders feel confident that, you know, Bitcoin L2s will stick around,

that the market's going to continue growing. And I think that's just additional validation of the Bitcoin Renaissance really happening. At the same time, we see more L2s really go to mainnet. I feel that, you know, if you'd asked me a year ago, I would have said there was going to be more strong Bitcoin L2's live on Mainnet at this time in terms of at least, you know, sure, pre-bit VM, pre kind of final stage, but at least, you know, scaling and growing. We've been quite surprised by essentially, you know, the amount of teams that have actually, you know, focused on like ecosystem product development. There's not that many. I think, you know, if you kind of split

the ecosystem into kind of three groups, some, you know, playing the pure TBL game. And, you know, we've had a few of those, or like pure marketing plays. And then you have on the other side of the spectrum, the very, very technical, strong technical teams who, you know, there's a few of those as well that, you know, will build maybe more for ethos and for product and take longer to maintenance, which can work, but doesn't always work. And I was expecting to see more teams in the middle that, you know, take a product-oriented approach and build things that work with the tools that are available today while also building on the tech, but kind of have this go-to-market strategy where, you know, have things that you can do now. You start building a user base, you work with your partners,

your clients, your apps to gradually improve the security in the system, adding better bridges, adding better security models. So I was expected to see more optimism in arbitrary plays, I think, or Polygon plays. And I haven't seen as many so far. I mean, I guess Bob kind of fits into this group and we're seeing slowly a few other teams kind of come out of stealth and go to Mainnet, I think, over the next couple of weeks. But that was quite a surprise. But I have a lot of Having said all that, what I do see for 2025 is that's a year when essentially it's proven out. Like, we're going to see filtering of between teams that actually can't deliver a product

ecosystem and also tech and teams that can't. And we've seen already quite a few people, you know, projects kind of, you know, fall behind, just leave the ecosystem again, many of which, you know, are former L2s, L1s that were in other ecosystems quickly pivoted into Bitcoin and then realized it doesn't really work. Then they pivoted to AI and now they're going to do probably quantum next. So a lot of those have moved on. So you kind of can probably see, you know, there's a few top tier teams that are going to be competing over the next couple of years. And then there's, of course, also Starknett and a few of the like well-established EF players are now kind of following Bob's approach of becoming a hybrid, which I find quite exciting because it proves a point that you shouldn't pick one or the either.

you need both Bitcoin and defy in the network of techs of Ethereum to succeed. What's the state of the art for a roll-up on Bitcoin today in terms of getting to a quote-unquote true roll-up with kind of the security guarantees that we want to see? All right. I see we're going spicy from the start. So I think it's important to state that similar to what we've seen on Ethereum, where it took us five years to really see the first kind of stage one, stage two roll-ups really rollout, go to main net. Bitcoin's moving faster, but we don't yet have stage one roll-ups on Bitcoin. We have stage zero.

Can you define that? What would constitute stage one? I mean, I generally tend to follow the definitions that we see on L2 beats, where, you know, a lot of the role of definition terminology has has come from because I do deal that that's a very good way of just aligning with the definition. people are used to. So there's always some nuance between stage zero and stage one. They gradually updated and make you more strict. But generally speaking, the distinction here would be, okay, does it, is it really a roll-up? Like, is there a fault-proof system? And if yes, if it works, it's up a great, if it's operational, then you're stage one.

And then of course, they're adding some additional constraints for like, okay, you have a fault-proof system, but you still have some training wheels on with a multi-sig. And then I think in 120 days from today at the recording. actually they're going to consider you a stage zero again because you still have some dependencies so they gradually make it more strict but i think for the listeners the the way to think about this is stage zero means that you are not really yet a real roll-up you're on track you may have like fault-proof systems active built in but they're not fully operational yet they're not fully activated yet there's still some centralized components there's still some bootstrapping

going on which is you know fair enough if you're transparent about it i mean base and many other big L2 players on Ethereum are still stage zero. Yeah, I'm looking at L2B. Of the top 15, L2 is on Ethereum, only two are considered stage one, op, op main at an arbitram. And even base is still considered stage zero. Exactly. So there you have it.

It's not something that only, you know, where Bitcoin is the only kind of ecosystem that's taking a bit longer. I mean, Bitcoin 1, we've had BitVM2, which I would consider really the actual design that allows us to build optimistic roll-ups come out in August. We released the paper with Robin in August 2024. So it's not even been a year since then. So what we'll see this year is essentially test net rollouts, right?

