S15 E46: Peter Todd on Bitcoin Covenants, Tail Emission, OpenTimestamps & Why Lightning Rocks
Search this conversation
0:40Vlad CosteaHey, you. Thank you for listening to the Bitcoin Takeover podcast. This episode is sponsored by SideShift.ai, StarkWare, LayerTwo Labs and Hodling.ch. Don't forget to subscribe. Hello there and welcome to the Bitcoin Takeover podcast. This is season 15, episode 40 something by the time I publish it. And I'm here in Mallorca with Peter Todd, longtime Bitcoin contributor, former Bitcoin Core developer. I'm not sure if he is still involved in writing code, but he's definitely reviewing code and sometimes he's very critical of what's happening.
1:18And I want to ask him because this is his fourth or fifth time on the show about all sorts of stuff that happened in Bitcoin lately, from Ordinals to BitVM, to stuff that might come to Bitcoin like covenants, and stuff that can be built on top of Bitcoin with covenants like ZK roll-ups and stuff that you expressed dislike towards like StarkWare and also drivechains. I want to hear you talk about that. So there's a lot to unpack. Hi, Peter. What are you up to? What do you do these days?
1:52Peter ToddWell, go hang around Bitcoin conferences and speak on podcasts, apparently.
1:59Vlad CosteaYou're on a full campaign to push for full RBF? And I guess you can talk a bit about that one.
2:07Peter ToddSo, I mean, full RBF is kind of a funny thing because... Right now, essentially all miners are running it, as far as I can tell. The only one I know of that says they aren't... For at least some of their hash power is Ocean, and that's just because Ocean gives you the choice of either running Bitcoin Knots or, you know, mining against a Bitcoin Knots template or a Bitcoin Core template. And Bitcoin Knots, of course, says full RBF on by default bizarrely... You know even though this fixes a bunch of issues with transaction pinning and whatnot, Bitcoin Core actually just closed my full RBF pull request.
2:44And... You know it's it's just one of those things like dude you you should just merge this months ago I mean why are you you know why are you creating yet more delays in this stuff. But. In fact, also, I can probably even say, given that this will probably publish after tomorrow, that I did recently find... You know so-called free relay attack involving full RBF you know I'm probably not the first person to find this because it's kind of obvious, but... If all miners are mining with full RBF and Bitcoin Core nodes are not.
3:19Obviously, that means you can broadcast a series of transactions that use full RBF to get a small transaction eventually mined, or at least in minor mempools. Whereas... Then the Bitcoin Core nodes that reject that by default you can then have them, you can then broadcast a transaction that only they're going to go broadcast. And since the probability of a non, you know, of a, since so few miners are non-full RBF your probability of the transactions you broadcast getting mined is so low that you can effectively broadcast data for very little cost.
3:58Um... Now I got to go point out, I mean, this is only one of like three free relay attacks I found the past couple months. And frankly, I don't think this class of problems is fixable. Like we just have to accept people will always be able to go and broadcast data on the peer-to-peer layer at relatively little cost compared to what we want them to. But it's interesting that even with this kind of attack, you know, Bitcoin Core doesn't seem very interested in enabling full RBF.
4:24Vlad CosteaYeah, to me it's very interesting because two of your friends started the mining pool that you mentioned previously, Ocean, and If I was to judge this politically, I would say that you are going to align with them and declare that the ordinal inscriptions are an attack on Bitcoin and they're a bug that needs to get fixed. But you took the opposite stance, maybe because of your previous experience with OpenTimestamps and maybe that you are thinking that they could also filter that out But you're also more free market than I honestly expected in this regard.
4:56And you're happy that the miners are making more revenue.
5:00Peter ToddWell... So, you know, full RBF is one of those things where it's kind of the free market approach. It's like, obviously, miners are going to want to have more money. If a replacement exists that pays more money, why wouldn't they... Take it. And I've been proven correct and that essentially 100% of miners are mining full RBF. Now when it comes to blocking inscriptions, You know, I think there's the same kind of rational, like there's a lot of money that Frankly, idiots are willing to go put up to go and get all these silly transactions mined.
5:28Of course miners are going to mine them. And also, The nature of the Bitcoin protocol is such that There isn't really a technical way to stop people doing this kind of thing. I mean, it's hard enough to stop people putting data in the Bitcoin chain, let alone prevent people from trading assets with transactions that inherently look just like normal Bitcoin transactions. You know, there isn't a long-term way that this can be blocked, so we should accept that and not waste a lot of development effort trying to stop all this stuff.
5:59Vlad CosteaAnd you're also a proponent of tail emission. And you believe that... I mean, it's a very... Interesting assessment because right now Bitcoin miners are being paid from the inflation that's being generated within the system. We have no empirical evidence that When this... Inflation is going to get diminished every four years. There's going to be a sustainable way for the miners to keep going. Of course, there are these... Guys who say, "Yeah, we should just let the miners go bankrupt or something and stop their operations." And see what happens after.
6:34But that's very reckless. And you have the most extreme of proposals and you say let's add this tail emission. And I believe that...
6:43Peter ToddThere are much more extreme proposals than what I've said in academia. So, you know, I think the Bitcoin world is just not seeing what people understand this stuff are actually researching. But, you know, the bigger point I'd make, though, is that... There's this idea, I think, in the Bitcoin world, you know, markets just magically fix everything. And that's not nonsense. Markets go and do what markets do. Based on incentives. And the problem that we have for Bitcoin is that there aren't clear incentives for decentralized entities keep Bitcoin secure.
