0:00 Hi everybody, welcome. This is the panel on transaction censorship. My name is 0:05 Craig Deutsch. I am the editorial project lead at Bitcoin Magazine Pro, and I am 0:10 joined by Nick Hansen, the CEO of Luxor Technologies. 0:22 Also joining us is Udi Wertheimer, co-founder of Taproot Wizards. 0:29 And Paul Sztorc, the founder and CEO of LayerTwo Labs. 0:51 Thanks all for joining me. So this panel is about transaction censorship, but 0:57 maybe it should be about wardrobe censorship. I do a little joking. 1:03 So let's go ahead and just start about how blocks are constructed. Right now 1:09 Stratum V1 is the main mining protocol for communication between miners and 1:14 pools. Nick, as our resident mining expert, can you just talk about how blocks are 1:19 constructed and some current limitations for individual miners who rely on the 1:23 pools to build those blocks? So blocks are constructed by analyzing the mempool. 1:29 So every node out in the world has a view of all the transactions that are 1:34 pending. Basically the way it works is all the transactions that have the 1:38 highest fee are filtered up to the top, and then we take that portion and create 1:42 something called a block template. That block template will be the proposal that 1:46 will submit to the network once the proof of work has been demonstrated. And 1:51 then once that block has been solved, it's added to the chain. The rest of the 1:54 nodes, you know, all of the nodes that I'm sure a lot of you are running, will 1:57 actually get that block, validate that it's accurate, and then the mempool state 2:02 will change. And then we just do that forever. So there was some criticism 2:09 when Stratum V1 was implemented because it removed the get block 2:16 template, which allowed miners to construct their own blocks. How does 2:21 Stratum V2 solve this, or does it solve this? Can you talk a little bit more 2:28 about how Stratum V2 changes the way that blocks are constructed? 2:33 Okay, so Stratum V1, basically all of that stuff that we do to 2:40 construct the block template, we send that out to the miners over a protocol 2:43 called Stratum V1. They have no input into what goes into that block. Stratum V2 2:49 implements something called job negotiation, where basically up front the 2:53 miners and the pool agree what the transactions will be within that 2:58 block. And if there is a transaction that isn't in the block at that time, the 3:03 miner can propose it through a protocol called job negotiation. Stratum V2 3:07 implements a lot of interesting things that I do think 3:10 Stratum V2 is a meaningful and important improvement. But when people 3:14 talk about Stratum V2, really the only thing that they're talking about is job 3:17 negotiation, because that does introduce a way to reduce the censorability of the 3:22 network and introduce transactions to the mempool that wouldn't have 3:27 otherwise been there. Because for whatever reason, maybe the 3:31 mining pool is censoring those transactions, either for regulatory 3:34 reasons or otherwise, the miner can propose them and add them to the template. 3:40 Thanks. So part of the idea for this panel came about in response to the 3:45 ordinals and inscriptions craze and whether these transactions should be 3:50 considered spam. From detractors there's been ideas for filtering transactions 3:56 using node policy. Right now we already have some examples of filters. So one 4:04 sat per vbyte as a minimum fee is a node policy. It's not a consensus rule. 4:09 The dust limit is a node policy. And now there's been a recent node patch which 4:14 is or disrespector. So similarly to the previous examples, it only prevents the 4:20 relaying of those transactions. Udi, what are your thoughts on these types of 4:24 filters? Would you consider this a type of censorship? I consider it role 4:32 playing games that people like to play in their basements with their little 4:35 Raspberry Pi nodes. I don't see how it relates to actual mining. 4:39 I've never heard of a miner who's running or mining pool that's running or 4:43 disrespector or a filtering patch like that. But of course if they 4:47 want to, they're more than welcome to. But I think they'll find very quickly that 4:51 it's not very profitable and fairly expensive for miners to try to censor 4:58 these. So I don't expect that they'll want to. I've never heard of a miner who 5:00 is interested in doing that. I have heard of a lot of Bitcoin enthusiasts 5:06 who are not necessarily miners who want to run those on their own computers 5:11 which is fine but doesn't seem to be really relevant for mining. And I guess 5:19 the big question for miners is whether miners should be maximizing 5:27 revenue at all costs. Like should that be their main goal? And if not, then what 5:34 exactly should they be doing and why? Paul, can you speak to that? 5:39 The answer is yes. The miners shouldn't revenue maximize, of 5:50 course. Anything else has the feature that it would be contrary to 5:55 what the miners want to do. So there's no reason why they would just, should 5:58 miners just give away free money in the form of not taking the ordinal, the 6:04 opportunity cost of not taking the ordinal transaction? So that's really 6:07 what it all boils down to is will miners just give away their own money for free 6:11 for no reason? And that is the whole transaction censorship question. I had a 6:17 question about a stratum v2. I was going to ask Nick but now we have the two 6:21 people with one microphone but we're gonna give it a try anyway I think it 6:23 will be about, the question is stratum v2, the job