SUPERHIGHWAY TRADING: The Fastest Blockchain Distribution Network | Uri Klarman

Transcript

you either make quite a lot of money because you validate and extract value or as a c5d5 Arbitrage and extract value or even a Searcher which has zero risk would kind of extract value Etc are you a sucker what's up everyone welcome back to another episode of zero X research we are joined today by URI the CEO of blocks route uh but before that I really want to give a quick shout out to our sponsor hexins they're one of the most hardcore security teams in web3 pioneering in ZK and novel cryptography they are trusted by some of the top tier Brands and cryptos such as polygon working on the ZK VM mantle risk zero Lido one-inch new bank and more um you'll hear more about them a little

later in the show but if you are coming to our event permissionless uh in Austin Texas uh September 10th through 13th be sure to stop by their booth at Booth 832 and say Xerox research sent you for a free web 2 pen test along with your quote but anyways Rudy we're really excited to have you on here thanks for joining us um do you mind starting just uh telling people what blocks route is and uh the main premise beyond the project sure well hi Dan Hi Sam great to be here I'm URI CEO and co-founder of blocks route tldr we're called blocks route because we route to blocks and transactions so the core of what we do we have infrastructure servers all around the

world which send information really really fast almost at the speed of light so we propagate transactions and blocks I think our average propagation is globally is under 140 milliseconds so we're not at the speed of light but we're pretty damn close to the speed of light and I did my PhD here at Northwestern University in computer networks with my then advisor now co-founder professor kuzmanovic and when we got started the idea was well we kind of like went down the rabbit hole of crypto coming from the research side and when we came up with the idea of blog so it was like

well if we can send information faster can we somehow help blockchains and there were some ideas around that but we really found product Market fit when the D5 summer happened in the summer of 2020 when everybody and their sister kind of like wait these guys allow us to hear about blocks faster and send our transaction to the ma back then mine equal nowaday validators send it faster to them basically allow people making better trades so we're kind of like um Flash Boys for D5 you know flash boy the guy who connected Chicago and New York like if you're one millisecond faster than everybody else and you know price skyrockets in New York then everybody rush and if you're just one

millisecond faster than everybody else to hit let's say Chicago then you buy there and you profit and you're better and so during the the D5 summer we had all these people who started to use our platform and our Network to see other blocks to see other transaction to frontrun this was prior to flashbot and because before Mev became a thing sending like we didn't call it MVP we were the first to offer let's say private transactions and the transaction directly to the validator don't get front run because nobody sees your transaction so this was like I think mid 20 mid 2020 like uh July or August or something like that and so over the

course of let's say five quarters from mid to it but basically over the course of 2021 we moved from you know being a team of 16 people we're kind of we built the infrastructure but we didn't really find the product Market fit and to wait a company that like makes 10 million ARR and like everybody are using anybody from jump and winter mute and the largest traders to you know an Mev Searcher in a basement in Taiwan nobody knows who is and everybody in between so they started using our services to see information faster to send information faster um flashbot came out we started to also run a flashbot relay so for people who

aren't very deep in the Mev landscape we'll probably dive deeper into that but um Mev the ability to see a transaction front running back Run It capture Arbitrage so when flashbot came out we also ran um a flashbot relay all the mbb serger became our customers because they all want to see transaction and they want to send it fast Etc roll forward to the merge and Mev boost um almost a year ago in in two weeks it would be two weeks two years well if I don't know when this is air but literally a year ago um then we are now a major Med block Builder and even more importantly an Mev relay so we send transact sorry blocks

from block Builders to the validators they choose the best one we broadcast it and so we are very deep into D5 we're very deep into the Mev landscape a lot of the actors that are using us I think we peaked in the bull market like everybody um I think we were routing one and a half billion dollar of Trades per day for our users and customers so that was like 40 of all D5 volume was coming from our users to be sent you know an ethereum on BSC on polygon on Solana awesome yeah that's uh that's a great uh

intro into into what blocks right I want to give a quick shout out to hexins as we explore today's blockchain landscape let's take a moment to recognize them as a premier cyber security provider in web 3. hexins is trusted by tier one projects like polygon including a Security review on their new polygon ZK evm mantle risk zero glido one inch new bank and more get a deep dive into your technology stack with the most comprehensive analysis in cyber security Consulting hexins not only uses widely known methodologies and flows but discovers and introduces new ones on a day-to-day basis with over 55 billion dollars secured they cover everything from Smart contracts to blockchain to web 2 pen tests yeah there's been nearly seven billion dollars of total value hacked in

crypto's nascent history so it's safe to say your team has a lot on the line don't skimp out take your security seriously and reach out to hexins don't forget to mention Xerox research for a free web 2 pen test with your partnership and reach out to hexins at hexins.io find them in the links in the show notes or reach out to them at permissionless they'll be at Booth 832 but without further Ado let's get back to today's episode do you mind taking a step back and just laying the foundation for the conversation and explaining you know what is the transaction life cycle uh through MAV boost today sure so I think it would be easier if somebody isn't really deep in the space let's start like from the basic you could imagine and then kind of like roll all

the way to me boost so you can imagine you know vanilla ethereum people send transaction out there and if somebody's about to buy 10 000 e for usdc that will push the price up if let's say you're uni swap then if somebody could see that transaction and get their transaction ahead of this transaction you know it's going to know the price is going to go up so they can front run this trend I can just get ahead of it and buy it low and then afterwards sell it high so that is front running back running would be um if somebody's about to make that trade and he only buys on uni swap price on uni swap will be super high but you know let's say on balancer it would still be low and the real price would be

somewhere in between those and you know actually centralized exchanges are the bigger Factor here and so the price in uni swap would actually be wrong it would be too high so back running a transaction is mostly capturing the arbitrust just behind the transaction and if you do both it's old sandwiching okay you run around the transaction buy before it then the transaction is a victim transaction happens then your background sells afterwards and you Prof as the Meb Searcher you profit so vanilla ethereum people were just trying to get ahead of other people Etc the next version was or one big negative externality of that was that gas prices went through the roof because