For bridges, that's already happening. So we have a public test net for the bridge. other teams are following and like rolling out. There's kind of the obviously competition who push the test and out faster. But then there's also this push to main. What I expect to see this year is emergence of side chain designs first that are really robust with trust minimized between B.m based bridges. So kind of a one point like zero point five stage where it's a side chain,

which has some level Bitcoin security, but already has. the bridge component implemented in a way that it's trust minimized. So you don't use a third party multi-sig or custodian, but the security of the bridge really depends on the security of the roll-up, which again depends to some extent on the security of Bitcoin. And then gradually I see then probably towards the end of the year, early next year, maybe a bit later teams transition to being a full stage one roll-up. the reason why we don't yet have stage one real optimistic roll-ups on Bitcoin using BitVM is

twofold. BitVM is still new, right? So I think it's important to just, you know, draw the comparison here between Ethereum that took five years and you still, as we just said, right, only have two out of the top 10 that actually have made the step to become a stage one. Because again, these fault-proof mechanisms are complex. You want to be extra careful once you have a couple of million or billion dollars locked in there, you really want to be cautious. You might run it in parallel with like some training wheels for a while before you turn off the training wheels and then you're really, it's like the code is a law and if something was wrong, the money's gone.

And in Bitcoin, it's only been one. Since BitFame paper came out a bit more than a year, one and a half years since the Bitfam two design was released, which is really the one where we can consider it being kind of that's what's going to production. It's been less than a year. Yeah, I think a lot of us agree that BitVMV2 is the key technological unlock. What would you say the specific innovation was that BitVMV2 enabled in terms of a roll-up on Bitcoin? So it's not just about the technological innovation, interestingly.

That's when Robin first came up with that gear and published this initial write-up, hey, this is what we can do and it triggered this gold rush for people to look into a bit VM and this hype, which was great because if you wait and publish a fully period academic paper, you lose a lot of time. And then it's written in a way that perhaps the community doesn't even want to kind of consider it because it's written so convoluted with a lot of proofs. So I feel that was the right approach because it got people interested in it. And then the difference to BitVM one kind of, which was never fully like we together with Robin and then a team from research team from TUVN and common prefix and a few other research groups,

is a write-up now with full proofs for bitvm-1 as well that is actually now probably deprecated but some teams have kind of forked off this bitvm-1 design and now have improved it to a stage where it's actually kind of okay it's quite good and i think the biggest kind of if you look at the from bit vm the initial release to bit vm 1 and then actual bit vm 2 and the further improvements bit vm 0 we could call it the first write-up was a protocol that allowed you to run some computation between two parties, like you and I could, you know, you tell me you're going to solve a Sudoku puzzle and prove to me that you sold that you're going to get some Bitcoin. And then lock up the Bitcoin, you solve the puzzle.

You don't give me the result, but you kind of, you know, you have a zero knowledge proof and you commit that you prove, or even if you show me the result, but you can prove to me that you've solved it, Bitcoin can't verify a Sudoku puzzle, right? But with BitVM, we could basically show that as example, like, oh, we can verify some simple company. or more complex computations actually. But it was a two-party protocol and it used something we call bisection. It's quite close to what Arbitum actually invented on Ethereum back in the days. The whole crux of BitVM is that we can't run the full program on Bitcoin. So we kind of need to slice it up, dice it up into smaller parts and only run a small part of it.

Now, when we do optimistic verification, we only want to run that piece of of the program that proves that you cheated, right? So let's say you have a program in, if you're smart about attacking the system, you're not just going to run something arbitrary, right? Because then you're instantly going to lose because you can't even prove the first kind of challenge step. You're going to change the program in a very specific way, maybe even change only one bit. But we'll still find it, but the challenge is, and that's why we have this challenge

response protocol or what we call bisection and what arbitrary and optimism also use on their sides, we are trying to find publicly in a verifiable manner that every Bitcoin node sees where exactly you modified which step of the program you made the change that now makes it invalid, incorrect. That allows you perhaps to steal the Bitcoin. And in BitFam1, we'd play this back and forth game where I would challenge you and you'd respond. And based on your response, I would know, oh, I have to search in the first or second half of the program. And then I would continue kind of zeroing in basically running a binary search. So we're searching, you know, left and right back and forth, like zeroing in on the exact

place of the program where you then actually made the change. And that would take over like 70 to 150 transactions on Bitcoin back and forth over a long period because ultimately between each two challenges, we need to give us some time because your Bitcoin fees might go out. There might be delays. So what, it's like a couple of months, which was of course not feasible. I wasn't feel it, but it proved that it's possible, right? The moment you prove that it's possible, then we can look at optimization.

And then BITVM1 cut it down already quite significant. Then that was the actual full design for how you can, you know, spec it out. There was a prototype. It was open now to a group of operators. So no longer only two people, but now you could have a group of operators. It could be 10, 20, 30 people operators. And the idea was as long as one of the operators is honest. So one of the 30, one of the 40, your system will be secure because the operators could challenge

each other. So now you have a system where you find a bunch of people. Like think of a big multi-sick, right? And that's, I think, a very good comparison to make. In a big multi-sick scheme, if you find 30 people, 15 plus 1, so 16 need to be honest. Otherwise, they can steal. So if 16 are malicious, so 50% plus 1, your money is gone. In Bit 3M-1, you only need 1 out of the 30.

because if you have one operator, that's honest, this operator will just challenge all the dishonest ones until nobody else is left. And then they're honest, so that's why the system kind of is secure. Then there was quite a bit of progress on cutting down the number of transactions and you had a few teams like Fairgate and a few others like spin out.