negotiation. The whole point 6:27 is the pool, the mean pool, the reason why everyone loves it is the mean pool 6:32 censoring and the person, you know, some small miner can like propose, they can 6:39 like trump the pool by getting the block template they want in anyway or something, 6:45 they can work on their own thing. And the tiny miner, if the pool and the tiny 6:50 miner disagree, the tiny client of the pool, they win. It seems to be how it 6:57 must work or perhaps not. But if that's the case then the pool can of course 7:01 just mine with 0.01% themselves and they can censor anyways and then they 7:05 escape all the blame. So that is my question to you. 7:09 Yeah, thanks for the question. I think there's actually a lot there. So actually in 7:13 stratum v2 when you do job negotiation there will be multiple templates being 7:17 delivered to each different miner. So, you know, the pool will deliver what it 7:21 calls the default or de facto template which is delivered to every miner and 7:26 then if you go through the job negotiation process, the miner that 7:29 proposed that, that will not go to the rest of the, that will not go, most 7:33 likely will not go to the rest of the miners. It will be that miner working on 7:37 it. If they happen to hit the block then their transaction is included. The reason 7:40 that I don't really think job negotiation is going to improve 7:43 censorship at all and the reason being is that the censorship comes from a 7:47 regulatory, most likely will come from a like a regulatory overstep, meaning 7:53 you're in an environment, you know, you're in a country that wants you to sanction, 7:56 let's just say OFAC transactions, you know, there's a whole 7:59 list of addresses that are sanctioned by OFAC. If you send Bitcoin to them you're 8:04 breaking US sanctions, which is a very bad thing to do. And so there's a proposal 8:10 or there's been some thought around should we, you know, should mining pools 8:13 censor those transactions if they see them in the US. Let's just say that they 8:18 do. Most likely in the job negotiation protocol what's going to end up 8:21 happening, even if the miner proposes one of these illicit transactions, we'll call 8:27 them, most likely the pool will just reject it and say actually we're 8:31 not going to accept that transaction anyway. And so what strategy two then 8:35 introduces a very easy fall over so you can fail over to another pool, propose 8:39 that transaction that way. But then in that way it becomes a little more overt 8:43 that a mining pool is censoring. Right now it's not really possible to tell if 8:47 a mining pool is censoring transactions because it's not overt. It could just be 8:51 my node's not, that pool's node is not very well connected, that hasn't 8:55 propagated yet. But most likely job negotiation, the portion of strategy two 9:01 called job negotiation, is not really going to improve censorship at all 9:05 because of the protocol effectively giving the pool the right to say 9:10 no to those transactions. The pool always has the trump card, I would think. I mean that must 9:19 be how it works. But then you have an intriguing situation where 9:23 if anyone, if you have a full node and you have had a transaction in your own 9:27 mempool that would have met the cutoff for inclusion as far as the fees. If 9:32 you've had it for more than like 20 minutes say, you must know that it has 9:37 reached every pool and that they have passed on it and in that way stratum v2 9:41 doesn't seem to be an improvement. See what people are, you know, everyone, what people love 9:45 about stratum, like as you and I agree that stratum v2 brings a lot of useful 9:48 things, but the people who talk about it on Twitter they just really think it's 9:52 about the mean pool is censoring and we get to do something about that. But it 9:57 doesn't really improve that situation I think. Maybe you agree. So I guess since 10:04 we're on this thread, yesterday mempool.space announced an accelerator 10:10 and right now when you go to broadcast a transaction you look at fee 10:16 estimation, you might check mempool.space or another block explorer to see the 10:21 current fee rate to get into the next block. So how does this type of 10:26 accelerator, which already exists with other mining pools, how does this 10:31 type of accelerator impact fee estimation as well as like potential 10:38 transaction censorship or does it? This is a very important just topic overall 10:42 because it's kind of a canary in the coal mine as far as if people aren't 10:46 getting what they want out of all the existing tech stack of Bitcoin like the 10:51 mempool for example, they will just build some other thing that gives them what 10:55 they want and that is why the whole idea of transaction censorship is very 10:59 simple which is as to the miners will revenue maximize because that's their 11:04 bread and butter and anything that tries to stop them. That needs to have a very, 11:09 this is often, I think this is modeled because it's like SegWit2x and 11:13 other things where the value of Bitcoin is maybe threatened by the 11:20 protocol becoming less useful but the difference when it comes to this 11:24 transaction censorship question is that the miners have the final say on what 11:27 makes it into, what was seen by the network by a certain amount of time. So 11:31 that is the whole game of mining and they have the final say on that and so 11:36 they can always say that they just didn't get a message and censor it if 11:40 they wish and they can also include anything, they can include a Bitcoin 11:44 transaction so this is where we