anybody who tried to get into the network everybody are spamming the network with super high gas transactions try to get into the block so even you know I just tried to send you a token or an nft or something like that it isn't super important to happen it doesn't have tons of in like value to the transaction but I am in the same game with like the hardest game of capturing Mev and D5 so it really pushed the gas price for everybody The Next Step was kind of let's say okay let's create bundles instead of just throwing it all into the mempool let's give it to the validator and tell them oh if you include this transaction like let's say a sandwich transaction or a Arbitrage or Etc so if you include this

bundle up let's say a sandwich F1 running transaction a victim transaction and back running transaction it would pay you let's say a dollar and somebody else does the same thing but he's slightly better or he pays more it pays a dollar and a half and somebody else is paying two dollars so the Val the back then it was the mining pools now it's a validator can like choose whoever paid them the most it's kind of like style this entire competition outside of normal usage roll forward even further to Mev boost and POS and the merge and the idea there was less than if each Valley they would do it internally or let's say the flashbot idea was that if each

individual value would do it by itself then if you're a solo validator you don't have the resources you don't have expertise you're not going to be very good at it so you'll make four and a half percent apy and let's say coinbase or Lido or attestant or any of the largest like validator or block uh demon or pigment any of the largest one they'll have an expert team in-house to extract them as well they'll be making seven percent apy and if that's the case then not only they will make more money of course they'll make more money they'll make more money because they're bigger they'll make disproportionately more money and this is a centralizing force if you have to if you have 32 East

what do you do with it do you solo validator do you run a solo validator or do you give it to Lilo to do it for you one would give you four and a half percent the other would give you seven percent of course they're going to give so of course they're going to give it to the largest actor who would be better at it so this is a centralizing force and the idea and this is I disagree with flash but about a bunch of stuff but I do agree with them that this is a bad thing and therefore um the idea was instead of each validator building their own block by themselves they open it to third parties to block Builders okay so instead of it the value of the block being dependent

on just how good you are at extracting Mev and finding Arbitrage and front running and sandwiching all these kind of stuff you just say oh anybody who wants to send me a Blog do send them my way and one would offer you a block and it pays you a hundred dollars the other pays a hundred and one this one pays you 105. you just take the best one you Outsource it and which makes it that even if you're a solo validator or your light or or whoever everybody are making the same apy and so that prevents this centralization force of validators obviously the other side of it is that that just pushes the centralization into the block building area okay so you have no like you have Beaver you have rsync

and you have Titan which are three Builders who are now responsible for something like 70 of all blocks out there which is wild and insane um and you could say it's not good is that bad but stopping here for a second this is now the landscape of the transactions um as to the transaction life cycle you could imagine or one more piece worth adding is now we have private order flow people used to send transacted to the vanpool they got exploited validator extracted the value from them so people said well we don't really I I don't like getting clubbed in the head every time I try to

use D5 also for me like it makes me terrible if I will never be able to compete with C5 that way and that's like a big part of why I the area where I disagree with flashbot and so a new form of actors started to to form from this um which is private order flow and Order flow auction and these kind of things basically I send the transaction and instead of it just going to the validator I said we had a service let's say we have a background me okay you send it to us we promise we won't rank on you okay which front running is bad I would get a word surprise as the user I will get the worst price but we will look whether you're doing a really good job with your transaction and if you're

not let's say you only buy a newly Swap and you know you should be capturing it on multiple venues and you're leaving money on the table we will try to pick that up and the user will get some of it and the value they will get some of it and we will get some of it and the surgery will get some of it so these if if we're talking about the life cycle of transactions instead of transaction going to the general member now they're growing whether these are services like background me whether directly to the Builder whether directly to a search or like older all sorts of auction mechanisms that are being built around it I don't necessarily like we're offering it and it's better than dollar alternative like you're already clubbing the head and losing all the money but

everything actually is basically like oh you give me the transaction I'll Club you in the head and I'll give you ten dollars and this guy offers to Club you in the head and give you 50 so you still get Club in the head like you're just getting me paid for it it's like by definition worse than not getting clubbed in the head um taking it money-wise like some of that value is being extracted you need to front run it or back run it or capture arbitration like there's definitely friction there the Searchers also want to make money out of it other people so by definition it's worse than like if the transaction executed without all these stuff around but it is better than oh just throw it to the memo and get run run and so stopping here for a

second if you're thinking about the life cycle of transaction you know you're sending a transaction to somebody you know send an nft to somebody or make it like it doesn't really matter there's no value to extract there it could totally go to the to the rental D5 transaction which in my opinion are the most important the most valuable and the thing I'm really bullish and excited about in crypto carry with him a lot of value so if you throw it into the mempool you're going to lose for sure um you could imagine maybe it's worth even stopping here for a second there is a price on uni swap it is worth two thousand dollars let's say now

block someone ethereum every 12 seconds so six seconds later price on centralized exchange is already changed right it's not 2000 maybe it's 2001 maybe it's uh 1900 or a thousand and 999. I it changes all the time but when you're about to make a trade and you send the transaction to trade on uni swap then you don't know what the price is going to be when the transaction execute in another let's say five seconds or 10 seconds so you kind of have to guess if you're about to buy it let's say if for usdc and the price would shoot up then your transaction would fail okay you want to buy two thousand but really two thousand one and what do you know you don't get to buy at

2001. you know so you have to kind of like if you don't want your transaction to fail you have to allow well a bit of slippage um will I want 2 000 but I'm willing to accept up to 2002 that's it just throwing a number however if the price would actually go the other way around to if it drops it's like it's at 1900 or something you're still going to pay the highest amount you were willing to pay right because you said I'm willing to pay up to 2002 and if the price went over went down by a hundred dollars or something like really like five percent crash people a Searcher would see returns I said oh this guy is willing to pay that much money sell it after him Etc so yeah so when

you're thinking about the life cycle of transactions we moved from vanilla ethereum you know just introduction to the mempool valley there see them they put them on a block largely based on when they saw them and the fee that you pay highest fee on the top into being like you send the transaction it will get front run so you send it through like a back um either an auction or background me or something like that to avoid that and just get it to the builders not even to the validators builders will try to make the best the best luck and offer it to the validators um now you have a bit of a trust problem between the proposer the validator if you're about to create an Xbox and the Builder they could each screw the other party

