MEV IN THE INTERCHAIN

Transcript

we are live welcome to the interop this is a show where we talk about the decentralized economic networks that make up the inner chain and I'm here today with braps okay how you doing nice to see you again yeah likewise good to hang out yeah um on on on a pretty on a pretty I guess you know significant day since today is the day of the merch merch day this is my second live stream today my third podcast in two days so um yeah it's pretty exciting where were you for the merch uh in bed oh okay yeah I just had like my laptop on top of a blanket and it kind of was um half sleeping half waiting for uh the collapse which never came and it was

like enormously successful and then I just watched the our Bots fight over the decline of die so oh so actually this is something that I I haven't followed up on today because I was over at zika Summit but this morning I was here we're talking with you know folks over here like Martin Coleman and Stefan and other people and um yeah we were sort of speculating about what would happen with uh the ethereum proof of work chain and so what happened well it's I think I don't think it's live yet I think I think that they're just releasing the source code now and uh it's a pretty uh Stark divide in communities from my understanding uh between the people that are sort of pro proof of work merge surviving versus

just moving into the future that was kind of set out upon seven years ago yeah um no but it's it's a really interesting uh case in sort of technical or uh it's a milestone in technical history where you have this huge transition of uh of an entire ecosystem which is comprised of a bunch of economic actors that have like enormous invested interests in uh in sort of economic composability sort of Shifting weight you know onto another foot uh all at once it was pretty incredible to to observe so so we haven't witnessed any of the you know like cascading liquidations or like the dye emergency shutdown mechanism and we that that

hasn't happened yet no I haven't I haven't seen that I think uh so so people like like Searchers were mostly uh eyes on Mev boost and seeing how that infrastructure was going to play out like on but on on on the proof of stake on the proof of stake too yeah yeah yeah so like I guess this is a big day for flashbot so if you'd call it to them for all the hard work they've been doing and integrated with multiple clients and getting the Searchers to reintegrate and even Fielding a lot of the concerns that the community has had over uh the approach of proposer Builder separation and uh yeah I think the first the first Mev boost uh block landed maybe two hours after the merge and uh we're actually seeing some initial numbers of like the

distribution so now in MAV boost it's not just flashbots so for those are unfamiliar um flashbots essentially in in proof of stake provided like a slot auction to sort of uh democratize mov Revenue to validators and then when it switched to uh proof of stake they introduced this uh PBS system called Mev boost which is essentially to ensure the economic viability of small validators yeah right and um of course when proof of stake went live it went live with with uh with with um with map boost and and we're seeing I mean what we expected we saw it in test net that uh that it was going to work so

it seems to be going all right and in the the current architecture it's no longer just a single entity that is running the infrastructure so it would prove mistake it was it was mostly flashbots uh but in oh sorry improve of work it was Flash plus proof of mistake it's now block x route has two relays and um and then flashbots has has their sort of auction system and you could actually see some initial numbers on who is winning the slot auction like with the with the builder so right now both flashbots are spending most of the time but yeah so let's let's just put a pause on on that for a second like we're gonna talk about that but first like who are you oh and what are you working on and uh you know for our listeners who

perhaps are familiar with you uh tell us who you are you know you were previously at at uh sure interchange foundation and then informal history and tenement and and Cosmos and so I'd like to start maybe right there sorry I think I just got too excited for murder yeah it is exciting like the thing about crypto is that things move so quickly that there isn't really a history it's kind of like an a historic ecosystem so when you like zoom out and actually have to say like where are you from what is your name or whatever you're like wait what does that have to do with yeah okay no uh I guess I go by the uh of raps and um initially from Montreal but I lived in Berlin for 10 years where we probably first caught um cross paths and I was working at a

company called subcloud and built out their data science team so my background is pretty pretty technical um and statistical in in many ways and as you mentioned probably around 2017 I got or 2016 I got involved in crypto mostly um to to work in financial inclusion started a company named topos which does uh banking for migrants currently deployed in Nigeria doing the the financial Court the migrant Corridor between Nigeria and the UK in particular but also the rest of the world and uh did that for a couple years until I kind of um realized that you know