get into maybe stamps instead of ordinals but 11:48 because of something called Alan Turing's universality of computation you 11:52 can just reinterpret any data as something else so it's like if you have 11:55 a JPEG file and you just, you're like on Windows and you change it to like a .txt 11:59 file and then you open it in notepad, people will always be able to just 12:03 transform anything they want, any kind of data, image, whatever into Bitcoin 12:08 transactions and then just broadcast them on Bitcoin. So everyone who is 12:12 trying to say that there is something other than a Bitcoin transaction in a 12:16 Bitcoin block and anyone who is trying to stop the miners from profit 12:21 maximizing or revenue maximizing, they're just asking miners to give away 12:25 money for free and that's basically the whole story as I see it. I feel very 12:31 unsportsmanlike holding the one microphone so I'm going to give it back. 12:34 We can switch chairs and then it'll be easy for you too. Totally. You know 12:43 on that point of should miners you know censor ordinal transactions or JPEG 12:51 transactions or whatever, we talked about this before, I don't even think 12:54 about it in the context of should or shouldn't. I think they cannot censor 12:58 them because if they do, it hurts their profitability and eventually just they 13:03 would be competitive. So mining pools cannot censor it or they're gonna lose 13:07 in the competition to the miners who don't. So it's not even, there isn't 13:11 really a choice. Mining pools have to include those things. Yeah exactly so 13:17 I was on a Twitter space talking about you know the big wizard and other 13:21 ordinal stuff and the very angry Bitcoin guy got on stage and asked me well why 13:27 aren't you running or disrespect or you should turn that on you're attacking 13:30 Bitcoin why would you turn that on? I said well I'm the only CEO of a mining 13:35 pool that's here talking to you about this stuff. All the other ones are stoked 13:38 that this is happening because they're making a ton of money and a ton of 13:40 revenue for their miners. You're asking me for to effectively reduce the 13:46 revenue of the customers that I have which are the miners. My job is to 13:50 steward their hash rate and produce the most value for that hash rate as I can 13:54 and by turning on or disrespect or I would be effectively you know doing a 13:58 disservice to my miners and you know effectively my customers and that's 14:02 gonna put me out of business because none of the other pools are going to do 14:05 that either. So there is really like you know like Udi said you're effectively 14:10 asking the mining pool to reduce their potential revenue and it's just not 14:14 going to happen because that is I mean that's the way the consensus is built is 14:20 to maximize revenue and to maximize profit from the network. So I'm gonna ask 14:25 you a follow-up question on that. When Luxor mined the first taproot 14:33 wizard, the very large one that got all of the Bitcoin maximalists up in arms 14:37 part of what some of them at least were upset about was the fact that the fee 14:43 was set at zero sat speed byte so it was an out-of-band transact payment. So I 14:51 guess one of my in going back to the like the mempool.space example with 14:56 the accelerator are those payments like made out of band is this a way that 15:01 mining pools could be taking additional revenue without potentially having to 15:06 disclose that publicly on the time chain and for people to for their miners to be 15:13 receiving the fees from that? I think it's very straightforward that this is 15:17 an example another example of what I mentioned before of the existing system 15:21 doesn't work so if you have if people are paying out of band then miners want 15:25 that miners control they have a certain amount of block space that they want to 15:28 sell someone wants to buy it we had this public mempool but if it's not going to 15:34 facilitate that transaction then they will just do the out-of-band payment so 15:39 yeah I think if the you know if the laser a maxis want to have efficient 15:44 transparent reporting of how transactions flow and you know fees and 15:50 so on the mempool will have to become more useful right now you know there's a 15:54 lot of people who want to you know for whatever reason they want to inscribe 15:58 four megabyte transactions there's a bunch of them they cannot do that with 16:01 the mempool so all of that you know that information will not be available in the 16:04 mempool before it's mined after it's mined you're not gonna know for sure if 16:08 the fee that you see is the fee was actually paid if you think that those 16:11 are things that are valuable then the mempool is gonna need to you know be 16:14 improved to accommodate for this because people are gonna keep doing that because 16:17 they want to we can't stop them up the protocol allows for it so the question 16:21 is are we gonna let are gonna have the mempool serve those people as well or 16:25 we're gonna say that that they shouldn't be part of that mempool for whatever 16:30 technical reason which is which is also okay but that would mean that they would 16:33 have to bypass it because they're still gonna do it yeah transaction accelerators 16:38 have been around for a long time but they were mostly for like a transaction 16:43 you put you accidentally put one sap per V by and you know you need to get it 16:46 through so you can go through and pay the pool out of band this has been 16:49 around for a long time by BTC has it on their front page a lot of mining pools 16:53 have done