um if the validator if the Builder just give it to the value so like oh here's the best Vlog the value and you know maybe it pays the validator a hundred dollars and keeps to the Builder let's say ten dollars um then the value screw that takes it takes a block see what the block looks like just unbundle all the transaction extract is the most value yourself don't pay anything to the Builder and just keep all the value to yourself so like if the Builder just give the proposal the block the proposer could screw him over if it's the other way around if the bill is like oh I would pay you one each just sign this thing blindly for me that if the proposer does that the value they're signs blindly and maybe the

validator sorry the Builder doesn't actually pay it says I'll pay you a million dollars in fact he didn't pay anything dollar took all the money to himself and disappeared into the sunset and so you have this trust problem between these two entities which is why we also intrude or Flash but introduced A New Concept of relays we run a relay for example basically we're somewhat of a trusted entity that sits in between and we get the blocks from the Block Builder and we see that it's fine and we say oh here's the best quote that you could get sign it blindly and instead of trusting the Builder trust us there are seven relays out there blocked throughout ultrasound flash bot agonostic block native istas

and Eden and that's it it's not a really good design in my opinion but we're gonna sit in between to facilitate all of that so validate make more money and Builders are doing it they're not afraid of being screwed over that was beautiful completely set the scene for what the current state of the the transaction life cycle is how what Med boost is how it operates and most importantly who are the major players in there and so you kind of mentioned all of them right Searchers Builders uh relays and validators and I have a quick question on a couple of those to be honest but uh to begin so Builder Searcher integration there's been a huge rise in in this and ultimately you know you've mentioned a couple of those builders that are dominating the block building Market um my presumption is a lot of them are

integrated Builder Searchers I definitely don't know that for sure we actually have good data both like intimately we know these actors you actually know them too you just no no that you know them like all the big Builders are actually large signify D5 arbitragers you know like if oh this is this is them and this isn't like you you know these names and so it's like no to the people like who are like deep in everything like like it's not a big secret but it's somewhat secretive um so there are two things worth saying here three one Builders generally speaking are c5d by arbitragers okay so if you have a lot of

let's start from a different angle C if I defy Arbitrage is the biggest piece of Mev people say well Mev and cross chain or whatever nothing is even close to matter as much as C5 D5 Arbitrage that's the thing you know C5 moves in real time D5 is somewhat of a side guard running behind it and so the prices on D5 are stale right we just like uni swap there was a block here are the prices but 12 seconds later prices are completely different and whoever gets to all of these I would love to trade against like stale prices it's 12 second sale prices all day long you always make money you never lose money that way um

out of all of Mev something like 60 is being paid by C two validators by C5 D5 arbitragers but unlike Atomic Arbitrage where there's not a lot of risk and you know if there is an imbalance between uni Swap and balancer there's a lot of competition everybody like everybody try to make it the most efficient way and pay the most to the value to get included so validators of that extract 91 to 99 of the atomic Arbitrage and stuff that just happened on Jane is a completely different Beast first it has risk okay you're trying to capture

that Arbitrage it doesn't like either work or doesn't work you could be you could find yourself capturing it on one hand and missing the other leg of it so it has risk you also need pardon my language shitload of capital on C5 and on D5 right let's say oh I want to buy a million dollar of Eve here and still a million dollar if there so I need to have inventory I need a million dollars worth of if on finance or on coinbase or probably both and elsewhere and on the fight so you need not only to have a lot of money you need to also have positions and managers so this is really done by like large C5 defy actors so if this is not something that you could do from your basement somewhere without a lot of

capital this is like super serious big money so it has risk obviously if they don't do it and they're like oh there's no Edge they just do it for efficiency but they don't make money if they don't make money out of it they don't do it and so they don't pass 91 to 99 of the value to the validators on average I think the mean or the median one of them is like 65 of the value go to the validators the range is between like if I remember the numbers from Frontier correctly between 37 and 77 so it's not oh let's extract all the value and we'll try to bid it to the validator and try to get it included it's a lot more about the different

positions of the different act c5d5 arbitragers if I happen to have a lot of eth on chain and a lot of usdc on coinbase at a given moment and I want to make this trade and a different C5 arbitrator sees the same opportunity but he doesn't have the same amount of money in the same position or maybe he has the money but you know they have like super complicated like Risk Management engines internally they can't make that trade because that exposed them to if going down it'll it's funny like it's bad even if if goes up okay it's kind of like if you're exposed to something it's bad even if you bought something and it goes up it's a life of like a strat art and

so the winner would be the one that happened to be the most positioned it's not super competitive as you might think so only 55 or so of the value being paid to the validators now take going back to the builders so block builders get all the bundles from Searchers who capture RBC um Atomic Arbitrage Etc but it is the Builder the c5d5 Arbitrage who wait until the last possible moment if we say that the value they would ask for the block right let's propose the best block It's Time Zero they'll wait until the last 100 to 200 milliseconds or so try to make the best bid because only then do the let's say differently they

try to minimize the risk if they've said like oh here's the block I will make it this way five second in advance they can't really make a competing a competitive bid because maybe price will change so they like wait until the last possible moment in order to make the thing that makes most most sense for themselves they kind of know the price on C5 they kind of know the price on D5 and try to make the best bet and so Builders block Builders are the c5d5 arbitragers like like Point Blank and so this is a somewhat centralized nut it is a very centralized um um landscape there there's a question

does it matter or doesn't it matter um I remember speaking with vitalik on this when he did like an Mev Salon in Vienna his take was okay we're going to have economies of scale we're going it's going to happen anyway let's break the connection between validators and this right what I said earlier we don't want this so larger validators make higher apy we kind of like there's still economies of scale and the larger actors will be better at it it's just that it happens only at the Block building space and whether that's okay or not okay is for anybody to decide is that into the question or did I just move to a completely different trajectory of rambling no no no no you I thought you nailed it and one of the

things you mentioned is actually the you know you can see in real time like when those bids come in and it's very close if not directly after that 12 second Mark uh where you see the largest in winning bid come through and there's a great website I want to shout out called payload.de it does a phenomenal job it's incredible it's absolutely incredible we'll put this in the link in the descriptions if you ever need to procrastinate time if you happen I mean if you're listening to this podcast you're already Deep In The Weeds so you're probably like us and you're one of us so check this website out 10 out of 10 recommend you already opened friendtech you already responded to anybody who responded to you Etc and you have like three minutes there are stuff you need to do but you don't want to do

it so it's kind of okay let me burn a bit more time doing something else go to payload.de and yes it's like their animation of like building the block and how it Progressive over time is super cool it's super super cool so I want to talk so we talked about the integration between Searcher Builders and how that's beneficial to that operation but if you look up two steps further down the supply chain you also see the integration with Builder relays so a couple questions about relays but first and foremost like to me it feels like relays are effectively subsidizing at the med boost system because there's no economic value or incentive to doing so into running one of these and you as you mentioned there's seven or so relays I think uh I wrote this down so six relays that have