Use that design and start optimizing it further. And by now it's actually quite good. Like they've made a lot of progress. And then again, kind of got closer back to like the I guess the canonical BitVM research that then came up with BitVM2. And then BitVM2 really took this where you had again. You had still this bisection protocol multiple steps, let's say 40 transactions per round per person. So like 30, let's say 40 to 60 probably in total, back and forth, still too slow, still kind of too many steps.

And only the operators can challenge. So BitVM1 had already one out of N operators, but it was permission. in terms of challenging and still too many steps. And then BitVM2 changed it quite significantly and that would say instead of trying to run by section, we just have three steps. And what we do is we basically have some bigger on-chain transactions, but it's only three instead of many, many small ones.

So the amount of communication and the chances of, you know, the challenge are just going offline throughout the period and missing a step and then the whole system fails, you just get rid of that risk and now the system is as follows we take a big program which we first put into a zK kind of pre-processing step so you get a snark prove or verify out of it which is compressed much smaller standard size much easier to process and then we split up that program into many pieces and what big fm then two does is it says okay like the operator will commit to these smaller chunks of

of the Zika verifier. And if you misbehaves, if I'm challenging you in the next, like in a follow transaction, so there's a challenge transaction and then there's a response. And in that response, you give me all the intermediary steps. And with that information, I can in the next step, prove that you're wrong. So we only need three steps. These are some bigger transactions. And sure you can optimize again, you can like find a middle of ground between between

one and two where you kind of don't do 70, but you do like five steps. but it's then cheaper for the individual kind of step for the individual user. So obviously a lot of like optimization space there. But the idea was we instead of using some like really execution level by section, we just take a snark and we look at the snark. So we can then just standardize whatever program we're plugging in, which helps a lot already because anything that you can verify on as Zika snark proof, you can plug into BitVM, which includes entire chains.

And we've reduced the steps to three, right? Maximum four in some edge cases, but that's already like, like it just takes way less time and it's much, it's much easier to actually implement. And the big kind of step here was permissionless challenging. So anyone could challenge, right? The whole goal of BitVM2 was to make it in a way that it's very close to what you see on Ethereum where any user could.

challenge, even you yourself. And that really gives us this proper optimistic roll-up security or optimistic roll-up bridge security where, you know, as long as there's one honest user somewhere, your funds are safe. So that's the latest state of Bitcoin. There's still obviously, like challenges and open research to kind of improve it further. But essentially we're at a state where from theoretical perspective, from research perspective, we are on power very, very, very close to Ethereum optimistic roll-ups on Bitcoin. without a fork. And I think that is very impressive in this short amount of time and with a fraction of the funding that etherlops have actually raised.

So what does this look like from our practical perspective? What would one of these kind of hybrid ZK optimistic rollups actually look like on Bitcoin in terms of, I know you touched on it, but trust guarantees and exit window lengths, things like that? So the interesting part is that design space is still quite open. And there's like different approaches. So it's quite hard to say like, okay, here is like the definitive exit window. And also like because on Ethereum, right, there was a seven day kind of challenge period and then people discussed and realized that actually we could do it in three days because

there is no reason for it to be seven. It's just like a conservative value. So it's same for Bitcoin, right? If we say, okay, well, we just follow Ethereum's lead here and we also say, three days around, then BITVM just kind of has two steps here. So it's like instead of it's a bit longer. So it's, you can estimate like 2x of that of, of the Ethereum kind of challenge period because there's just a bit just the way how BITVM is set up, you need like, it's a bit longer.

So you'd say what like seven to 14 days for like full finality, but without optimizations, right? You could cut this down. You could, you know, go lower after the system has been like for a while and your confidence that, okay, you know, this works. There's challenges on the system. There's like challenge response system works. People are online. There's enough people to actually make a challenge.

You run through it. There's no bugs in the client software. Because obviously what happens, if you start running the challenge and then you discover a bug, you need to fix the bug, come online again and to continue the system. So I would say it starts probably at two weeks plus

and then it will go down over time. And that's fine because you ultimately, well, there's two considerations here. use it for bridges you're probably fine to be slow because the way the bitfam bridge is design is it's fast for users it's slow for operators and operators are going to be professional providers of like liquidity and whatnot so they're not going to you know they can offset the time like they're used to this also on ethereum for retail users it's going to be fast like retail users just swap in and swap out they don't really care um i think a comment here