7:16And the simple nature is that the way you know The way mining works is that it becomes unstable when all of your revenue comes from fees because you can play games like reorganizing the blockchain to go get fees that other miners didn't. And the problem is, The bigger you are as a miner, The more profitable those games are. And we do not want that kind of incentive. And tail emission is just one of those things where, had Bitcoin been designed with that from the beginning, We wouldn't... You know, maybe to a degree we might have some of these issues, but a lot less.
7:48Whereas right now we have this very reckless design of Bitcoin where you start off with Bitcoin working in one way. With subsidy always providing a guarantee that there's incentives to create a new block. And the assumption is, well, we'll just Go to the extreme opposite and we'll just hope it works. And no other cryptocurrency has ever actually survived this transition. There's no reason to think, you know, Bitcoin is going to be magically different there. And ironically, a lot of the people who are who hate the idea of tail emission the most, they seem to also be the ones who hate the inscriptions, hate all these fee-paying transactions.
8:25What exactly is these people's plan? I don't get it.
8:29Vlad CosteaI do agree that these people seem to hate the idea of miners making more money. For whatever reason, and they still regard them as evil, I guess this comes from a hatred towards Bitmain that goes back seven or eight years.
8:44Peter ToddI don't even think it's about hatred towards Bitmain or anything like that. I think it's just They've bought in so hard to the 21 million meme that they just cannot accept that Satoshi might have screwed that one up You know, as it is, like the idea of having halvings is just stupid. You know, having the shocks to the system, we're lucky we've... survived these shocks to the system But the idea of putting it in the codebase was stupid. It was a terrible idea protocol-wise. But, you know, Satoshi made that mistake.
9:13Satoshi also made the mistake of... Cutting the subsidy every single time. And, you know, we just have to go live with those things and either hope those mistakes don't bite us in the ass or fix them.
9:25Vlad CosteaWell, you can also argue that if it wasn't for this promise of a fixed supply and this consensus that there should only be 21 million Bitcoin according to this inflation schedule, The adoption would have been much slower, and especially from this libertarian Austrian economist school of thought, because cypherpunks, I guess they could find Bitcoin very useful. But these Austrians, maybe they would look at it and say, "Ah, It's not that interesting because there is no fixed supply.
9:52Peter ToddThat's an incredibly stupid argument. Because... Having a supply That grows at a fixed rate is a fixed supply. Like, you know, had Bitcoin been designed with 50 Bitcoins per block forever? The Austrians would all rally around Bitcoin saying, "Yeah, it's a fixed supply." Look, 50 bitcoins per block forever. Much the same way as the same Austrians rallied around gold, even though gold has an even less fixed supply, in that the amount of gold that miners can mine actually can go, you know, up quite a lot as technology gets better. But gold, they were willing to say, was a fixed supply, You know, it just happens to be that we had this 21 million meme, and people rallied around that.
10:31And they just cannot step back and look at it rationally and say, all right what is the economic properties of a system that works a little differently?" And the fact is, economic properties of tail emission is a fixed supply. It's a fixed supply of circulating coins. Same thing, demurrage is another example. Demurrage, you wind up with a... Basically fixed supply. You know, there's some subtle differences economically, but it's roughly the same thing. And had we had that from the beginning, that's what The Austrians would have rallied around.
11:01Vlad CosteaI usually think about Litecoin and the fact that This... Scrypt coin should have had issues if it wasn't for the Dogecoin, which became much more popular. And it was merged mining that actually saved it because the number didn't go up as expected. They didn't double in price Over the last. 7 years, I guess. They're still at 60 something dollars right now. And if it wasn't for merge mining with Dogecoin, they could not survive.
11:30Peter ToddSo... See, I actually disagree there. It's the real thing that kind of saves Litecoin is that it's so insignificant that no one's gonna spend a bunch of money to attack it. You know, this would be a very different discussion if Litecoin was the number one currency. But it's not. You know, it's not because it's not really all that useful doesn't get much usage, people don't go hold onto Litecoin, so there isn't really a strong incentive to actually go and try to kill the damn thing.
11:57Vlad CosteaWell, BitPay reported that they have more payments with Litecoin than Bitcoin and Lightning combined, so... You say there's no usage, but at the same time, we don't know who's using it.
12:08Peter ToddYeah, I'm always very skeptical when I hear figures like that. And particularly Who is holding Litecoin? Because after all, what drives the price of a currency is not people doing payments, it's people actually holding the currency. Now, you can argue that one reason to hold the currency is to do payments. But, you know, just as an example on my own, I've on occasion had to go pay for things, where Litecoin was the most reasonable option available to me. You know, they might sometimes they don't accept Bitcoin at all. Sometimes they go charge fees on it.
12:38Well, what do I do? Well, I use something like flip.me or FixedFloat.io to convert Lightning to Litecoin. You know, it's not me holding Litecoin for any reasonable amount of time. It's me just doing the bare minimum necessary to go pay with that. And that doesn't drive price.
12:54Vlad CosteaYou basically mentioned two competitors of my sponsor, which is SideShift.ai. Check them out.
13:02Peter ToddDoes SideShift.ai still require accounts? I thought they did at one point.