10 more of the block share that actually blocks getting routed through them and you guys run two of them so there's really only five entities that have at least a 10 share of of the blocks getting relayed and so you could you could put those people at a dinner table so ultimately what do you mean you could we have like you go to Paris you literally meet everybody um so first of all I want to quote Matt Cutler the CEO of block native block native are our competitors but I also I like them a lot and we try to collaborate as much as we can and it's like in the Mev landscape you're either I don't remember you're either the victim or you're making your money or you're a sucker or no no he had he

emailed it better but basically you either make quite a lot of money because you value their extract value or as a C5 Arbitrage an extract value or even a Searcher which has zero risk but can like extract value Etc are you a sucker you're the user values guy you're the uni swap LP money is attracted from you or you're the relay which as he said it's great we're a a small group of like a few entities and we're subsidizing the largest validators in ethereum who are who have 25 billion dollars of each stake and they make like a let's say five percent out of that so that's what like one in aggregate that would make like 1.2 billion dollar per per year and okay we got them we and

block native and ultrasound Etc so there is no incentive for to run a relay we don't make any money out of it a lot of people are like confused then why do you do it and for us like okay if there was the merge it was either flash but just running a relay and their oh [ __ ] transaction like like censoring that's not a really good thing and so we said let's chip in let's try to help this we were subsidizing yourself but we're now trying to actually find a good funding or a good um sustainable model for this I'm actually in the opinion that we could be charged for this service than anybody who wants to use a display and choose which we which relay do you want to use but that's a side note but to your point

about into vertical integration there is no incentive to run a relay if you're a builder really like we run a builder but I really is like our Builder makes I think it's like two percent off the block space we could have made it a lot better if we wanted but we don't want to we're we don't want to compete with our customers like everybody like all the other builders are using our services Etc we mostly run it to provide additional services for some of our customers and also to be you know Hands-On and know exactly what we're doing so we we understand what our customers are doing but so we're running it like flash what are also doing it block native are also

doing it there's no like we're not C by defy arbitragers we don't make stupidly large amounts of money especially when the market is volatile and keep half to ourselves it's a thing that we also run you could imagine that if any of us who were running a builder were not neutral enough okay so it's kind of good yeah I run it that but I don't give my own builder any advantage if my Builder if there is a vertic integration between the Builder and the relay then the relay isn't really this neutral entity that sits between the Builder and the validator right it is just the Builder giving to the validate so as I said earlier there is a trust

problem each one of them could majorly screw the other side um and people say well the Builder can just trust the validator right and the answer is no a validator no I solo validate I wait until it's my turn I get the block and I take all the bundles all the sandwiches where somebody front run and then background Etc again bundle it I just take the front running from each of them and then Crush whoever did so it's as if you're trying to capture an Arbitrage but you only do one leg and you don't do don't do the other one um it is very similar in Dynamics to what I don't know if you no sandwich the reaper later renamed to to um low carb the Crusader but like he

asked his name to be changed in in exchange for him revealing another Anvil security vulnerability in the Mev boost like setup long story short he found a way that the validator could screw up basically sign something and then like misbehave or something you could he extracted 20 million dollars disappeared into the sunset okay so like any solo validator who just gets the block from the black Builder could do something like that it really really protects both the Builder and the validator from one another so if there was integration between a relay and a builder that protection is gone it's kind of like well you're really just trusting the Builder so that is less of a real vertical integration

pressure I would argue people just won't use it like like any of the major values um values say like yeah we don't trust that relay we'll use all the other relays but not that one because it is strongly integrated with the builder so again we do run a builder but it's not really a lot we almost do it as a service for others Etc we had the opportunity to make it a bigger better Builder we're not in that business that is true for us in Block native and flashbot some um agnostic and ultrasound and I so don't even run a builder so they're like they're they're safe from that direction so it sounds like you kind of think ultimately right now in today's ethereum D5 users are ultimately suckers so I guess like

what do you think the ideal end state of this looks like and like what's the best way to internalize this Mev that doesn't just incentivize sucking value out of users transactions I probably have this I was about to say I have the strongest unpopular opinions but I think Vlad zamfer beats me with his idea of time traveling trades which I only have understood when he explained it to me speaking with Vlad zombie is always a bit like that um I have a very different take on what how the Mev ecosystem is affecting defy right now how bad it is and what we should be doing so I think the key Insight that I have that most

other people don't share yet is that Mev is a negative sum game okay not even a zero something a negative sum game the more value validators extract from D5 whether that's lpes or users front running whatever if they extract more value then D5 is less useful right by definition that value is not created from they're not creating value by offering services or whatever no no they're extracting value clearly any of the um um minor extractable value and so they extract value they make D5 worse if D Phi is worse and not as useful then we're going to have less liquidity and

less activity in D5 because if I have a thousand dollars and I want to buy if where do gonna go buy it on uni swap or on coinbase in coinbase I will get the current correct price I don't even know what's the current correct price right I am not a high frequency Trader with seven monitors in front of me trying to like okay here's the liquidity on each one of them the price right now should be this and put the limiter yeah I just like try to buy it at the current correct price but because we have these high frequency Traders on C5 who constantly are of any opportunity there then any venue you go wherever you buy you're going to get the current correct price if you go to D5 you will get the worst possible price right you'll try to

buy it at some price and if price goes up it would fail and if it goes down you'll still pay the higher price like you're not going to get the best price by definition and so D5 won't be able to compete with C5 despite all the good stuff that Diva has so I'm a giant D5 bull okay D5 has transparency it's like this single always on open in the context of how like the internet is open anybody could participate single like Global Financial system that we're building here its value is immense I could spend some time like jamming on that if you guys want but like its value is immense but it's not going to compete with C5 and it's going to stay this like sidecar that runs behind it if it's going to be