bitvm comes with a built-in liquidity bridge whereas an ethereum being have these slower you know if you use a native bridge to arbitraram or optimism it takes to seven days to go back and you can use third-party liquidity bridges to go faster bit VM comes in with a built-in one by design so for users it's generally quite okay it's the operators who have to deal with rebalancing you know between bitcoin and the l2 but generally we've seen that there's product market fit for searches and these kind of operators on ethereum already um i guess the biggest concern it really isn't time or like trust assumptions we can get right. It's the cost. And there was a very good report by Galaxy that studied the Bitcoin

fee market under the assumption that you have a couple of these full-fledged optimistic role of L2s posting data to Bitcoin because here is the challenge, right? If you want to be able to do a challenge response game as a roll-up on Bitcoin, well, you've got to get the data from somewhere to be able to verify if something's wrong and to then trigger the challenge. On Ethereum, think of it pre-hard fork. It was very expensive to post the data. That's where Celestian, a lot of these alt-dealiers emerged. Now, Ethereum Forklin is actually very, very cheap. Well, Bitcoin is very expensive, right? It's the most expensive database in the world. So if you start throwing four megabyte block chunks of data in the regularly, well,

that's going to be expensive. If you're the only one, and if there's multiple teams doing that, you just got to drive the P market up. And I don't remember the exact numbers, but I think we ran our internal estimates and it's 100 to 1,000 times X more expensive than what you see in Ethel 2 is right now, which is kind of a problem because we're not competing in an isolated market.

I don't think people and specifically retail users care that much whether the thing is secure by Bitcoin or Ethereum. They might prefer Bitcoin. It's easier to sell it to them. I'm explained, but if it's 100 times more expensive, well, I'm not sure they're going to like that. And so this really works once you have either we've improved the DA mechanism of Bitcoin. And maybe like, you know, there are ways to kind of do this with like some tradeoffs and like, but you always introduce trust assumptions that we tried to optimize there.

They might be okay because they use like against Bitcoin staking or merge mining. So things that still give you Bitcoin security and you might be okay with it. But then it's like, you know, it's already like maybe 0.7, stage 0.75 or 0.8, it's not stage 1. But it's okay. Or, well, you have to grow and be big enough and have a lot of users. And then you can basically split the costs among them. And sure, if the cost per user is a cent or two cents more, people are not going to care. So that's the kind of two options that we have to really. And that's the biggest challenge. Like now, apart from you, getting everything running and so on, but that's pure engineering work. The bigger kind of theoretical hurdle is the cost of posting data to the Bitcoin L1.

And that's why with Bob, we've taken this multi-phase rollout approach, where we start with a design that works and gets us to stage 0.5, arguably. And then we can go to stage 0.8, where we kind of are very close, but we don't post all the data fully to Bitcoin. And eventually you can say, okay, you know what? We're big enough. Our users, the gaps really want. that it is affordable, okay, let's start posting the 4 mega, like let's start posting data

to Bitcoin. And by that time, we actually might already have better ways to compress that data essentially. So Bob Maynett is live. Obviously, you know, things are going to change a little bit in the coming months. But tell me about the traction you have so far and also what kind of applications people are favoring on Bob right now. So Bob Maynard has been like. lives for, I mean, not quite a year, but almost. So since May last year, we've seen 430,000 unique users and use applications over time, which is quite great, especially like given, you know, we took the approach of going to main it very fast, in a very short test phase, we didn't do these big predeposit campaigns.

We felt, you know, we want to go to mainnet and let apps build and then figure out what actually works. So we really started seeing growth from, I'd say, November last year. And that was when we started working with Bitcoin LSTs and this kind of Bitcoin staking ecosystem and allowing people to stake their Bitcoin and Bitcoin staking protocols like Babylon, but then still use it in the form of LSTs and defy, something that was previously only available to Ethereum and now still in its infancy, but already available for Bitcoin. And that's also when our TVL numbers started to grow. We're around 200 and let me see what the latest is in L2Bit,

280 million in on-chain deposits, which puts us in position 1, 2, 3, 4, 5, 6, 7 on L2Bit if you don't count like native tokens. So just pure like real like stable E for Bitcoin TVL. And that's just like we're just getting started. So we're seeing like this thing start picking up steam. We have unyswap live on Bob. It's the fifth largest uniswap deployment ahead of B&B chain, ahead of optimism. We have Euler that just went live and they've been growing really, really fast and doing a really good job.

Bob is also, this is all public in the forums, is in like very progressed stages of conversations and like governance discussions with the Avedow for a Bitcoin focused deployment on Bob. Of course, still subject to final governance vote, but we've had a lot. very constructive conversations with the app and doubt, you know, how we'd like to roll this out and make this also like a place for them to get more Bitcoin exposure. And we see also a lot of new startups that are looking at Bitcoin Defi also really oriented at different geographical locations where they really focus on retail users borrowing

starting stables against Bitcoin, creating Bitcoin-backed stable coins, but also projects on the intersections between Bitcoin and RWA's like really looking for ways to generate long-term sustainable yield from Bitcoin, but also being open to being like C-D-5, which you know, if you know the risks that can work quite well. So we see essentially a lot of inbound interests from new builders' apps to start building around Bitcoin. And just because how the timelines are, and you start with an idea and your startups come on board. We're going to see a lot of them start going to main at like Q3, Q4, 2025 and really just grow further and bring more and more users and unique applications.