13:09Vlad CosteaThat's optional as far as I know.
13:11Peter ToddI I mean, flip.me, it just works, no accounts, and you do it over Tor, so... I can't speak for SideShift, but maybe I'll go try it sometime.
13:21Vlad CosteaYeah, it's pretty cool. Check it out. SideShift.ai. Anyway, I want to ask you about covenants because that seems to be I guess the hottest topic right now. People disagree in regards to which covenant soft forks should be implemented. They propose CTV, OP_CAT. There's another one that I'm missing right now. Anyway, the idea is that...
13:44Peter ToddThose are the main two. Technically, SIGHASH_ANYPREVOUT is a covenant. If you jump through a bunch of hoops, um... And then there's also a bunch of sort of more special purpose covenants. There are essentially variations on what CTV does like, you know, SIGHASH, or sorry, not SIGHASH, OP_TXHASH and things like that.
14:06Vlad CosteaYeah, Yeah, there seems to be an agreement that Bitcoin needs covenants to build some... New stuff. Are you sharing the same opinion? Do you think that covenants are necessary for Lightning, for Ark, For trust-minimized sidechains.
14:21Peter ToddWell, I mean, so... You have to parse that question very carefully because Bitcoin covenants are certainly necessary to do certain things on Bitcoin. But there's a bigger question of how necessary it is to do those things in the first place and, you know, something.
14:37I've been pointing out to people is like you can go open roughly a billion Lightning channels per year using available on-chain capacity. So it's likely you could maintain You know, in the hundreds of millions of Lightning channels, of actual users using it for all their stuff. And while that may not sound like the whole world population, it's a very big chunk of it. Keep in mind, like, out of the entire world population, which is, you know, roughly, I don't know, 10 billion people now or something, Only about 4 billion own smartphones.
15:08And if you don't have a smartphone, you're just never going to participate in Bitcoin directly anyway. So, I mean, It's just not... Like, it's just not relevant. All those people. What's relevant is the number of people who at least own a smartphone, at least can Participate in Bitcoin with an on you know with a a client that actually has self-custody. In that number, that 4 billion number, there's a good chance it's actually going to go down as world population decreases because If you ever looked at fertility rates, like, Fact is, basically every country in the world, which is halfway developed, is going to lose population and going to lose population quickly.
15:44So... You know, how much do we really need to rush into potentially changing Bitcoin in ways that are bad to get covenants for things that are only really strongly relevant? When Lightning has a significant percentage of the entire world. I'm not sure what the rush is.
16:02Vlad CosteaBut for some reason there was this report by Antoine Riard, back in November of 2023. And he presented this fatal bug in the Lightning Network, which requires some sort of soft fork to fix.
16:18Peter ToddWell, so he was talking about the transaction cycling attack. In a very clear way to fix it is my OP_EXPIRE opcode, which is not a covenant in any way, shape or form. So. And also, I mean, transaction cycling, you can certainly mitigate that to a large degree with changing how Lightning clients work um Basically, the way I put it is every time you do a round of the Transaction Cycling Attack... The attacker has to go spend money. So provided you have a client designed to force attackers to do a bunch of rounds of the attack, The attackers will run out of money.
16:55Um... There are probably edge cases where it isn't quite true, but you know this isn't a This isn't a serious enough bug that there's very strong incentives to do a software change. And like I said, the software change you would do is one that has nothing to do with covenants.
17:11Vlad CosteaI'm not sure if it was this disclosure or the emergence of Ark or other Layer 2s, you know, It seems like there's this What's the word for it? Not resurgence, but there is this movement where ZK rollups are fashionable for some reason with Citrea and stuff like that And Lightning has lost the spotlight. It seems like it lost steam. There are fewer nodes, there are fewer channels. Maybe for efficiency purposes they decided to close some of them. But there are not more users compared to last year.
17:45Peter ToddI can't agree with that statement of not more users. I mean, You know, Wallet of Satoshi stats as an example clearly show that number of Lightning transactions per day just keeps going up. Um... You know, I mean, the nature of Lightning is that you do not actually need very many nodes and very many channels support a very large number of users. So it's really no surprise to me that public node stats are going down because there's an enormous pile of interest in running Lightning nodes. I think by a lot of people who had You know, who just had the wrong idea about them making money through routing, which is not really going to happen when everyone and their brother runs a Lightning node and is willing to route for free. So, it's no surprise that's going down. But it does seem that Lightning usage is still going up. And it's not like switching to Ark is going to magically make the situation any better. Like Ark has many of the same ultimate problems that Lightning does, expresses some stuff differently, but the biggest problems I think are not really about tech, they're about Bitcoin adoption.
18:46Vlad CosteaI remember reading this blog post by Eli Ben-Sasson, who is a co-creator of Zcash and now he works on zero-knowledge rollups and this whole StarkWare, and he expressed this idea that he can bring these roll-ups from Ethereum to Bitcoin, and you are very much against it. And I did not understand clearly your point of view. Was it about miners being able to do reorgs with MEV or stuff like that or What is it that you don't like about the idea? Because honestly I interviewed him and he told me that you can basically sync a full node with a ZK rollup on your phone and I think that's pretty cool.
19:24And he also brings some...