less useful and so going back to by the negative some game validators extract the most moves that they can so the D5 sucks so there's less liquidity and there's less activity so there's less Mev and so D5 stays like a niche if you have to you want to trade Pepe you can trade it anywhere else okay you use D5 but anything that you could you use C5 and my favorite example for that is actually Bitcoin trading so BDC the highest largest market cap in the world not only that it's C5 volume is twice as big as if I mean huge e-head and this is like I am very supportive of you but fact is BTC trading volume is much higher whether you like BTC or

don't like BTC and the question is do people take their BTC to rap to BTC or something like that and try to trade it on D5 and it's not about the oh well you know Bitcoin they don't trade the whole no they're trading they're trading it on C5 the question do they also bring the liquidity and activity to D5 and we're seeing that 99.5 of all BTC trading volume happens on C5 and just like half a percent or so happens on D5 and the reason is that D5 isn't that useful if you're trading you're better off trading it on centralized exchange then decentralize the chain and default right now and this isn't inherent but that's the current situation so defy isn't

winning right now from where I'm standing if we were with like it could be 10 times bigger or a hundred times bigger or a thousand times bigger okay I want to divide to compete with C5 and it's not happening right now to a large extent because D5 isn't useful enough and so my own my chip on the shoulder is that D5 needs to be useful rewind understand right we want everything all right but if it's not useful we got nothing okay buildings the most decentralized system which isn't useful and nobody wants to use because they're better off using C5 unless they actually actually really have to use D5 isn't a good outcome and so with this Insight in mind of like Mev

being a negative sum game I actually hold the following opinion which is somewhat contrary contrarian everybody in the current Mev landscape are aiming at Mev auctions okay right the valley it's the validator term he would make try to extract the most value as it can from the Block we have third-party block Builders trying to feed them that Etc and the result of that is that those validators extract the most part the most valuable that they can they're making their slice as big as they can but it's the biggest slice from the smallest pie in the world okay like this

isn't a very useful D5 ecosystem and so two point first is to say the fact that we have that well we have mbv Boost so the Mami doesn't go to jump or winter youth or any other it goes through the validators means squat to the default users and LPS right it's kind of like that oh I bleed money but it doesn't go to Winter view it goes to the valley I feel so much better right like this really solved a problem for me and we're going to have Mev burn right so he doesn't even go to them it burns and value a cruise in theory at least to all ethereum holders again as a D5 user this doesn't help me one bit I bleed money and now I bleed it to all ethereum

holders that that doesn't solve anything and so that's one point and the second was um that instead of validators trying to maximize this their slice of the pie but in the process making the pie that much smaller they should be doing the other thing they should be thinking of themselves as service providers and instead of an Mev auction ensures that they get the local maximum right like they get like one somebody bids one is somebody 1.1 somebody 1.2 is there so they they are in a maximum point it does extract the most value but it's a local maximum and

will actually make a lot more money if they don't extract 90 of the value or 60 if they extract 40 percent but of a much larger pie they're better off especially if that pie is 10 times bigger or 100 times bigger and that larger pie in my opinion is if we can make D5 compete with C5 specifically and more important than anything else you could imagine that if validators give what I call instant inclusion so think of it like as a pre-confirmation it's I'm a validator it's my turn and now I like okay this is one this is two this is three this is four this is five order it in real time and get paid for it okay capture half a basis point or a basis point of all

volume then that isn't theoretically optimal it won't maximize the size so why capture just one point you could capture 1.5 like it would capture a smaller slice but it would make D5 useful you could imagine again I take the PTC example you could imagine if we just get BTC volume to actually move into D5 because it makes sense to help BTC you don't just bleed money you actually make a higher API it makes sense to trade there and not everywhere else then the volume on D5 would roughly double maybe more because again BTC volume is twice as much as if trading on on on C5 and what if D5 isn't five percent of all

trading volume but if it's 50 and what is not 50 what is not half of the trading world what it if it dwarfs C5 what if it goes to being like 500 and which is the future I'm aiming for not even talking okay future New York Exchange runs on ethereum base what not Etc okay nothing's there but like we're talking about very realistically being capable of allowing this to happen now a million technical challenges to solve there how do you prevent front running how do you propagate transactions so nobody could run from one another and there are ways to to resolve but what is fair and received order by all the validators again tons of stuff I'm not saying well I have a solution from start

to finish but I have ideas about how to solve anything it would probably evolve over the time the major question in my opinion is does it pay validators more money if it pays validated more money to offer services instead of extracting value from their users if they make more money that way they'll do it okay like they would do a lot validators are great actors they don't like extracting value but if they don't somebody else would and then everybody would move to stake with them at least in theory so they're kind of like forced to do it but if we offer them a path to make more money in a different way they would love that and so my own personal Crusade is to prove that he pays more money that way if we

can prove that providing instant real-time pre-confirmation Fair ordering things if they validators could offer that and it pays better we can spend the next five years okay how do we build it in the most decentralized way without Builders and relays and MV sergers or whatnot Etc a lot closer to vanilla ethereum okay that's super interesting so basically the idea of the theory around instant inclusion um would kind of be you mentioned mentioned this earlier but you described D5 as like this side car that every 12 seconds a new block gets uh created and then it catches up to C5 so whatever the you know right now whatever the price on binance is that's that's the new price

on on all these dexes uh due to that sexy decks are driven by the builders and now it's caught up but oh what one seconds passed two seconds pass all right now we're now we're lagging again so with instant inclusion the validators would be providing some sort of service where they could say okay I'm gonna give you this price right now even though the block isn't complete but once the block comes then you'll get in is that kind of the core idea here yes let's think of a unisop LP for a second okay and unisotope put this money in there and then every 12 seconds make a really terrible trade right it's kind of easy is at 2000 uscc 12 second pass and then price moved to let's say 19 or 1950 okay and then the C5 arbitrager makes a

single trade capture all Delta and again passes a lot of that value or half of that value to the value doesn't really matter but the lp made one terrible trade every 12 second and then you know there's the non-toxic order flow of like okay people actually trading against it but that matters less um if the effect moved in real time then the volume during these 12 seconds is something like 10 times higher than between starts to finish right price goes up let's go with that price goes up price goes down price goes up but let for 12 seconds which is an eternity in trading um um terms and you could think LPS will be trading on the way up trading on the way down so

they'll be making a lot more trade smaller trades half of it would be good half of them would be bad on average right they'll train them both ways instead of one single bad trade we'll make a lot more trade capturing more fees in the process even if their fees are let's say like um three basis points or instead of five okay so you can actually make it cheaper for the user to participate with it but they'll capture a lot more of the value if we find a way to make D5 operate in real time so from the uni swap LP perspective as an example trading in real time means instead of one bad trade where he's always being taken advantage he's really letting Arbitrage Arbitrage the way up Arbitrage on the way down