So with that, like we've been very happy with the growth of Bob. And we really see the biggest opportunity I think here is this defy still. Bitcoin Defi is still in its infancy and you know, and still it is huge. right? It's billions of dollars already that have been bridged over to Bitcoin L2s and Heath and others L1s using still, yes, centralized and custodial bridges, but it's still only a fraction of what we could access. And if you compare the numbers and the metrics, if you look at Ethereum, for example, the market capital TVL ratio, it's at around 30%. Bitcoin's at what? 0.3 maybe

getting closer to a percent, but it's still like so far away, so we have like a 30 X to 100 X growth opportunity here if we just reach the levels of Ethereum. But this is a multi-tenth, hundreds of millions billion dollar market that we're looking at. And the interesting observation here,

Nick we discussed this like last time on I think on Twitter as well, we see that with the US embracing Bitcoin and many countries following as well, that a lot of treasuries are building up on Bitcoin. They're like buying BDC, but as a treasury manager, you have to earn yield. Your mandate is to earn yield.

And sure, Bitcoin might appreciate in price, and you could factor that in and by all means, I'm not a treasury manager. But the feedback we're getting is from big custodians, big funds, big LPs, and there's a new liquid fund of funds being launched, specializing in Bitcoin. There's a lot, a lot, like hundreds of millions,

of dollars worth of demand for Bitcoin yield. And they're looking for ways to actually get that. And that is where Bitcoin L2 is like Bob are just positioned in the best way possible to accept that. But at the same time, of course, paired this with retail adoption, where Bitcoin is far ahead of any other assets, at least in the digital space. Do you think Bitcoin Defi looks in any way different from what we've seen with ETH, defy or other chains? Or is it just, we're just going to rebuild everything that's built on Ethereum, but just, on a stronger collateral base on Bitcoin. Well, we already have, right?

It's already all there. Like that's the cool thing. And that's why we firmly believe in hybrid design where you don't have to rebuild everything from scratch using Bitcoin script or like some derivation thereof or like come up with any programming language. No, like we accept that the EVM, even though, you know, technically speaking, they might be better virtual machine designs. It is the best because it is the most adopted and that's just what makes it with.

and by supporting that, but still providing tools and access to Bitcoin and access to Bitcoin security and liquidity, well, you get all the Tier 1 Difa protocols and things that exist in Ethereum easily deployed and build out with Bitcoin Focus. So that already exists. So everything you can do on Ethereum, almost everything you already can do with Bitcoin and Bitcoin L2s. What you see now is specialized Bitcoin Focus defy emerging, as mentioned, the Saur of the call that is very interesting because these are things that specialize in bitcoin and the asset

and the properties of bitcoin as the asset and are targeted at the different user groups that have btc of which there are many right many more i guess so it's much more diverse than the ethereum user base where you have defy degen's and defy funds and power users now bitcoin has retail users they have tradfine that you know isn't a defy acquainted yet You have a lot of like people who are in centralized exchanges and you know they know Bitcoin so that's the one thing they can you know associated with usually. So we see a lot of essentially I feel like yes you see a lot of the same but I do feel that there's a much stronger push towards retail and user basis that previously were not attracted

to defund. I think that's the biggest opportunity it's really growing the pie because ultimately every defy user is a Bitcoin defy user, like almost. The majority of eth users have BDC. There is a lot of Bitcoin in DFA on Ethereum. Yes, it's centralized forms in WAPN, and sure, it's hard to access and whatever. And by, you know, improving security and accessibility, you can convert a lot of these users and let them use more of their Bitcoin. But at the same time, you have users that have never heard of Ethereum before or never

touched Ethereum, not because they're anti-eth. It's just because they're just not experts. You know, Bitcoin, that's what they hold. And maybe they hold a few altcoins is because they got airdrop through like finance or some other exchange. But it's really Bitcoin that they understand and that they can associate with. Because that's the first thing that they saw when we got into crypto. And they have like a backstory of when they bought their first Bitcoin. So it's much easier to onboard these users where you don't have to explain to them a very complex proof of stake mechanism or how your L1 is like super fast.

It makes these specific tradeouts. Well, it's just, okay, it's secure by Bitcoin. you can use a Bitcoin here in the best possible way. That's very easy. So in the ETH community, there's a lot of bell-haking around the roll-up-centric roadmap. People are pretty down in the dumps about it, to be honest. I don't know if that's just a function of ETH BTC, ETH-NOT really participating as much in this cycle, or if it is like a real fundamental concern about maybe ETHL-2 is cannibalizing fees from DL1

or ethel 2s maybe not playing that nicely with the L1 and launching their own tokens and kind of siphoning away some of the demand for ETH itself as an asset. What do you make of this? Is this something that concerns you as we start to bring roll-ups on a Bitcoin? Do you think we're going to have the same fate? Or, I mean, is this even a problem? I mean, that's a very interesting question. I think it's a problem in Ethereum because Ethereum right now is kind of in the middle.