19:26Peter ToddI mean, the way he's talking about syncing a full node, the nature of that math means that you aren't really running a full node in the sense you are today. Like he may, you know, he can go argue you are in the sense of validation, but there's more to nodes and validation. It's also about data publication, ensuring that the data behind Bitcoin is widely distributed. And really good ZK-SNARKs actually screw up the incentives of Bitcoin, because I can prove validity of a block to you without actually giving you the block.
19:52And that's extremely harmful. Now, I don't think the tech is quite there yet. But, You really got to take all the what all these guys say with a grain of salt because they've huge amounts of investment money riding on this tech being successful. And the fact is, that it does change incentives of Bitcoin in negative ways. And that's before I even get into the MEV stuff that will probably come as you get more and more complex financial shenanigans happening on top of Bitcoin.
20:21Vlad CosteaOkay, but more directly, what's your objection? Because You said that there's a lot of VC funding. You can also say that about Lightning or Liquid or something else. But what is it about their ZK setup that is wrong Did you look into it or something?
20:36Peter ToddWell, it's not about the ZK setup per se, it's at the... I mean, you know, there's potentially issues with that too, but like the fundamental issues is that the things they're trying to bring to Bitcoin change the incentives around Bitcoin. That's not necessarily good for Bitcoin users. You know, it may be good for the new users they want to go bring in, But, Those users aren't us. If you want to continue to do payments on Bitcoin and hold Bitcoin, What do you care about people doing fancy financial contracts for assets that are not Bitcoin? Like...
21:06You know, it doesn't benefit you. It benefits investors trying to do a bunch of clever stuff. But... That's not you. You know, that's not you doing payments. It's not about number go up Bitcoin. So, All this stuff we got to go ask. Does this actually benefit Bitcoin users? Now with Lightning, Lightning very, very, very clearly benefited Bitcoin users meant you could go send Bitcoin around the world nearly instantly at very little money. You know, that was not a hard argument. The only decent argument against Lightning was that potentially it would take transaction fees away from miners.
21:39And of course, We don't live in a world where miners are paid for transaction fees. We live in a world where they're mostly paid. From subsidy. So, if you want to fix that, the argument there is you go add demurrage or something, or tail emission and pay miners that way. Plus, of course, If you do have a Lightning channel, The cost you're willing to pay per transaction can go up a lot because you're using one channel for so many more transactions.
22:05Vlad CosteaI also want to ask you about drivechains because it was a big topic. I watched that panel at Baltic Honey Badger last year. And you made a very interesting statement, and I don't want to misquote you, but what I remember from it is that you said that the cost of increasing computation is much lower than the cost of increasing human politics. And I don't understand what this means. Is it that you mean that we can always increase the block size if it's needed or... What do you mean by increasing computation?
22:35Peter ToddWell, I'm just saying that In if you really do want to block size increase doing it through an actual block size increase is much less costly than doing it through a drivechain model which invites politics into low level consensus. Ultimately, you know, drivechains Unlike ZK-SNARKs, Does not have an actual mechanism in math to determine validity of anything drivechains assumes that you ultimately determine validity through politics, through voting. You know, it's really comes down, you know, it's really a miner voting proposal where we assume miners will vote on what chains are actually valid and what ones aren't.
23:11And that's always going to be political. And... It also introduced a very ugly type of politics where Miners Can easily go vote. To accept chains that are invalid and confiscate people's money. You know, currently... Miners don't really have much ability to seize funds. You can kind of argue, yeah, they can kind of play games with Lightning channels, but You know, even in those cases, they need to be a part of the channels like if you have a Lightning channel No matter what miners do, they cannot take the money out of the channel without colluding with at least one of us.
23:45Whereas in the drivechain model, If you and I have funds on drivechain, they just take the money and that's that. That's a much lower bar to seizure, especially like government ordered seizures. And we really don't want miners to be in that position.
24:01Vlad CosteaWell, but... There's there's always this argument that who are the miners? You seem to have an issue with Foundry being too big and getting too much of the hash rate. But honestly, there was always this argument that That you can always become a miner, I can become a miner, we can decentralize hash rate by contributing actively instead of pointing fingers and saying this is too big.
24:25Peter ToddWhy do you want to go and take a system where that's already our defense and say, "Oh yeah, the fact that in an emergency, when you do all these incredibly costly things, go try to maybe fix stuff?" Why don't we just rely on that more? Like, you know, this is like saying, well, you know, firefighters exist. Why don't we go make all of our buildings out of gasoline? And, you know, firefighters will go fix it. You know, we'll just spray some water on it when they catch on fire.
24:50It's like, dude, no, you do defense in depth. It's nice that, you know, you have these options if things go really badly, but you do not want to rely on them. You want to go and not need them in the first place. And the fact is, If if that if all this was true, You would kind of hope that maybe Bitcoiners would have already made hash power get more decentralized, you know, reduce influence of Foundry and AntPool and all this. But that's not how it's played out. Even though this is a risky place to be.
25:20Because the cost of doing all these things is so high, In particular, even if you're a small miner, a small man of hashing power, The cost in terms of variability of income is huge unless you go to some big pools. This is why, as an example, Braiins no longer makes their own block templates. Because they found they just did not have the... Consistency of finding blocks necessary For hash power to be properly sent to them. So they wind up partnering with another bigger pool. Who's the one who actually creates the block templates?