this throughout history now the problem is if it becomes prolific if 16:58 that becomes the way that fees are paid to mining pools that starts to look kind 17:02 of like an attack on Bitcoin the other pools so I don't know if people don't 17:06 know the most pools now are paper share meaning you pay the expected value of 17:11 every share that's submitted to the pool regardless of whether you find blocks or 17:13 not it you know maybe 94 to 96 percent of all hash rate in the world flows to 17:18 PPS based pools that PPS calculation includes transaction fees so if there's 17:24 a mining pool that's taking a bunch of out-of-band transactions they're 17:27 effectively reducing the PPS rate for the rest of the network and the rest of 17:31 the pools which then will drive all of the hat all of the hash rate to the one 17:35 pool that's doing all the transaction acceleration and so like with what men 17:39 pool is do it men pool space is doing they're going to effectively make it way 17:44 more centralized or they're going to put apply a centralizing force to the hash 17:49 rate that exists today because they're going to make all the other pools less 17:53 profitable now Luxor hasn't subscribed to the transaction acceleration program 17:58 that men pool space has I believe they announced it was foundry I'm not sure 18:01 but that's just going to continue to accelerate the decentralization of hash 18:07 rate if that is the case if that becomes a very prolific way of doing you know 18:13 transaction propagation and transaction fee payments so let's talk about the 18:18 centralization of mining because it becomes more easy for transactions to be 18:25 censored if miners are centralized and yes but what does it mean that what is 18:31 the definition of minor centralization it's a good question it's a good 18:36 question we right now we have three pools that have a significant share of 18:41 hash rate and I don't know what the percentage of that is offhand but it's a 18:45 it's a lot but does that is that because that's just a certain grouping you know 18:48 there's just like maybe they have it's like just just because someone runs a 18:54 pool like he just gave you a perfect example of someone who controls a pool 18:57 they run they own a pool but they have very limited control over that because 19:00 they have a they will lose clients immediately to a rival pool if they if 19:07 the pool doesn't you know so what we want is it to be as cheap and easy as 19:11 possible for someone to start a new pool new competing pool that stops the pools 19:16 from misbehaving the same reason why we want miners to revenue maximize in 19:23 contrast to an older dispute such as SegWit2x we want the software developers 19:29 to write software code that maximizes the value of Bitcoin it's a very it's a 19:34 parallel situation but I think it's the lesson of the SegWit2x era was that 19:39 miners should like be very tame and just do whatever they're told or something 19:42 like that there's still this cult lingering cultural baggage from SegWit2x 19:47 that has carries through to the present day but the miners must at 19:52 revenue maximizing we wanted it to be as cheap and easy as possible for someone 19:55 to start a new pool I think sometimes people use a definition of minor 20:01 centralization that goes something along the lines of if one minor becomes more 20:05 profitable for any reason and puts another minor out of business then that 20:09 is mining centralization which I think that that definition cannot possibly be 20:13 useful because that is like just happening all the time every two weeks 20:17 every time the difficulty resets it fires the bottom half of the mining 20:23 performers so so I guess I guess my question stems from like a concern 20:30 around that being so an accelerator being so successful that so much of the 20:37 hash rate goes to one particular minor or a couple particular miners because 20:41 then it opens up the network for potentially more censorship I think we 20:47 should bet everything on the censorship question but isn't it interesting that 20:50 the idea that if you revenue maximize you're going to include as many 20:54 transactions as possible and if you can only include so many you'll put the ones 21:00 that pay the higher fee first to me that seems to be the definition of not 21:06 censoring so the the fact that someone would block a transaction for some kind 21:11 of a political reason or that we give these other people influence over like 21:15 someone in a bureaucrat in a different country influence over the Bitcoin 21:21 protocol that would be like one kind of censorship but then also the or 21:24 disrespectors you know we're doing a similar thing if you just say that or 21:28 disrespectors are all from a different country and they're all lawmakers or 21:31 something like that so the question of it's really all just about the revenue 21:35 maximization of the miners which is why I think it's well I mean don't you think 21:39 that like what do you think of like is it mining centralization when very many 21:43 people join a few pools if all the other pools are irrational and if in 21:49 particular if they censor what if all the other pools censor and there's only 21:53 three that don't and everyone joins those three that's is that mining 21:56 centralization I mean it can become mining centralization if them if it's an 22:03 overwhelming majority but I think like it goes back to like is that are they 22:08 acting in their own best interests and does that also act in the best interest 22:13 of Bitcoin I mean one of the fascinating things about Bitcoin is