We have Solana, which is specializing being fast, cheap, and retail friendly, and it's okay to make tradeoffs in terms of decentralization and so on. But it's secure enough, right? For many people, many applications. And Bitcoin is Bitcoin. It's a currency. It's an asset. Nothing challenges Bitcoin as, you know, digital currency, digital, whatever you want, as money. Nothing comes close.

And the ETH kind of landed in the middle where it's like neither that it failed. to kind of win or it's failed to challenge Bitcoin on the money side. But now it's losing to Solana with its roll-up-centric roadmap where, okay, you have a lot of fragmentation and Solana is this one big monolithic chain that's fast and cheap. And that's and it just, you know, focuses all its resources at execution and has less fragmentation. I think on Bitcoin, we will not have that big of a problem because, you know, it's a problem. because well, sure, if L2s are taking away use cases, application, transaction volumes of

Ethereum, well, you don't, none of this exists on Bitcoin. You don't have defy on Bitcoin. You don't have a lot of transaction volume on Bitcoin. You had Ordinals, so people consuming block space. Okay. But they're not going to go to the L2s with the block space. They want to have it on Bitcoin at 1, right? At best, actually, what's rather going to happen is that they can, you know, Ordinals users can

write and post the data to Bitcoin, but then use like have a lot of, you know, have like hybrid designs, which they can actually use on the L2s, which gives them more things to do than just posted and then be able to transact it every 10 minutes. Like Bitcoin is like Ethereum was such an upgrade. And I mean, I'm going to get a lot of hate for saying this, but like in terms of defy and usability, Ethereum is just way ahead. Defy tooling, usability applications. Ethereum is like way, way, way ahead of Bitcoin.

But Bitcoin L2s are in part. In fact, they're better because they're cheaper, they're faster. they may be not are not yet as secure but once you reach stage one well arguably you know you're pretty well off that's enough you're secure enough right your bitcoin security with proper proof you're good even with like 0.5 right with like side chain designs or bitcoin staking you're good right this is better than anything else on the market especially if you look at the ethel two kind of ecosystem with a lot of stage zero is out there and i think the the biggest difference between like the Bitcoin L2 centric roadmap and the eth L2 cent the ETH roll up centric

roadmap and the Bitcoin roll up centric roadmap is there is consensus in Bitcoin that that's the only way there is no other option like we have no functionality on the L1 it we don't want that it's not going to happen we're fighting and bike shedding over minor upcode upgrades for years the only way that Bitcoin ever is going to get real defy is through rollups and working around bit VM and ultimately hopefully getting some level of additional upcodes that make it more secure. Whereas for Ethereum, it's pure scaling.

For Bitcoin, it's actually use cases. And I think that is a subtle, but very important difference. Bitcoin-L2 is not taking anything away from Bitcoin-L-1. There's nothing there to be taken away, right? You're bringing a use case. You have a chance of actually bringing a lot of users, consuming more block space, and seeing more applications on the intersection of the L1 and the L2 like swaps and like, you know, we're lending where like more logic happens on the

Bitcoin L1 with things like BitVM, but still. still a lot of the defy stuff happens on the L2. So I think there's a lot of synergy there to be explored. And you see like DLCs and more and more like RFQ based and intent based swap systems like Bob build and there's Optimix and there's others now coming out that are specializing and also like leveraging what the L1 has to offer. So it's kind of in the middle and it drives more transactions and more blocks based amount to the L1.

I do think long term, we as a Bitcoin focused L2 project, even though we're hybrid, we have to think about how do we contribute to the long-term sustainability of the Bitcoin demarket. That is certainly a concern that's been on our mind ever since we started researching Bitcoin 10 years ago. And that is something that we'll have to figure out because, of course, there's a balance between paying a lot of money for posting the data and then suffocating and losing to Solana because you're more expensive. So, you know, we have to find a balance where, yes, there's contribution, but it doesn't, you know, it's not extracted early on.

Like, I don't think people should expect Bitcoinotis are coming here to immediately dump a lot of data to Bitcoin to make the Bitcoin fee market. You have some miners make enough money to keep mining. Bitcoinalties are still startups. So you don't extract value out of a startup. You have to help them grow and work with things that, you know, are somewhere in between that kind of help both the consensus, participants in miners and the L2s and then gradually win the L2s have won the market against Eith L2s and so on and everybody else. Sure, then there'll be enough fees coming in.

Yeah, I want to talk about fees in a minute. It's a fascinating question I get a lot is what does the fee environment look like on Bitcoin post true roll-ups? But one observation first, I guess, is, you know, another distinction is Bitcoin almost never increases the quantity of block space available. Well, they did one time, really. with Segwit, arguably. You know, so they quadrupled the block space.