25:54You know, and then Paul Sztorc comes in and says, well, I mean, this situation is so bad. Let's just rely on this even more. It's nuts. It's completely reckless. And unfortunately, I think he's doing this because this is a way for him to go make money, and he just doesn't have any other good ideas in Bitcoin. Like, this is one of his two ideas. One is the, um... Prediction markets. Which is not really profitable idea. It's already been done and there doesn't seem to be a great way to do this with crypto.
26:22And the other one is drivechains.
26:25Vlad CosteaWell, it seems like Polymarket, which is on Ethereum, is catching a lot of traction these days. I don't want to get into that. It's just that prediction markets seem to have cycles, you know. All of a sudden nobody cares about them and then they have resurgence.
26:39Peter ToddThe problem with prediction markets is there just aren't isn't good ways to reliably do them in a decentralized fashion was ultimately the best you can do is turn them into a voting scheme. In which case they're vulnerable to people ganging up and voting the wrong way. You know, at least human-run prediction markets, it's clear who's ultimately deciding. Whether or not a prediction was true when you have When you try to do it in a decentralized way, it's really a voting scheme. Who decides who's going to have that pot of money?
27:08Vlad CosteaI recall reading a tweet by Super Testnet. Who figured out how to create drivechains using covenants. He used OP_CAT to construct these sidechains. And I think if we get this covenant soft fork, we're going to get drivechains without the BIP 300, possibly.
27:24Peter ToddWell, it depends on what soft fork. I mean, this this class of thing is a very good argument against OP_CAT. It's just such a general purpose tool that can do basically anything, especially if you combine it with a few other scripting changes that There's very good arguments that we shouldn't implement it. You know, you just, you allow too much stuff. And it's much better to do something simpler like CheckTemplateVerify.
27:48Vlad CosteaAndrew Poelstra said the opposite, that The negatives are outweighed by the benefits and It's 90% good stuff that's gonna come to Bitcoin.
27:58Peter ToddLook, 90% good stuff's all well and good, but the question is, what is the 10% and how damaging is it? And the 10% can be pretty damn damaging. You know, you are you are dealing with... Trillion-dollar assets. What is it? What's now? 1.5 trillion, half trillion? I don't know. I don't... haven't looked lately. But, you know, you're dealing with enormous amount of money. Are you really going to go risk this for people's pet projects? When it's clear that we can do covenants sufficiently good to do a bunch of simpler things that are clearly better, you know, clearly good for Bitcoin with a much simpler covenant model like CTV.
28:33Vlad CosteaSo, do you agree that CTV is great? Because I recall a couple of years ago when Jeremy tried to go through this activation process. That he got a lot of... Shit, I don't know how else I should call it.
28:46Peter ToddHe got a lot of shit because he didn't bother to do his homework. I mean the CTV BIP just Wasn't good. It had a bunch of examples that were clearly bogus. And, you know, if you're going to go and do a good argument, you have to go and do a good argument. If you're going to do a change to Bitcoin, it has to be well researched, has to be argued well. And the fact is Jeremy just hadn't done that. There's also other problems with CTV around fee payments, because CTV is just too restrictive in exactly what transactions it allows to be able to efficiently pay fees in a lot of scenarios.
29:19But, You know, when you compare it to something like OP_CAT, Which is both inefficient and enables potentially far more stuff than we want. Between the two of them, CTV is a more obvious pick. But like I said before, This isn't a rush like Bitcoin can already scale to hundreds of millions of users. A very big percentage of the entire world population that could conceivably be using Bitcoin with self-custody anyway. So, You know, we're just not in a position where we have to rush this.
29:50Vlad CosteaWhat do you think about the fact that we have Cashu and projects that bring Chaumian eCash on top of the Lightning Network? I think it's cool that you can receive money without having any inbound inbound liquidity, right? And you have some privacy, but on the other hand, it seems like we are reinventing banking. And to me, it sort of defeats the purpose of why I got into Bitcoin. Do you think it's cool? Do you think it scales Bitcoin?
30:15Peter ToddLook, if you don't like it, don't use it. I mean, it's as simple as that. Because we have Lightning that lets you go turn a Chaumian eCash token into a Lightning payment, and immediately get into your own self-custody. It's people's choice whether or not they use it. The fact is there are circumstances where... Cashu tokens make sense for some use cases. And people who are willing to use that are willing to take on that risk of having custody. In much the same way is that people use exchanges all the time. People have money stored in exchanges all the time.
30:47It's their choice like there we shouldn't we shouldn't try to go and prevent them from doing that And you know, in the case of something like Cashu, it's not like we can prevent them from doing it. It's just Lightning software. How would you stop Cashu developers from doing their thing?
31:05Vlad CosteaYou can't, but the market might. I don't know. It's just that it seems like it's overhyped right now.
31:11Peter ToddGovernments might the market isn't going to go stop Cashu because obviously Cashu is useful for some use cases. I mean, you know, I have a Cashu wallet myself. I go use it all the time. And one reason I use it is because there's obviously good privacy with that. And also because by me using it, it provides anonymity set for other people to also gain good privacy for that.
31:36Vlad CosteaOkay, so you like it. I also want to ask you about BitVM because it's super interesting. It allows for computation to be made off-chain and And it's pretty interesting. People have been tinkering with it and thinking of ways to build trust-minimized sidechains and stuff like that. I know that you have been thinking about sidechains for longer than a decade. And you're one of the first people to... Say that Bitcoin needs them.