Etc and in retrospect half of the trains will be making were actually being good trades and half would be better they'll be making more on the fees so that in that I think is the spiritual or so that's half of it if you're just you know know you have a thousand dollars and you want to buy if you want to buy and you don't know what the price is but you know that if you send it to the instant included if the price was off some arbitrager would already come and correct it and capture the Arbitrage so arbitrars actually work for you they provide a really valuable service so when you send the trade you get the car in correct price you don't care if it's on uni swap or balancer somewhere else you don't care what's happening on coinbase or buying like you would always get the credit

right because they would move in picoseconds to try to correct it and they won't front run you which is like the third piece people can like hate high frequency Traders and arbitragers you know we don't want to make Trad fire again we don't like the latency War there the bad thing about Trad fi is front running what's bad about the you know the top high frequency trading firms kind of like operating and they're faster than you the fact that you're trying to make a trade they see what you're about to do and said oh he's going to buy eth I will buy Eve before him so the price so this is the bad thing now we moved like okay we take this into MEP and we say oh no no latency is bad we don't want to you know try to find latency games oh but

front running is fine like let's normalize that no that's the single most terrible thing that's the thing why we don't like high frequency Traders that's the bad thing about Trad fight we put we said that's okay but the latency where you know arbitrages are actually a good thing a valuable thing they provide a valuable service in correcting the price so you don't care how much liquidity there is here or there any venue you go around the world or I don't know wherever you are okay X you'll always roughly get the same price because if there is a mismatch somebody would correct it same for D5 if the prices are correct if prices are corrected in real time that when you

send a trade nobody sees it it gets first come first serve nobody front run you so you don't get if I'm making a trade and jump is making a trade and they're faster than me then their trade got in front of me that's totally fair that's okay I'm not trying to beat the jump and identify a pic or second opportunity but if I send a trade and they see my trade and then regular trade they make their trade based on the fact that I'm about to make a trade that's not okay at all that means I got the worst price because they thought that's the thing we want to sell for and maybe the third and for me the most exciting thing is that in if D5 moves in real time then it's not C5 is the real thing and defy every

12 second kind of catches up with it instead you're going to have D5 affecting C5 and vice versa the way coinbase affect binance and they affect okay like it's going to be a bi-directional effect okay somebody buys on uni swap that changes the price on coinbase and they're going to affect one another in real time as equal not D5 being like a side card running behind C5 and never actually being able to compete and offer an alternative there's just one question in my head that I can't get over and maybe there's a simple answer to but what's the point of solving this at the base layer of the ultimate roadmap of ethereum is moving this activity to l2s that is a fantastic

question and I I've been thinking about it a lot thank you lately there's this notion that centralized sequencer are kind of okay you know if we were to talk a year a year and a half ago like everyone no no like we have an L2 and it needs to be this decentralization this is what it's all about they're not gonna well it's hard so let's start with this centralized sorry centralized sequencer and they're like well okay it's not working how about we just keep it at that and you know we have the if one the L1 security what not so they can't cheat they can't steal your funds if this if people is to become a thing and it's major or why not and everybody

are using it and then Ukraine situation flares up and there's giant tension between the US and Russia and the USA you know what a all that here's a list of any individual Russian person we could identify touching any of them is illegal it's overking compliant you can't do it coinbase where is your sequencer if you touch any of it like you're out like you should not do that same goes for like it's not specifically for bass or anything like this is true for any single sequencer centralized sequencer so yeah they can steal your funds but we just threw them out like okay let's exclude all these people from participating in this L2 because like there's a single entity that can enforce

all sorts of stuff or saying like oh there's pepper coin that's a security you like don't touch it like don't include any transactions it does if you do you go to jail so I love Motors I think they should be decentralized I think we could achieve a lot with decentralized l2s but the current landscape is that they're not and so a lot of people want to I don't blame people who said like no ethereum is decentralized and I trust it and I have like a lot more trust in it this is why most liquidity and most activity still happens on L1 and not L2 and when l2s are mature then this should happen there maybe so like maybe it should happen on L1 maybe it should

happen to L2 who am I to say but I would also argue you know that's the user choice if I am a user and I trust ethereum decentralized Network and I do not trust definitely not a centralized sequencer he said well arbitrum if they decentralized what would that look like well I don't know I need to see the details but to then like it could happen there rather than anybody anywhere else but if you take an L2 and decentralize it you had exactly the same you just replicated L1 so everything that I just said true okay let's not do it on L1 let's do it on an L2 it's the same thing so the L2 should do first come first serve itself few questions to be asked is I don't know if I'm a validator and this paints giant loads of money do I want

to push it away to an L2 because if this is the most valuable use case of the blockchain should everybody else move to an L2 because this is the real thing and blockchains inherently always only do the most valuable thing and push less valuable use cases out two l2s are not to happen at all should this happen on an L2 there's a great reason for this to happen on L2 should this happen on email one also like who am I to say if this should happen here or not who am I to tell the user oh I know you want to try to to trust ethereum you don't trust bass or op or or arbitrum or any of the others but no no

take your trading and put it to there and I don't know too aggressive like it could happen on L2 if we do if all the trading happens on nl2 we need to have exactly the same solution for exactly the same problem there so it changes little from from my perspective it's about solving what do we do when we have a decentralized network of validators who produce blocks together it's almost like if you build the working Solution on L1 maybe you can map that to L2 in some way and I have a similar question because we had a conversation with Eugene Chen from Ellipsis Labs building an order museums what he's the man I mean incredible I like him a lot I totally agree but he actually so your first point in one or two questions ago was okay well if we