it 22:16 is economically you're economically incentivized to act in the best 22:19 interest of the network at large so I want to I do want to go back to the the 22:24 mempool and how that is specifically not able right now to function with like 22:29 four megabyte blocks and things like that and can you just explain why that's 22:33 the case because for people who might not understand right so so the way 22:39 transaction propagation works by default in addition to you know 22:45 validity rules that just check that a transaction is valid there's also what 22:49 you know is called in the technical terms is a standardness rules and that's 22:53 basically policy rules and they basically affect how transactions are 22:57 allowed to be propagated through mempools of the Bitcoin network node 23:02 Bitcoin no network into eventually a mining pool and that peer-to-peer 23:07 network does not allow for propagation of transactions that are above 400 23:13 kilobytes so again it's not because the transaction is invalid it is valid but 23:18 the peer-to-peer network will not propagate it unless it gets into a block 23:23 so that just means that if you do want to get a four megabyte transaction into 23:29 the network like we did and like you know a bunch of others have done after 23:33 us then you have to do it you basically have to contact the mining pool directly 23:38 to do that and that kind of reduces the efficiency of the mempool yes it's just 23:44 magic ignore it it's not it just it kind of it kind of makes the mempool a little 23:51 less sufficient because it means that it doesn't serve everyone needs some you 23:56 know some users of the Bitcoin network cannot use it and it doesn't fully 23:59 reflect what the actual you know state of transactions that are waiting into 24:06 inclusion into a block so it's it's inaccurate and it is also creating you 24:11 know eventually I think for mining pools the the equilibrium they would probably 24:17 want to strive for is that everyone gets an equal opportunity to mine those juicy 24:22 forming about transactions and not just you know the the the pools that get 24:27 those sent directly to them right so those kind of things clearly can hurt 24:34 you know decentralization of mining but it's really up to mining pools to you 24:40 know they could they could all connect in and decide to receive four megabytes 24:45 transactions they could start their own alternative mempool there's a lot of 24:48 things they could do I think that as Ornels become more popular as 24:54 inscriptions become more popular and I fully believe that they will you know I 24:56 it's funny I've heard a bunch of people saying on panels that oh this is a fad 25:00 I've been hearing for three months that it's a fad 25:03 I've been hearing for three months that it's a fad and the fad seems to keep growing I really think they're here to stay and I think that mining pools are gonna have to start thinking about this more seriously because again this isn't just I feel like right now people in the mining space that I've talked to seem to think of it as like just a juicy booster to their revenue which it is but also it kind of requires involvement because mining pools that will not optimize for this will 25:30 end up being uncompetitive compared to the ones that do yeah the so the mempool policies there's 25:36 actually an alternative implementation called Libitcoin by Eric Bosco they remove that entirely 25:41 because they think that any any transaction that is that is valid and meets the consensus rules 25:46 should be allowed to be propagated the reason that we don't propagate those now by default is 25:51 because their DOS vectors DDoS vectors basically if you're spamming four megabyte transactions to 25:57 everybody you know it takes a lot to propagate and validate those same with zero fee you know 26:03 it's a very easy way for you to spam the network with zero fee transactions that's why those don't 26:08 get propagated it is it those would be considered you know DOS vectors and that's why they remove 26:13 those is it time for me to invoke MEV do we want to talk about MEV okay so MEV on Bitcoin that's you 26:23 know MEV so from ETH they did they basically there's this huge business called flashbots that 26:31 they actually integrated into ETH to help combat minor extractable value basically the idea is like 26:37 in DeFi stuff you can add or remove liquidity let a trade occur absorb the ARB opportunity and then 26:44 add liquidity back there's a bunch of different attacks and I don't know attack is the right term 26:48 a bunch of different techniques that you can do generally it's up to the mining pools to do this 26:53 because they're the ones that actually get to order the transactions now for BRC 20 and all 26:58 these inscriptions coming mining pools are going to be incentivized to place those transactions in 27:03 a particular order that could be considered a type of censorship I'm not sure if it's censorship in 27:09 the classic sense of like binary but you are effectively modifying the consensus without 27:15 outside of the standard rules of the fee you know the fee rate so that's a very interesting 27:21 potential vector for for mining pools to become vastly more dynamic than they are today where 27:31 generally a mining pool is plus or minus a few bips within all the other mining pools 27:36 are deficiency and things like that but if MEV becomes a real thing that's a very meaningful 27:40 like hundreds of millions of dollars of revenue have been generated by MEV and ETH and if that 27:45 comes to Bitcoin the mining pool game is going to change very