technology as a as a mechanism for inclusion is always going to run up against regulation and that's like a whole a whole other Beast that is not sort of in my wheelhouse of uh of technology and statistics yeah so I decided to um sort of move the The Cutting Edge of the industry by by taking a position at uh The Interchange foundation so this is before IBC so I spent some time at when did you arrive at internet Foundation 2019. okay you were here in the Berlin office I was here in the Berlin office yeah okay never in this room but uh close by yeah close by right right back here yeah yeah um so then worked on on tender mint and and shipped IBC

and uh and also worked at informal so Building Systems for formal verification of IBC and tenderments obviously uh and then probably around um May of last year I left to start a company with uh with Alex Emerald called Mecca Tech um so I guess our our motivation for uh for starting mechatech was was really looking at IBC as sort of like a socioeconomic political system and not just a technical system and trying to fill in some of the gaps of what was missing in terms of incentive alignment IBC is a socioeconomical system yeah sure please elaborate well like you have people uh if it's going to be a

decentralized system so Cosmos the the cosmos thesis is all about sovereignty so Sovereign communities Sovereign people acting independently but with lower coordination costs and sort of reflect their common values yeah organizing self-sovering communities and uh I think in many ways it does this in a very different way than say ethereum which is like one giant Community one giant or not one giant Community but at the consensus layer let's say is one group of one now enormous group of people um that that well that's not true it could be a bunch of different people but actually it's mostly just Kraken coinbase and Lido uh probably because um it wasn't the consensus algorithm was not conceived

um purely through the or like with sufficient sensitivity to the socioeconomic and political dimension of of running these systems right um so seven years ago or when when did the eth2 sort of road map get laid out was like you know a long time ago yeah I mean I think I think the white paper already talked about proof of stake and I mean certainly the white paper already talked about transitioning to proof of stake right and then you know very shortly after those come I'm not exactly sure like when these conversations and when things actually start being built but yeah I think it's it's a long time coming they had many false starts and and one of them was like when tendermint was released when when

jayquan uh sort of designed tenement and took pbft and added some timeouts and had a very practical system for for doing proof of stake which could have been a viable option for uh ethereum but ethereum decided to go a different way and they decided to go a different way um in in a very very specific Dimension so tendermint shows safety over liveness yeah and um ethereum wanted to do liveness over safety so the big difference between I guess those consensus algorithms was is that um there is a period of time of which ethereum 2 will not have will not finalize the blocks versus tetrament has instant finality yeah and there's certain advantages to that like you could have a larger validator set so validator sets in in tenderment based

chains can be anywhere from 125 to say 500 people versus the idea of of B2 is that you have like tens or even hundreds of thousands of validators that was the theory so when the research roadmap for ethereum 2 set out it was looking at these different dimensions different trade-offs chose a set of the trade-offs that would give them higher decentralization right from a technical standpoint but from a political from an economic and political standpoint in eth2 today you know you have like three parties that could uh I think it's two parties that could threaten liveness of ethereum too as it is today and I think it's just Lido and coinbase and if you add Kraken um you can actually uh threaten safety as well so I think my

co-founder and I saw this playing out we knew that there was this this new field of economic study of blockchains about how do you make them decentralize and how do you um form communities around them and a lot of that work was happening underneath the sphere of Mev I mean to be fair though I mean this is also the case in in many Cosmos chains where you know a handful of validators effectively control Beyond 30 percent or Beyond even maybe 50 of the general like I haven't looked at the actual numbers but many Cosmos chains have highly concentrated uh staking providers uh in the top ten yeah which you know has has hindered governance I think in some instances uh absolutely and uh could perhaps even

um threaten uh like safety safety of the chain yeah yeah so if you look at the cosmos chains there's like four or five really really big ones um Cosmos Hub osmosis Juno and fmos would be would be the biggest um the centralization across them is not even it's actually probably probably the worst on fmos but it's also the youngest so you guys and they're trying to fix that it's probably in some ways the best on on the cosmos Hub the cosmos Hub is probably one of some mature evaluated sets um in terms of infrastructure providers but it does have a lot of uh top validators that are institutions so uh coinbase custody is top elevator Cosmos Hub