can do instant inclusion then we get more trades so LPS earn fees on the way up he's on the way down and effectively um you're you're trading both directions instead of that one bad trade and that makes a ton of sense because then you're earning fees wow uh before that one bad trade occurs in in the 12 second block time so his pitch on why I was Solana over ethereum was okay well with 400 millisecond block times versus 12 second block times you get that inherently with uh with Solana and so you don't have to kind of create this solution to go around it so what do you think about like lower block times and how they impact this lower block times reduce Mev for sure from the get-go you could imagine that like if you're thinking about like LLP perspective or whatnot

you trade up your trainer instead of one bad trade every 12 seconds you make all sorts of trade and then like get smarter and smaller the smaller blocks are um I like Solana I like Solana a lot and I like the ecosystem there and we have Solutions there and I like Eugene and what they're building Ellipsis labs and and Phoenix and a giant shout out to what they're building there it's hard to it's a different architecture to begin with so it's harder to get traction so you know it's this 10x thing but like it's is it so much better that it will be able to get the liquidity and activity and everything around that and not only that the tooling the ease of a build like evm is so broad right now that it's very easy to get something up and going and it isn't

that sorry as easy to do on Solana right now with the way that they do accounted like like just the learning curve around it so that's one thing um second thing um in Solana I hope it's this I always had this thing with GPU Labs so G2 are building like an Med solution I love them I think they're great okay like Buffalo and and the rest of the team like there's they're good they're super smart they're super technical and I hate their product okay I hate it I absolutely hate it because the entire value proposition of Solana is just like a real-time trading what not Etc if you take all the important transaction and you stole them for 1.2 seconds or something like like any not

important transaction goes through immediately but anything important let's put it on hold and try to run an auction for people to extract value from it you're hurting the value proposition of Solana itself so if that is the way that being going to happen on Solana I think it's hurting that and I told that to the G2 and again I think they're great I think they're awesome they could disagree with me we could have a disagreement and I'm sure they're going to have like their own like argument in in their favor but I think it's hurting and I told that to totally uh to Anatoly as well like I think it's hurting the valuable position of Solan is like the thing that works in real time and allow for bandwidth Etc um now take ma it's called my solution instant inclusion which you could always

almost think as virtualized trades right it's kind of like the block is going to come in 12 seconds but instead of waiting until then we're doing okay this is going to be one this is going to be two this is going to be three and then block is just like checkpoints through these pre-confirmations so it doesn't really matter when they come in it would happen in 12 seconds as long as I know that the order is already decided upon then nothing else really matters um these are very similar Visions now the question is which would get traction and liquidity and ability and compressibility with everything else faster I honestly wish Solana would do well here okay like I'm actually I think highly of the Solana ecosystem

and the founders and I think the tech is great and like I hope they're doing better I think what they're building is awesome would that be what I'm currently imagine you could think of my what I'm trying to Envision here instead if this blows up and this becomes like compete with C5 and 10 000 transactions well at that point we need Salama again right we can't like what's the battery how many transactions could be pre-confirmed and at what rate and whatnot Etc so even if I'm correct with instant inclusion they could come and say okay ethereum will take you here and Solana will take you here right like like you still need Solana if you want this to be like the the New York Stock Exchange like for real so that is how I would like it's not I think me and Eugene agree on

99 of things so like we have very similar views I think of this um it just I am trying to take the ethereum ecosystem and push it in this direction which is I think it better it's better and there's somewhat of like lack of self-awareness and like self-importation like yes I am worry and I think everybody else in the right including some videos I think I love italic and I think it's great but I think he's wrong on this and I love Phil Diane and a lot of the people on the flashbot team and Etc but I think they're wrong like um um um the go step um

um blanking thank you on his name um Uh crap so if you hear from them that uh uh there is no good and bad MVP Mev is just like F4 so like I disagree front running is bad and Arbitrage is good and whatever so we have this agreement and yes I am like icon it's like I think you're all wrong and let me try to prove and show you that and you know I might be proven wrong but I'm going to try it one way or the other because I think it it needs exploring and it's need proving and say like oh no well we have Mev and we didn't come up with any fancy way of solving it so this is the direction that

we're going and it makes Diva useless but like like let's do it anyway because we have a good understanding of the auction mechanism and how it works and we're going to have an alternative L1 right okay we have sway so users don't make transactions they have meta transaction and they go to an auction where Searchers try to create bundles and give them on an alternative L1 where block Builders participate on multiple say what users make transaction they come first like the complexity around that and the stuff that are in Soviet like I'm not sold on it okay this it's I don't want to be offended like this

gives me strong like lightning lightning Network Vibes we're going to have this I'm going to have like um um watchtowers and something this behaved we're going to see it's gonna it takes Simplicity has power now that doesn't mean that complex solution don't work or are invaluable or they might turn out to be correct and I'm wrong like it is a possibility or whatnot Etc but understanding the complexities around it and what it would look like and how many new actors are we inserting and we are relays right we talked about we don't have an incentive my real incentive is to rug everybody and just you know Ram 20 million and run into the sunset and I'm not doing it because I have bigger vision and more money which

I want to do in different ways but I am running a relay and you guys are all trusting me to do it there is no coherent like incentive plan around it and so I think I'll stop you yeah that's such an attractive solution it's it's super interesting but I do have one more question maybe as a wrap up here is what's the counter argument against instant inclusion what what do the naysayers say no no no here's why it's wrong and and how do you what's what's the biggest it won't work because the argument would to be you can't have real first come first serve you don't you don't have a notion of fairness and you don't know how to achieve it and there will be front running and what is what is there and does it actually work and could so let me

my PhD is in computer network okay so even before being a crypto guy I am a networking guy so my thinking is is that if validators make money that way and they are service providers you could say validators instead of see instead of a single proposer building a block and giving it the ability to order transaction extract value all validators together create a new block as an aggregate of their perceived order it's called femis It's A system that already joules from Cornell and chain link kind of like um and built basically the proposer doesn't decide on the order he received all the orders from all the other validators and say oh here's the order of their aggregate it