32:00Peter ToddI mean, maybe I'm missing something more recently, but my understanding is there isn't a way to really build a sidechain with BitVM you know the BitVM model is closer to like a Lightning channel which lets you verify computation but that that doesn't directly... Allow you to do a sidechain the way that people normally use that term.
32:23Vlad CosteaWell Sergio Demian Lerner I guess you're gonna disagree with him because you don't have a good opinion on him but I had him on the show and he created BitVMX which is purposely created to scale Rootstock and allow it to reduce some of that trusted model.
32:42Peter ToddRootstock has had a trusted model for so long. I mean, obviously, once you add trust, there's all kinds of things you can do, but... You know if you don't involve trust and centralization that's when you get an interesting sidechain and So far, everything I've ever seen... About this has shown that you can't quite do that.
33:02Vlad CosteaWhat is it that gets you most excited about Bitcoin right now?
33:08Peter ToddI'd still have to go say Lightning. If I'm thinking pragmatically about things that you can actually use Bitcoin for, Well, things that you can use with Bitcoin to do real-world transactions, like or not... Lightning is a way to go and my suspicion is that Time will go show that Lightning was sort of a fundamental... Type of mechanism in computer science, in much the same ways that proof-of-work blockchains seem to be a fundamental type of mechanism you know and once you've discovered it There's this whole set of things you can do and...
33:44It'll be a very long time before it's obsoleted by anything else. Potentially never. Um... Now, similarly, like accounting in general. The idea of accounting was discovered... Couple hundred years ago and it revolutionized business and we've never gotten past it. We may have gone past some of the details how you do accounting, but the idea of, yeah, you go write all the stuff down on ledgers and like... Think about it that way, that will probably be with us till the end of time and I suspect Lightning something like Lightning with that kind of model of channels and something like HTLCs, is going to be with us till the end of time.
34:21Vlad CosteaWhat is something in Bitcoin that you think sucks and is a waste of time?
34:26Peter ToddWell, V3 transactions. It's like, it's, you know, it's kind of a niche thing. It's, um... You know, ultimately about transaction pinning, which is such a clumsy, ugly way to go fix a problem, and it doesn't fix the problem very well. You know, just says, well, why don't we have a new type of transaction version number? That We go put these artificial limits on to try to prevent transaction pinning from happening. Although, because the limits are fairly big, You know, if you use a v3 transaction, even in the best case, like I can, It's a typical sort of Lightning case.
34:59Depending on how mempool fees are, if they're reasonably flat, like I could, Increase your cost to get a to get your transaction mined by like 5x. That's just an ugly thing to go have. And to do this while also messing with the transaction version numbers, which are quite important for consensus upgrades. Without having a clear plan of doing that, when replace-by-feerate is just so much cleaner and simpler. Like, I think it's... You know, it really says that Bitcoin politics, um, sorry, Bitcoin Core politics has gone into a really bad place where people are afraid of calling each other out.
35:33You know, people are afraid of stepping on each other's toes, and there's an enormous sunk cost fallacy. People have worked on v3 transactions for a long time. Haven't gotten much external review. Then they get external review which says, "Yeah, this is a bad idea. We can do this much better." They say, "No, no, no we got to go with this thing." Like... Sometimes you got to admit that your bad idea is bad and to go move on and I I think we'll probably see v3 transactions get shipped and then we'll just forever have This ugly wart in transaction version numbers, where some for some reason version 3 is allowed, but there's no difference.
36:09And we'll end up with replace-by-feerate in the end anyway.
36:13Vlad CosteaThere are a bunch of people recently who discovered that exchange of emails that you made with John Dillon about 12 years ago or something, and they're calling you a spook. Amir, who is somewhere around here, said that...
36:26Peter ToddYeah, the... Well, the Amir thing bothers me a lot, because Either he can't read properly or he's just straight up lying. Because, you know, he was trying to claim it as like I was working with, you know, This Dillon guy, to go prevent him from getting his CoinJoin bounty. It's like, dude, Dillon donated like five Bitcoin to this bounty. And... You can easily read this. Just read the fucking leaked emails. It's not hard. But, you know, I think Amir knows that...
36:55If you just go say something on Twitter, people aren't actually gonna go back to the original sources and spend the time reading through them because It's real pain to go do that you know and I challenged him on Twitter over this like dude cite your sources he just hasn't so the whole thing's the whole things very stupid. But... You know. It's always it's it's been a long, long BCash kind of pile of conspiracy nonsense, and BSV as well, you know, and of course the They go bring in like replace-by-fee in here, which, you know, of course, replace-by-fee. That's a thing that really killed Bitcoin on-chain payments, or I should say Bitcoin merchant payments. You know, nobody can accept Bitcoin now as a merchant.
37:33It's like, dude, This You're just so far off how people actually use Bitcoin, but... It is what it is.
37:41Vlad CosteaI've interviewed Rafael Cordon and he really praised you for OpenTimestamps and advising the Guatemala, it was Guatemala, right? The Guatemala government during the elections, which were contested multiple times and it was A big deal that they were able to verify the validity of the votes with OpenTimestamps. So can you explain what you did for the government and how this works and how it can be used for US elections possibly?