um and then uh Paradigm is also a a top validator in the cosmos Hub and those particular validators so binance coinbase custody and Paradigm are are you know they're good infrastructure providers but in particular they don't vote yeah right so they don't really represent the sort of community spirit that is sort of core to the sovereignty thesis of of Cosmos yeah um versus other chains like uh osmosis was a little bit uh more decentralized but the point is this that even though there are some challenges to Cosmos in terms of decentralization in terms of uh of governance and participation it is better insofar as those are like chain

specific problems as in the cosmos Hub can solve that problem in a cosmos Hub specific way versus fmos can look about solving that problem in an ethmo specifics what's the beauty of app chain sovereignty and every chain having to deal with with its own issues uh sort of independently the beauty of it because it's all using because all these chains are using essentially the same software exactly they can sort of set practices and maybe even like set standards around how these things they can try to or they could even [ __ ] up in parallel in the ideal so insofar as they are independent uh validator sets insofar as like the validators on um the cosmos Hub are not the exact same validators on osmosis right we could actually learn as a community faster the

problem that Mecca Tech is trying to solve is that this is actually not the case and actually the incentives are not aligned for that to continuously be the case it's actually there's a disproportionate advantage of being a validator on multiple chains so one advantage is that you like you know how to run a note that's one advantage you have another is that you have reputation reputation that you develop on one chain will help you build a delegate and build a community on another chain and the problem is that the more even though the economic incentives are for that um the community incentives are against that so as a community the more the communities they take part in the question is do you really represent that Community yeah are you really allowing communities to be Sovereign and to evolve in parallel yeah and I think there's also the question I mean you

make a good point that you know so there's economies of scale in in running a validator where you know once you've once you've established you know your your stack it's very easy thing Cosmos 2 then like just spin up new validators on new chains and like that that's like mostly a solved problem then um then there is the yeah the reputational aspect and certainly validators uh have an incentive to build reputation so you know you see validators like building Brands and having like clear um clear sort of like uh voting uh practices and things of that nature um you know we're thinking of we're thinking of spinning up validators that fund public good and that's so that would be like kind of our marketing

Spiel right it's like our validators find public good and so hopefully like people will will delegate to us and that makes sense for us um and then there's also there's also this idea that cross cross Across The Interchange ecosystem uh you know if if we have um most of the chains uh with the same balloters like is the ecosystem really decentralized like Across The Interchange so if like say I don't know validator a validator b and validator c uh letter D and E are in the top ten of like all the major chains um does that represent a does that represent decentralization across the interchain and I think like some folks are doing research around there around

understanding the interplay between um the security of different chains and like how interdependent uh the validation sets are is that something you've looked at I think they will be as long as they're like to be in Cosmos just means to be connected by IBC then there will always be sort of like uh asynchronous and ideally interdependencies between chains as they compose functionality that's like a good thing the bad thing and the and the sort of economic problem now we are uh most concerned with at mechatech is that um if you have like the unique privilege of sequencing blocks on multiple blockchains then your execution risk of co-sequencing or doing cross-chain

operations is dramatically lower than everyone else so for instance if you're a market maker on osmosis as well as on on Crescent on Crescent decks and you want to like do a trade to arbitrate some small discrepancy or you notice a price on osmosis is far lower than a price on Crescent tax right if you have the unique privilege or if you have the privilege of sequencing both those blockchains at the same time so your proposer of a block on osmosis at the same time that you're a proposer of the block on um on Crescent decks right then you could do that operation dramatically faster which is with dramatically less risk and more profitably than every

other person so that means that there is the economic incentive to centralize the network and this is the field of cross-train Mev where mechatech has been sort of researching and developing products for and this is this is what you mean when you say that Mev is a centralizing force that it it forces validators to or it incentivizes validators to become bigger and to have more uh more of a probability of of being a proposer on several chains at the same time that's exactly right it's it's a it's a field that gives you a disproportionate Advantage like a super linear advantage of being bigger so like as a small validator you're probably not going to run operations to extract Mev and you're probably not going to sort of like manage bilateral agreements or