can't touch or change anything around it and so the question is can that be fast enough and the second again going be to being a networking guy if I send my transaction to all these validators directly but I don't just do it well I'll broadcast it and hope for the best and then get front run what if you start by sending it to the entity which is furthest from you and gradually send it closer to you around the world over the course of 100 milliseconds or so that means that you're being protected by the speed of light okay the one thing that jump can't solve and the biggest enemy of jump and they actually have like an article about it is the speed of light you can't beat it so even if they sit next to me and I give them the transaction and they

within a picosecond they identified it and send it to the other side of the world if my transaction has an advantage of 70 millisecond they're not going to beat me because it's going to take them another 100 milliseconds to reach there like they can't make it faster so if you send it to the other side of the world and gradually closer to you I I'm fairly certain we could prevent completely front running and if you could every so often front run what would be the cost to achieving it and how much like again this is this is my argument needs to be proven with real time but going back to my so the NASA is you would say what is even fair and you can't achieve it and we can manipulate it and you can prevent prank running so whether it's actually doable

or not which are Super valued point I'll go back to my previous argument I want to prove that it pays more money if it pays more money we can solve all these like like it won't work how do we actually do it how to make sure it's decentralized like everything around it and you know people say we can't do proof of stake either right I don't know if you've been long enough in crypto then nothing at stake problem is a problem we cannot solve right we're all doing proof of work and then like proof of stay can be because you could be doing also but we solved that okay like we could if we have heart problems and we want to try to solve them we have good technical abilities and iterate over them you know we had plasma we had charting and we had Roll-Ups like all

these kind of stuff but we only it's only worth exploring if this could be done if it actually pays validators more money so that's my agenda like the main role and first objective if we can prove that he pays more money then it's worth doing all everything else if it doesn't pay more money value it won't do it do it anyway so what's the point of trying to solve all the other technical challenge all right I got one more one more question slash attempt to poke a hole in your your beliefs here so your primary reason for wanting this kind of Architecture is so that the defy experience is better because you're arguing that users are leaking value you know by making transactions in defy but that sex to Dex are will always be there like no matter what into the future because it's just

different venues so do you not expect our Bots to you know run in there make gas fees super high and then ultimately you're left with the same problem that you started with and that users are paying exuberant amounts of money in gas fees instead of you know Mev so that's an excellent question what I expect two things for that or I do accept Arbitrage birthday I want them to participate and they will affect just like binance affect coinbase and vice versa so we're all going to have this Dynamics instead of see by running ahead and then every so often Okay just grab almost risk-free like trade against sale prices the risk is only okay who gets to cherry pick and take all the

money from the from the D5 ecosystem so yes I I want them to participate if we build such a system just like how block Builders and others could take transactions or bundles and if they revert not include them you could have exactly the same thing Arbitrage will try to send transaction and get included to try and these would say like I'm trying to make this trade and if it doesn't happen here revert it has to happen in this block in this state here it's in calculus and if it reverts because he missed it right three different Arbitrage tried to capture an opportunity the first one get in there as I don't then we don't include that transaction until like we're first of all we're in a post 1559 uh world so like gas behaves differently so there

will be usage and that would increase the base fee or whatnot but to the level of the usage that people are willing to pay but it won't drive gases the way that auction did that back in the day where all transactions got in everything you know one transaction hit 90 transaction revert behind him just waste gas instead of it it won't be in the incentive of the validator to include reverting transactions because they'll be getting basis points on volume if the transaction reverts it doesn't paint anything so it's not really worth doing so that would be my my come my two counters to to your to your poke right I do want Arbitrage to participate but to participate the way they are participating in C5 and just

like we're preventing waste of gas after reverting transaction right now with Mev boost and previously with map gef and in the future potentially with sway or whatever you could do exactly the same thing here the validates who are creating transaction decide on order and eventually don't include everything that reverts because it's wasteful and doesn't pay them in this model you're basically asking the validator to do more right it's got to run this uh piece of software that will effectively be making the instant inclusion impossible pop making that instant inclusion possible um so I guess the question is like ultimately how much is that added requirement and does that make it harder to run he's actually an X that thing's an excellent question actually I touched upon it a bit with Justin Drake when I

spoke at Laura Shane's podcast I'm like well won't this won't only large actors be able to do it in smaller wires won't Etc and the short answer is no the complicated part is let's simulate all the potential ways of transaction being included in all different orders and trying to extract that is a very convoluted they require a lot of like computation and heuristics and expertise to do it a software which receive transactions and put them first come first serve and broadcast okay this is one this is two this is three this is four to the proposer and the proposer aggregate that is super lightweight super that's very easy you could run it on your Raspberry Pi no problem

what about bandwidth internet doesn't just like sense you send transactions it's just sending these transactions it does not add a lot to that and maybe third well do you need to run something additional short answers yes something very lightweight but yes do you have to know but it would pay you all the Mev value that we're trying so like if you have 33 if to stake and run that so 32 East let's say what fifty thousand dollars or or more yes you can pay on your validator instead of running it on a hundred dollar Hardware run it on 150 Hardware or something like that and you have a proxy and you could accept transactions

and you could actually if this is a thing we spoke about Solana Solana have like jump building fire dancer okay like a a different client for Solana and this is from the people who optimize every single piece of the hardware and software coming from high places and trading firms this could be running on the networking card you could optimize this to run on this to be faster than your fastest computer if you build it and do it like on the firmware of the networking card and we could do that we have the expertise this is just identical like building like what what fire data are doing very doable LR a large task just like building a client is a large task but we could do it if we find that he pays more

value and we want to do it and yes will it happen in a month no will it happen in two years yes you could like it's not that hard to make these you know get packets drop them if they're already irrelevant all these kind of checks Etc you could do it on the hardware you don't need special Hardware stronger computer pay more well that was an awesome conversation honestly Yuri like really appreciate it this has been one of my favorite conversations I've had in a while so I'm going to kick it over to you uh where can people learn more about uh blocks route and yourself really find us on Twitter URI and block throughout everything related to trading and defy and private transaction and everything both in the current ecosystem that we're participating we're a large a large actor and both in these futuristic stuff

that we're trying to achieve and improve the D5 ecosystem I love it if nothing else here is a great [ __ ] poster on Twitter so an excellent follow we'll be sure to put those we'll put those links in the show now so they're easy for the to the listeners to find but thanks again Ray this is a fantastic conversation this was fun thank you so much thank you