38:07Peter ToddTo be clear, I wasn't working directly for the Guatemalan government. I was working for... well, I guess it would be like a subcontractor to a subcontractor to the, you know, voting, you know, electoral commission agency, but. Yeah, like, you know, what we did, well, I mean, you know, for background, Guatemala, like any sane country, it's a paper ballot system. So... You have paper ballots and they get put in boxes and the boxes are taped shut. And at some point after the polls close, those boxes are opened up in front of a pile of observers and people count the vote manually.
38:42Now, the results of those per box counts go into tally sheets. Yeah. Piece of paper. Like any other, I think it's... A quadruple, triplicate, one of those, like, multiple thing forms, whatever. But The important thing is when people actually physically write down the totals, against every candidate in that particular You know, for that particular box. Then what happens? Since this is like triplicate paper, One copy of course gets kept. I think another copy got signed in some other way. And then one copy Ends up getting scanned in to a scanner, connected to a computer, That sends the results up to election headquarters.
39:24And that's the moment when the timestamping happens. So all of the computerized stuff, that's a preliminary elections results database. And that's when we do the timestamp. That's also when it gets recorded in sort of the main databases that of course feed into the TV programs announcing who's You know. Who's winning what? And the relevance of timestamping here, Isn't Well... It's a little less obvious than you might assume, because In the Guatemala presidential election, there was voting fraud that happened. But, I may be butchering this bit because of course I'm I've heard this second hand from people who are actual Guatemalans.
40:06Basically speaking, what happened is the ruling party realized they're going to lose. And what they wanted to do was commit just enough voting fraud To throw out the whole election and get a do-over in another six months. Because, you know, six months more campaigning, maybe they'll win next time. Well, with OpenTimestamps, We it was one piece of evidence that showed hang on a second nearly all the country The ballot boxes are getting opened and counted when we expect them to, right after the polls closed. You know, usually within 10, 20, 30 minutes.
40:37Counting a pile of votes is not hard. It can be done very quickly. But, in some parts of the country, This scanning process happened hours later. And on top of this, we have eyewitness evidence saying that the boxes were stolen and returned later. And we have video footage, evidence, etc., etc., etc. And timestamps, of course, we're just one part of many pieces of evidence saying, "Yeah, There was voting fraud that happened there. And of course some of the... Pieces of paper there that even have ridiculous things like the numbers crossed out and new numbers written. Like, dude, come on. But... For most of the country OpenTimestamps is one of the pieces of evidence that said, yeah, all the other votes were done correctly. And the part of the country with voting fraud was small enough we don't need to rerun the elections. And long story short is after months of agonizing legal bullshit and people I worked with, you know, some of them fleeing the country, some other people in election commission getting arrested, etc, etc, etc, Eventually, the party... Who won, legitimately, finally got into power. But, You know, it took a while.
41:47And fortunately OpenTimestamps was there to add one more piece of evidence to let this whole process happen. And since then, my understanding is Mexico, They have a very similar system. The way it's all organized. They also had the results timestamped, although not as officially as with Guatemala. El Salvador really should have. But my understanding is... In El Salvadorian politics, they didn't really want to mix up the election with Bitcoin, Because from sort of the external anti-Bitcoin, anti-Bukele, yeah. From that kind of side of things, they think, oh, they're using Bitcoin to steal the election.
42:26They just don't want to touch it. And maybe this stuff will wind up in parts of the US election too. But we'll see.
42:34Vlad CosteaSomething for which you never get credit is that you indexed with timestamps the entire internet archive, and that was in 2016 or something. And I just want to get it on record because I guess You get a lot of hateful comments sometimes, but you did some very cool stuff that doesn't get mentioned. But how does that work? How do you timestamp the entire archive?
42:56Peter ToddWell, so they had, so Archive.org had an API that let you download metadata for every item in their collection. And... While it's not publicly accessible, The metadata is available for their web crawls. So, you know, every time The... Wayback Machine does web crawl. That gets collapsed into big files essentially of many different web crawls and that metadata is available plus the Internet Archive. It's not just about the Wayback Machine, not just about websites, it's also about things people upload. So there's many, many, many sources of... Data of all kinds of things you know scans and magazines like videos it goes on and on and on Old computer games, all this stuff, is the metadata is available.
43:41So of course I download all the metadata. I put into a giant Merkle tree and I just timestamped it. And then I create a software to make those timestamp proofs available for every single item of metadata. In the in You know, not quite the entire Internet Archive, because due to Due to flaws in the archiving system, not all the metadata could actually be downloaded. And then I also added a system that at least tried to Watch new things as they came in and then also timestamp them. Unfortunately, I just...
44:14Didn't have the budget to do this fast enough with fast enough computers to actually keep up so in the end I you know I gave up because I was my my server running all this was falling behind but you know I'd love to go revisit that with better software and faster servers so You know, it's been quite a few years now and One of the things that has changed a lot is NVMe.
44:38SSD drives are much are much easier get and they have far higher IO bandwidth and the stuff I was using back in the day so it would actually be much more feasible to do this project again
44:51Vlad CosteaAnd it's really useful even though it's not the data per se it's just timestamp to prove that the data existed there Right?
44:59Peter ToddYeah. Well, essentially... The way it was set up was that any file in the entire Internet Archive you could go and paste in the hash, the SHA-1 hash of it, and you would get back an OpenTimestamps proof. Um. Obviously some of them would go fail because I couldn't get quite 100%, but for a while it was pretty good uh These days, of course, I've missed a lot in the year since. But I would love to do do that project again. I just need I just need funding, ultimately, for... You know, dev time into...