option agreements with Searchers to sort of sell your block space at an efficient price yeah uh versus a large validator like the the block demons um the chorus wants to call out anyone by name but these are very successful uh members of our community that do have this unique unique privilege and that as a community we need to develop sort of incentive structures and better protocols for mediating this centralizing Force so maybe let's let's just like take a step back here and you know because I think like Mev and like you talk about flashbots and I I feel like this is a topic that's very much talked about in the ethereum space but maybe like less so in the cosmos space or at least like in my Twitter feed people talk about Mev a lot less uh when I'm looking at like my

Cosmos list and like if you get my like ethereum uh evm list like it seems like everybody's talking about EBV maybe research and flashbots and everything um so maybe just like to contextualize what is the history of Mev in in ethereum and one one like when did the research start uh you know being being built around this in particular I think like flashbots was sort of a selling point yeah I think that the the research mostly came from uh like tradify either in uh like hft research so looking at how uh people buy order flow how fees are paid and and how those markets develop and how services around those markets are are structured so in particular in hft a lot of the

fees are very low because the doesn't cost anything but more most of the profits for for markets come from like co-selling um real estate sort of close to The Exchange so that people like hft firms can um can get lower latency execution so they pay for the execution through the um just being closer to the exchange yeah so much to the point where they actually like measure the fiber optic cables yeah um the thing in this creates all the servers have the same length the fiber optic cable so like no matter if you're closer to the door or further away from the door the length of the fiber optic cable is the same yeah because they're trying to provide a a commodity right which is access to the inventory this and this creates a lot of weird Dynamics like in for for retail users who might

use like a Fearless um trading service like um like Robin Hood for instance the question is like why is it free one free is like the benevolence of like Venture Capital backed startups is one idea another idea is that there's actually a viable business model behind selling order flow that has nothing to do with fees so that's how that market developed and that's how it centralized around you know big firms like Citadel who could pay Robin Hood for for that order flow at the expense of retail users you know that that's what happened and try to find what happened in ethereum because you know for a long time like blockchains are still very new the Market's still developing and even amms are were not built that long ago we used to have order books like each Delta

was it was an order book based system which would have had a completely different Market Evolution Dynamics if it was fast enough at that time but really what happened with with amms is that people started um racing in the mempool at the expense of the validators so in the same way that in Trad five people would buy fiber optic tables cables to get the best execution people would uh race on in in the peer-to-peer layer of ethereum to sort of bid as le as little as possible to get an opportunity to get execution at a really good price yeah and this created a lot of spam so the amount of the we call this an mvv like a negative externality of trying to buy something that you want so you want execution you

have money to pay for it and the only way to do that is go through the wrong place which is the peer-to-peer layer is that every node in the network has to Service as garbage traffic so there's a there's a transaction that is never going to get executed that is just an incremental bit of information is now being is now burdening every single ethereum note is what was happening so this was uh the research of of Phil Dion and Co in The Flash Boys 2.0 Paper so like from Flash Boys 1.0 was hft Flash Boys 2.0 was the ethereum of priority gas auctions that were essentially spamming the network and that's what sort of incentivized um Phil and and and uh and his team

to go step and go to put all of that garbage into an auction system to put that system where people could just pay for what they wanted because that way one the negative externalities of spam and of of garbage P2P traffic would be eliminated Traders could get exactly the execution they wanted and they would actually maximize the revenue for the validator so instead of it the utility being bled out through computational resources of just a price um so that that's where ethereum came from and there's a like there's this idea of like good Mev and bad Mev maybe that it's worth getting into here and um I was watching earlier like this talk that that Sunny did at nebular Summit and uh

um and where he he describes good Mev and bad muv can you can you break that down for us I think it's a it's a it's kind of an annoying topic sorry it's fine we'll we'll have some some fun with it in insofar as it for for the be any clear categorization of good and bad MAV requires such a preliminary understanding of like market dynamics and how this thing is is going to play out like you could say that a bad MAV is one that um let's take one definition if you use information from the mempool to decide on a trading strategy that's bad MAV which might incentivize you too um I don't know randomize the um the the the mempool or whatever so