I think that was it in the whole history of Bitcoin. Whereas Ethereum is constantly in this tug of war between trying to make Eath this sort of monetary asset and then trying to fight with Solana and others and increasing the quantity of gas or effectively block space, which they do periodically. I mean, I think it happens all the time, right? So with Bitcoin, there's this like social commitment to never increasing, the block space as part of this competition with other newer blockchains, whereas with ETH they're sort of like constantly pushing down the fees that validators get by increasing

block space. So I think that's one reason also why Ethereum's like struggling. I mean, I think that's a very interesting take and it definitely contributes. I think there's more, I mean, as you mentioned before, it's not just about the fees coming down why Ethereum is struggling right now, but because it's in limbo between like different kind of markets. But of course, I mean, the fee revenues are getting, you know, sucked away by the L2s and then increase block space for when actually there is a lot of demand and like it becomes unusable again. So it's a tough it's it's a tough spot to be in no fairness, right? Yeah.

I think Eiff is in this, maybe, you know, Blackberry, Blackberry is a bit exaggerated. But it could become that if it if if if actually doesn't innovate and decide what is it going to go for like what's the focus. It's it's not even like the technology is bad or the governance is bad, though I certainly have issues with the youth governance. It's more like they have an existential crisis and they don't know what they want to be. And with Bitcoin, there was never even a choice. We were all we had our existential crisis in 2016, 17. That was the box size wars and the layer to. approach one, you know, the small blockers one. And that's just been the way. There's no other way

than doing the layered model, whether it's lightning or now roll-ups. So in terms of fees, and it's basically impossible to project fees in a dynamic system like Bitcoin. It's like so many variables. But is it fair to say that you think fees on Bitcoin will be structurally higher once we have a kind of a mature system of roll-ups in place? And also like, How much is the burden on main nut in terms of the data requirements for these roll-ups? I've actually just pulled up next in the other window, the Galaxy report, which I can highly recommend reading. So that's really well thought through reports. I mean, there's, of course, a lot of like, you know, debates about, you know, whether all the conclusions are indeed valid.

but I think we can all align, right? The math, even with that looking at the detailed numbers, right? If we start, like even if you don't post a lot of data to Bitcoin, Bitcoin L2 is existing. And there being, if there is demand for Bitcoin L2s, the number of transactions are going to increase because you, you bring in more users and these users have to bridge and withdraw that Bitcoin eventually, right? So just the deposit of the journal transactions are moving between different Bitcoin L2s, that alone is going to increase demand for block space.

might not be enough to sustain the Bitcoin fee market, but it will contribute to continuous activity. That is not as reliant on spikes where somebody launches in your NFT collection and then the if they flood the network. And then if you see the chains start to post data in some form, of course, that immediately means, well, think of it. You're just like from a very simple perspective, right, when does Bitcoin network clog or when has it been clogged in the last

last year when when you had ordinals of ruins of UC20, like inscribers, you know, go into frenzy and you know fill the blocks. Well, you have this. Yeah, now basically, this just spread out throughout the year. You have just continuous kind of bigger chunks of data being posted, which takes up space on the nodes, but that is nothing compared to, you know, what the actual, you know, costs that users will have to pay or like these rolls will have to pay. So I think, look, the math can actually math in the sense if you bring enough users, can afford to post data. There will be of course an inflection point.

And I think it's going to become a barrier of entry for like real Bitcoin roll-ups. Like if you want to become a full Bitcoin roll-up, you have to become, if to grow to a certain size. So I do also see a future where there's only a few Bitcoin L2s, potentially with a lot of L3s on top that, you know, with teams that don't want to deal with BitVM or frappers. It don't necessarily have the technological capabilities. It will be commoditized, but it is definitely more complex than Ethereum L2 designs. And you have to actually pay the cost, right? That is a big difference where, you know, and you saw L3s emerge when the L2 costs on Ethereum were so high.

And Bitcoin, they aren't going to come down unless you have an upgrade or we come up with a massive innovation on the data availability side. Or again, we go with this kind of semi-secure, like semi-roll-up route. where it's like roll up, but it's like an optimist, basically optimiums, but the data is where the data is posted to another network that isn't Bitcoin, but that network could be secured by Bitcoin's taking or merge mining, which arguably is pretty reliable. And we can actually, you know, verify that that network is executing correctly.

The data is there. So you can make it rather secure and kind of acceptable. It's hard to predict how much data will actually be posted to Bitcoin, but it's certainly going to be more. And there is just unless Bitcoin roll-ups completely fail, completely, categorically fail, but even then you'll have side chains. And if Bitcoin side-chains emerge and they are emerging, there is more demand. There is just you see it.