45:27Buy the servers necessary to do it.
45:30Vlad CosteaLet me ask you a question to get you cancelled. Name three cryptocurrencies that are not scams and might be interesting.
45:38Peter ToddWell... I mean, Bitcoin is an obvious one. Monero is an obvious one. And then it gets kind of hard. Kind of Zcash, although it hasn't really caught on enough. But yeah, that's probably your your best three that you can... Give an example of going If you asked me to do a fourth, it would be very difficult. Very, very difficult three is not too hard four is very hard.
46:03Vlad CosteaIt's interesting that you still like Zcash. You are part of that ceremony. I happen to own that laptop and I'm going to hold on to it because it's a very nice cypherpunk artifact. But for a while, I guess you are not in good terms with the guys, and maybe it was personal.
46:19Peter ToddIt was more than personal. I mean, when the Zcash people, you know, tried to go falsely accuse me of rape to go and get back at me for pointing out the trusted setup hadn't been done properly, among other things. So, like, some of those people are just fucking nuts. But, you know, it is what it is. Like, I'm not going to say that the tech is bad because the people behind it... Are just straight up crazy. So... The fact is the Zcash tech is interesting. It would have been more interesting had more competent people and less, you know more sane people being behind it, but the tech is still interesting and what I've heard is Monero is looking to adopt the same kind of proof that Zcash has.
47:01The technical details will be different, but the important thing from the point of view of the users is that the anonymity set for Monero transactions with this potential upgrade that might happen would be the entire set of coins. And that's always been the thing that Zcash had that no one else did anonymity set truly of... All possible coins. Modulo, of course, timing issues. I mean, good example is like, in this one, the failings of Zcash, because it's got so little use. If you deposit into Zcash, and then go spend Zcash.
47:30You know, a couple hours later. If I know roughly when you might have done that, I can, of course, narrow down the potential set of Zcash transactions quite a lot because there just aren't that many per day now Monero has enough usage that once they add this every coin could have come from everywhere. That'll be quite a nice privacy improvement. Currently, Monero is really like CoinJoin with some improvements. But they you know with this, they could do a lot better.
47:57Vlad CosteaDid you have time to look into the MimbleWimble extension block that they put on Litecoin. I think it's pretty interesting because they have addresses They have this peg in peg out system that is just like on sidechains So they They
48:11Peter ToddActually hired me to go do the review of this. I never actually sent them the bill because there wasn't really much for me to actually say about it. It was useful. The big problem with this is MimbleWimble just isn't that interesting technology uh... You know, it's ultimately kind of like a CoinJoin. It just doesn't... And because of how the CoinJoins are created, it doesn't actually provide very good privacy. So... The other thing too is I think people have often had a misconception that MimbleWimble provides really good scaling.
48:42And it just doesn't do as well as people think it does because it doesn't let you, you know, people think it lets you combine all transactions into one. But it's not true. It lets you combine... Many things into one... But you're still left over with a certain amount of data, essentially per transaction, that can't be compressed away. And that data adds up.
49:05Vlad CosteaYou've been in Bitcoin for about 14 years or something. Yeah. And I want to ask you, where do you see Bitcoin in the next 10 years in terms of technology? What do you think is going to be the next big thing? Are we going to be using ZK proofs? More or do you think there's going to be some new sort of cryptography that's going to make transactions smaller?
49:28Peter ToddI think the most likely big change that we would see in terms of what average people use it is probably something like Ark. And even then, there would probably be Ark being used to then run Lightning channels. You know, so I do not see Lightning going away.
49:45Ark might be a basis for making Lightning channels a bit more convenient and having less on-chain impact But... Doing better than this is hard you know, ZK is like A good example of this is, ZK proofs don't, like... For average of retail, if you will, use of Bitcoin, ZK proofs don't help that much. Because they still can't make your transactions confirm quickly. Lightning is what you need to actually do an instant confirmation. Just doing ZK proofs and say a ZK proof blockchain doesn't actually get you that. So it ultimately is very hard to beat Lightning's model. I mean, When you get down to if you want the fastest possible transactions, Ultimately, you're limited by speed of light, and if you're in an environment where you're limited by speed of light, You have to have something like Lightning because you're actually routing between physical entities.
50:38And Speed of Light. Isn't that fast? If you want like the instant payment experience, You're going to wind up with something either Custodial, like the eCash stuff, or something like Lightning, With very fast routing.
50:54Vlad CosteaThat sure is fascinating. Thank you very much, Peter. And... What are you currently working on and where can people follow that?
51:02Peter ToddWell, I mean, and I talked about this earlier today. I'm still working on the replace-by-feerate stuff with Libre Relay. I've I got this big paper I'm writing comparing like L2 protocols and covenants and so on. I'd really like to do more work on OpenTimestamps, but currently I'm trying to find funding for to go... Do the stuff I want. I applied to OpenSats and got rejected for that, unfortunately. But there's probably a way somehow. So, you know, I think that mostly covers it.
51:31Vlad CosteaThank you very much and best of luck. Thank you. Hey you! Thank you for listening to the Bitcoin Takeover Podcast. This episode is sponsored by SideShift.ai, StarkWare, LayerTwo Labs, and Hodling.ch. Don't forget to subscribe!