this is one idea that has been played out and created worse outcomes for the user so when you randomize the the mempool it means that you're incentivizing people to just Spam for the best chance of getting and you've recreated the exact system that you're trying to get away from right so these formulations of good and bad Mev don't don't really make make sense because there's oh they just push the externality somewhere what may not may not belong I think that a better way to think about mov is about overlapping incentives and to figure out where do they intersect and where do they maximize joint utility for the community members that are that are operating there in particular like if someone if like if you have

um if your mortgage is gonna is gonna go go bust the the bank is going to repossess your house if you don't uh file a mortgage payment you need to put a check in an envelope and send it to the bank 12 p.m sorry 5 PM and a business Friday let's say yeah and then you have a system like uh like a like a postal service right that just has one box it has envelopes and everyone's in there so the utility to you of like or say you know Columbia House is sending a CD to Grandma that doesn't need to remind some of the people listening to this what a CD is a compact disc is like a laser engraved data format that people used to put

Crystal Waters on yeah they used to put like CDs and they used to send them to Grandma like through the postal service and people will get them and you put them in the in your CD player there's like a CD player that you know had like play and stop and fast forward and stuff like that and they they kind of use the Postal Service like to to Great expense of a lot of people because it was just like it's one box and if you get in the box then your stuff gets sent and if you're not on the box and stuff it gets sent so the marginal like cost to like Columbia House in Tennessee is effectively zero like whether grandma gets it or not she's probably not going to listen to it right so they could just use that system but the marginal like cost to you who's

trying to like file your your mortgage payment at like 459. Friday evening right is the cost of your [ __ ] house right so the question is like is that a good system and like is the inability like who if if it's good if you have like a randomized like someone who gets in there is randomized so you like have a 50 50 probability between yourself and Columbia House of like Grandma getting a CD or whatever does that maximize The Joint utility of society I would say no because those those incentives just don't overlap so I think there's there's generally two camps there's like the uh minimize Mev at the protocol layer so either randomize or

encrypt or whatever and then there's like sort of the make Mev legible right and uh and have the market deal have the market like introduce a market to to to let people pay for for what they need and allow those overlaps of incentives to be rendered in price you know so those are the two camps in Mecca Tech referral we're firmly on that camp because oh much cap on the market cap on the market cap yeah yeah because um because we have we have you know 20 years of of research and Market Evolution to support the fact that people if people cannot pay for what they want uh in a an efficient way they will build systems around it and those systems are usually worse for small

people okay and I mean so coming back to to the cosmos ecosystem and you know you're talking about osmosis and and Crescent um you know osmosis uh implemented a mechanism to encrypt transactions such that uh the quote-unquote bad Mev uh couldn't happen or was much more difficult to uh to implement what what are so what are some of the externalities that this causes right so on one in one extreme you were talking about the fact that like randomizing just creates an opportunity for more people to span the system in order to get in but like encrypting transactions what's what are the external negative externalities that I could create well for one uh so friveo is the system which is designed which is uh uses threshold

encryption to provide uh pre-commit finalization a pre-commit um privacy to transactions right so it means that as a user you know you uh encrypt your transaction and then it only gets revealed to the user once the Block's finalized and when it's finalized yeah so I mean you can look at the the externality and multiple categories one is that um the current mechanism requires all users to bid the top of the gas price for inclusion so you can't like modulate your your gas price there's no gas limit in travail so every transaction whether it uses that amount of gas or not it's spent so it's not like a very efficient

mechanism in that way okay another way that it's not efficient is so first of all it's great I shouldn't just like it's amazing is the paper that was put out by uh Dev and um so I forgot his name is John from uh anoma it's it's fantastic work and and I I should really um praise it as an example of the flexibility that you can get with ABCI plus plus so the fact that you could Implement threshold encryption uh is like a is a tenderment specific Advantage right because you have access to all these callbacks yeah for for Block construction and gossip you could instrument at that level so it's it's very very exciting so I we need to recognize that you know we're moving the state of the art forward as a