significantly so you know I think it 27:54 was around when I think might have been before SegWit was introduced and adopt widely adopted 28:01 there were consistently high fees and it seemed like it was coming from a mining pool who was 28:09 just putting extra fees to drive up the fee market would you consider that like a form of MEV? 28:17 I would I would say no I wouldn't say that's MEV. MEV is generally more transaction ordering I think 28:27 with MEV it's kind of like you're doing something other than including the transaction so there's 28:31 like there's some other thing going on and you the sum of all that equals more money for the 28:37 miners I think it kind of cuts into a little bit about how easy it is to run a mining pool 28:42 whereas how difficult is it to like check the transactions to make sure they're valid and then 28:48 order them or something like that and then whereas with Ethereum they are doing many more steps and 28:55 this I think makes I think the reason why it makes people uncomfortable sometimes is that it does cut 29:00 into what I was saying about how it's you would want it to be very easy to run a mining pool and 29:06 even to audit a mining pool to make sure that it's doing what we want it to do whereas with 29:10 MEV it does seem to be it seems that there's like a victim with MEV you know there's like 29:15 traders are trading and some they would have gotten a better deal but instead the miners get 29:20 a better deal so it's a suddenly a very adversarial relationship between a miner and the end user but 29:26 I do think that you know miners will revenue maximize which is the bottom line as always. 29:34 It creates a very I think it makes mining into a very different type of business because either 29:41 it might not be the miners themselves it might not even be the mining pools themselves that 29:46 have to do this because maybe they outsource it but at some point someone has to be very 29:49 intimately familiar with the way that users are using their network what their motivations are 29:55 what are they trying to achieve and the miner kind of tries to extract value from that which 30:01 has not been really happening in Bitcoin in the last few years maybe for the first time again sort 30:05 of at some very specific points in the last three months and again yeah it's specifically about it 30:13 usually comes down to ordering transactions right because it's not really about letting a 30:17 transaction in or not letting it in you will eventually probably let it in happily but when 30:22 there's like this added semantic to the point in time that the transaction appears relative to 30:30 others then because miners are the ones that decide that then they might find that there's a 30:35 you know that there's a way to extract value from that and just you know like as an idea as an 30:41 example right if there's been this little race that ordinals people have been doing to get into 30:51 the first you know 10,000 inscriptions for example right now if you because they have this number and 30:57 some people had this added you know they value having a low number of an inscription so if you're 31:03 a mining pool and you see 20,000 you know inscriptions coming in and you need to pick the 31:09 first 10,000 then of course you know that might be very valuable for the for the people are creating 31:14 those transactions it is nothing to it's very different from what usual Bitcoin transactions 31:19 are where people do not necessarily care about urgency and or you take a little bribe on the 31:24 side to get in the first 10,000 yeah listen I pay you a little extra yeah I'm gonna be there yeah 31:30 so that the issue I mean I don't know if it's an issue but this this happens maybe happens when 31:35 people ascribe meaning and semantics to the order of transactions and not just the existence of them 31:41 yeah certainly I mean if you saw this burst of 10,000 transactions let's say we're minting a 31:47 PRC 20 which what are you referencing here maybe the mining pool just swaps out 5,000 of those for 31:53 their own because they know like wow there's a bunch of interest in value here that's it I would 31:58 say that's MEV and potentially a censorship vector MEV is very centralizing that's a big problem you 32:05 we talked about you know we're talking about mining centralization we can get into this you 32:09 know we can get into the semantics or the minutia of what centralization is in my opinion it's one 32:15 entity producing more blocks than the other that's kind of the way I would define it and in MEV is 32:20 very centralizing the reason being is that it's highly it's highly technical and very difficult 32:26 to do that's why flashbots was built flashbots was built to basically make it easy for everybody to 32:31 do MEV but before that there was a huge centralizing force towards pools that were doing MEV because 32:37 they were massively more profitable like to the tune of 20% when we negotiate fees with miners 32:41 you're talking about like single bits you know the difference between 1.01 and 1.05 percent fee if 32:49 you're talking about adding 20% more volume revenue that's you know 2,000 bips that is an 32:54 incredible amount of revenue that could be added to pools which will certainly perturb the end and 33:01 in cause centralization towards a pool that's able to do MEV effectively and I'll just add that when 33:08 in this discussion about MEV there's kind of three groups that I see that are kind of playing that 33:13 game there's end users there's application developers and then there's mining pools for 33:18 end-users MEVs obviously end-users