There is a lot of focus on Bitcoin. Even with just side-chains, you're going to increase the number of transactions on Bitcoin, moving in, out and between them. And with roll-ups, you just add additional transactions on top. You bring more and more users. I mean, you have millions of users. that currently can't use Bitcoin properly because it's too expensive, too slow. So I think no matter which way this goes, unless it goes completely to zero,

and side chains are all dead and nothing short. Then it won't really add much. But in all the other scenarios, it's going to contribute to the Bitcoin market. Two questions. You've been generous with your time. This has been great. I'm going to put you on the spot here. Any soft forks that you think would make life easier for Bob?

but you can see me because there's no there's a video here but you can see my expression so I see the following from Bob's perspective we've taken the roots of pragmatism and least resistance and like product oriented thinking will work with whatever we get if there's a software that gets activated it will make our lives easier certainly no matter what we get it not be necessarily the one thing that we like for BitVM. I personally think that lightning already got two or three upgrades and still isn't really working and it's like, you know, actually losing an adoption right now again. It's going down your wallets and like exchanges actually going back and turning it off. It still has a use case. And I think it deserves to exist,

but it's not the ultimate solution for Bitcoin's scalability and use case problem. So I think the next software should be driven at roll-ups and L2s. Like, that's what we should be focusing on. The problem is we're still a bit early to determine which of these potential upcodes helps us the most. We debate about OPCAT, TX, H, CTV, and so on. There are potential combination with CTV and like other upcodes that could be useful. CTV itself doesn't help us as much, unfortunately.

Like, it's great. It's better than nothing. And the one concern that I kind of initially was, you know, I made fun of it. And then basically I guess like Eric Wall made a very convincing argument of like more institutional adoption means more justification. We don't have that much time to, you know, activate more of course. So there's, I think, two scenarios here. Either we take a long time and it becomes very controversial and maybe we get one in the next three years, four years. And that's it.

And then like you see as you yourself said in it, right, the times between upgrades just will increase and increase and increase. The other probably less likely option is that we get one and then a few others quickly after where there's this window of innovation where institutions are open because they're still onboarding and they kind of see the demand of like and the opportunity to, you know, actually also make money of Bitcoin L2s in the ecosystem. And they see the benefit of adding a few simple upcodes that. you know really allow bitcoin to compete with ethereum and all everything else i'd give this like

maybe a percent chance to be fair but it is possible yeah to me it looks like a geometric series where the time doubles in between every uh a soft fork so we can expect one um in 2033 maybe hopefully not um i think it's going to be sooner but by 2030 fingers crossed Last question for you, Alexi. What can Bob users look forward to sort of through the rest of 2025? What's on the roadmap for you? So we have kind of two focus points on the roadmap product and tech. Tech-wise, Bob is becoming a Zika roll-up.

So we're going to be, because Bob is a hybrid, right? So we are connected to both Bitcoin and Ethereum. On ETH, we're going to be stage one ZK roll-up. So that already improves security a lot. it's a very important component to the BitVM bridge. And then we're going to be rolling out Bitcoin staking security. So really transitioning to this full hybrid design, we're an ether roll up with Bitcoin security and the trust minimized BitVM bridge.

And that is basically the end state for the end this year. That's where we're heading. We're on time with the roadmap. We're very optimistic of rolling this out essentially through like Bitcoin security in the summer. And then the BitVM bridge will follow shortly after. and then obviously training wheels and gradual roll-up, but we're really heading there at full speed.

And then that will make Bob the most secure chain, ultimately in the entire market, apart from Bitcoin and Ethereum, which I think is a very solid argument and just allows us to be very convincing within users, retailers and institutions that, you know, this is secure, this is going to work. On the product side, we have, of course,

more Bitcoin-Def applications, stable-quant integrations, on off ramps, smart accounts, which really allow you to send and receive Bitcoin, paying fees for cents for fees within seconds or even faster. So because Bob benefits from a lot of the innovation that happens on Ethereum,

we'll see our block times likely go down to sub second block times, which just makes us feel instant. So we can do things that Lightning wanted to do. And it's perhaps not as perfectly secure just yet, but it's gonna be more secure than anything else in the market and good enough, for micropayments if people really want to do that but also with stable coins and boring stables

against bitcoin another big thing will be of course thou governance and participation in the network and the ability for users to really then vote on which upgrades going next and i think that's very important for us it has always been because building just for ethos and having the best tech doesn't make sense in our opinion our goal is really to onboard a lot of builders and users and work with them and their needs to make sure we're building the thing that the market wants. And that means giving our builders and users a boat and a voice in the system. And I feel that's a necessary and very important step in the life cycle of a network. And we're essentially moving in that direction and are very excited to essentially see this

all unfold over the next couple of months. Well, very exciting stuff, Alexi. You've given me some amazing explanations of roll-ups, which is honestly not an easy task you know as I said in the introduction we're investors were supporters of yours we're very excited to see where you take it over the next year amazing well thank you so much for having me and yeah super excited for the rest of the year big corn nuisance is really happy