would prefer that MEV doesn't exist because MEV basically means 33:24 that miners extract value that could otherwise have gone to the end-user so end-users would 33:29 prefer MEV not to exist but there's not anything they can do immediately other than trying to pick 33:34 applications that do not have excessive MEV so and that brings us to application developers they can 33:43 try to design their applications and their you know protocols and you know BRC 20 for example 33:47 is very MEV heavy but I wouldn't be surprised if BRC 20 will be you know over time adapted or 33:53 replaced by protocol that is not as MEV heavy so application developers can help solve it in that 34:01 way mining pools however will probably have to be on top of the MEV game you know from now on 34:09 forever because even though I would expect application developers and users to push to a MEV 34:16 light world there are always going to be these you know these temporary period of times like we 34:24 have had in the last month where you see fees going up very high because basically of these 34:29 MEV phenomenon phenomena so mining pools will have to be ready for it it is gonna happen even if it's 34:36 not gonna stay you know gonna happen all the time it will happen sometimes and this is what we've 34:41 seen in aetherium too it's not like you know flash bots is making buckets of money necessarily 34:47 right now but if you look at a year ago when new protocols popped up on aetherium like every week 34:52 and there's been like this entire defy conundrum and NFT and and and whatever then during those 34:58 periods of heat MEV revenue goes up a lot so I think pools will have to kind of adapt to it 35:05 this is all very healthy thing though I think that you think like because well for those of 35:10 you don't know something weird happened at once at scaling three where a miner left scaling three 35:15 early to signal for a bit kind of limited with nine percent of their hash rate and this was led 35:20 to a kind of blockade of bit nine activation of SegWit and then it started this whole like miners 35:26 versus developers like access or something but really the miners are just when the miners revenue 35:32 maximize they are just trying to make it so that you know the customer is actually whoever's paying 35:39 out there this money you're trying to maximize like the end users use of the the protocol they 35:45 want the miners want the price of Bitcoin to be as high as possible miners want the transaction 35:50 fees to be as high as possible so this has always been a kind of very healthy and very rational 35:55 feedback loop that we sort of like we kind of like took a break from for like five years and 36:02 now it's like returned so I don't know like that's just kind of like you know what do you think is 36:08 like the next what's the next for what's where's organo's going to go we're going to go like how 36:13 high are the feet see because the thing is the fees are very low that's part of the reason why 36:16 ordinals could break in is it didn't really cost that much to actually fill up the block which in 36:21 a way implies that the people were not people were not actually taking making use of the blocks right 36:27 you know I would say the reason that inscription got popular and we're seen as so appealing to 36:33 such a you know broad audience was because block space was underpriced for for a while 36:39 artificially it's very cheap and then you know we see the fee rates fees per block like total 36:47 US dollar fees per block that will go up sometimes like during these bubble periods but on the whole 36:53 it then returns to a very very low rate only with the ordinals has it been anything above that right 37:01 it may it may return back to its lower rate again or I may not but it's kind of interesting that the 37:06 one time the fees have stayed slightly high has been because we've done a kind of kind of like 37:15 non orthodontic non laser I type of yep do you think many pools will clear again it's a great 37:24 question I would say yes I think I think the default answer needs to be yes but we're very 37:29 far away from that there's like two days worth of transactions over 20 sats so and they keep 37:35 coming by the day so the reason so early on inscriptions were driving fee volume a bit I 37:41 mean blocks were all full but people were like oh wait why is it still only five sats per V bite 37:46 that's only a little bit high the reason is because art inscriptions those type of inscriptions are 37:51 what we would consider to be low time preference they don't really care when it gets in just as 37:55 long as it does what drove the big volume last week like I don't know if you saw but like course 38:00 I printed 64 Bitcoin in a day that's insane they usually do like 30 something the reason that there 38:07 was so much fee volume during that time is because there was high time preference transactions coming 38:12 into the mempool in the form of BRC 20s basically I need to mint faster than everybody else so I'm 38:17 gonna set a thousand sats per V bite and it was really important to get those transactions in 38:22 the top and basically get them included very quickly and that was that is what causes fees 38:27 to go up the last time we saw a fee spike like that was during 2017 when everybody was trading 38:33 shit coins around different exchanges and trying to get you know Bitcoin from Cryptopia to Bittrex 38:39 and whatever so that he could buy Pepe whatever that was really and that was again another type 38:45 of high time preference activity and that's what you need to see we need consistent high time 38:50 preference activity on Bitcoin to continue driving fees and continue driving transaction 38:56 fees up for for mining pools yeah but of course that's kind of a paradox because in the long run 39:02 people are patient right so it's kind of like you need people to be panicked on a continuous basis 39:07 I think and I think that's very good but I when I was talking about organo's I kind of just mean 39:11 this more abstract like I kind of include BRC as ordinals is like ordinals is like the name of the 39:17 technical 20s built on top of one and also yes so but they're like there are ostensibly not a 39:29 Bitcoin transaction so like stamps would also be it's part of the ordinals movement the ordinals 39:36 time I think we if the if the men pool does clear I think that's would spell the doom of the there 39:43 was always a long you know this idea that over time transaction fees would replace the block 39:48 subsidy so if the men pool will always eventually clear and that does seem to imply that the fee 39:55 rates will never rise above a very small amount there always be a time when it's only $1 per 40:01 transaction for like you know 2,500 transactions or whatever you can fit into a block so but that's 40:07 using dollars as a unit of account yeah but I think we so it is should be like the the total 40:13 amount of money paid by users you can use in whatever you like it's like it's the purchasing 40:17 power of that unit that makes the difference so it's what did the users sacrifice an opportunity 40:22 cost like how many how many hamburgers do they give up to broadcast a Bitcoin transaction you 40:28 know but that's the however many hamburgers they give up is how many hamburgers the miners get to 40:33 eat so it's the same number so that value in terms of its abstract purchasing power that value 40:39 well you know though users want that to be low and they'll wait they get annoyed when the fees 40:43 are high but the miners obviously want that value to be high but it is the same value and the blocks 40:46 when the block subsidy goes away that is the same number but if if we're but if we believe in hyper 40:53 Bitcoin ization which is a default assumption then Bitcoin will theoretically be worth a lot 40:59 lot more so if the block subsidy goes away but the fees are still at 22 cents per V by that might 41:06 actually still be a lot in terms of revenue the fee revenue is in purchasing power so it's 41:13 independent of the Bitcoin exchange rate it's 100% independent the Bitcoin exchange rate could 41:18 actually be increasing at an accelerating rate and that would make no difference when the block 41:22 subsidy is zero in the interim the block subsidy gets halved every 48 four years famously but that 41:31 would the Bitcoin price cannot actually double at a rate that would make up for that without it being 41:37 the case that I mean this is like a spot I don't know if this is possible or not but one day like 41:41 you know 99% of world GDP would just be like the value of Bitcoin I'm not sure you know maybe but 41:50 I think yeah so that's an interesting question but my guess is that we want we want there to be 41:55 we want miners to be allies of people who actually want to pay fees because they are the actual users 42:03 and we want to be as respectful as possible of the actual end-user not necessarily someone on 42:09 Twitter we want people who pay they really are they are the customer customer is always right 42:16 with that from the point of view of the of the end-user who's creating their transaction the 42:22 reason they would agree to pay you know whatever $20 per transaction is not related to how much 42:28 it is in in SATs or how much what the Bitcoin price is it's just a question of do they get 42:33 $20 of value out of creating that transaction and the the answer is we don't know and it's 42:42 very hard to predict what if someone will be able to get $20 of value out of making a transaction 42:47 tomorrow and it's definitely gonna be difficult to answer that question in you know 10 years into 42:52 the future right and that's why people now are paying a lot in fees for a BRC20 token because 43:00 they are they believe they are getting more value exactly these are currently worth so we're running 43:04 out of time so I just want to do a quick run through of like for final thoughts do you feel 43:11 like transaction censorship is a reasonable or like valid concern for Bitcoin no not at all all 43:20 right the miners can and will maximize their revenue because of Alan Turing's computational 43:26 universality among other reasons so I also say no but we should build something that checks to make 43:33 sure that every there's no transactions that would have been in a maximized block revenue 43:37 maximize block that are not that are being passed over so that we cannot be warned of that but it's 43:42 very easy to be warned of that I think yeah I mean not only should miners optimize their revenue 43:50 miners cannot not optimize their revenue and that's what makes Bitcoin work I'm not losing 43:57 any sleep over this I know that it's gonna be fine of I've seen it work for 10 years and it's 44:01 gonna continue to work thank you Miami for the last three years in this amazing city the whole 44:21 world shut down but Miami welcomed us with open arms we want to show Bitcoin to the whole world 44:28 we are taking the conference on the road to set the stage for Bitcoin in a new city 44:46 Nashville Bitcoin 2024 is coming to Nashville in Tennessee a city that is known as a music 44:56 and freedom city Bitcoin 2024 in Nashville from July 25th to 27th