community um and that this is a step forward but there are still drawbacks so beyond the benefits right or the purported benefits right we look at it critically just because in the name of evolution okay so I should say that from from the get-go so from a gas efficient standpoint it's worse from a latency perspective it's worse so you have to um one encrypt data I have to figure out what to encrypt so who who you get the encryption shares from or the way that that threshold encryption works is that as someone who wants to encrypt something I trust a bunch of people a distributed set of people to provide partial keys right yeah and then I take those partial keys and I encrypt my stuff and I trust them that they're only going to reveal

those keys at the right time that's my those are the trust assumptions I Prevail yeah and in particular I think it's it's two-thirds you need to you need to trust two-thirds of the validator set um to um to to not pre-reveal to to not extract Mev in this case which has several drawbacks beyond the one about gas it's also pretty slow I think in the current estimates it's like 21 it's a 21x uh speed uh decrease so it increases the um the latency of transactions like like the time it takes to to get execution a lot which means like as a user who's trying to execute a trade you're going to get a far worse execution you're going to pay more for a

worse price because it's going to take longer is it worse than say getting sandwiched you know we got to figure that out there's a lot of arguments that say that actually sandwich attacks improve uh um routing and execution in aggregate um but so that's worse about another thing that's worth it is the premise that two-thirds of the validator set would not collude to extract Mev is categorically untrue in the case of of ethereum and flashbot so ethereum flashbots they went from like zero to like 75 I think in two months yeah as in like the the number of the the amount of hash rate that doesn't go through flashlights is like you know 10 or or less yeah yeah

so and uh and now that may be boost I haven't looked at the numbers um but the assumption is that the majority of validators are going to want to extract Mev because the consequence of not is getting diluted so if you have I think the current estimate for the the difference the APR on Mev is probably going to be about six or seven percent let's say so if you're making six or seven better returns in a proof of stake Network or Worse you're a validator and you're going to take like a six to seven percent haircut on your returns how long are you going to be in the voting set for you have a fixed amount of people in the validators had 150 whatever it is you got people who are compounding five to six percent yeah over you how long do you expect to be in business yeah right

so it's not you're just going down that value set day after day exactly as as your as your competition is compounding those returns yeah right so I guess the in addition to the commissions they're making here exactly and I guess the the um in in the lingo we say this is not uh incentive compatible but like it is not I guess the why do you think this is not this isn't happening on on osmosis yet why do you think that validators are not yet colluding or or are they I don't know the volume's too low volume's too low it's obviously low yeah yeah so I guess like like so you think that there's a there's a point at which like there's a threshold where validators start to say okay there's Meb here and there's something that like they go away from

this like sort of yeah they start to want to align their incentives like their incentives become compatible it's it's very much already happening if you look at Tara that was doing a billion dollars of volume from Dave we were in the market when it when it went collapsed and it was very competitive yeah right it was very busy blocks were full users were getting impacted by the negative externalities of the Met of the the token fee mechanism on on Tara when when it collapsed so it's real there are firms they're very serious [Music] um because there was a billion of volume per day for 10 million in volume on us mostly no one gets out of bed for that if you look at like the the numbers that the course one's posting we're talking like two to three k per day those those reflect our numbers as well we do it

mostly uh as market makers to just balance liquidity in pools it's like a utility for the ecosystem but for instance like jump is never going to to Market make for those volumes it's just not worth it for them but the expectation especially in Cosmos which is which is growing so fast right and only needs these singularities it could be DUI DX coming into Cosmos with with a billion dollars of a volume uh per day which those markets are going to evolve very very quickly so I would say like we can't take the current state of things for granted so you know in in um in Cosmos um well yeah let's before we go there maybe

let's talk a little bit about yeah Mega Tech and uh like what you guys are building and I like there's not a whole lot of information yet so uh I don't know like we're still about marketing when you're ready to launch or if you have launched or but we've launched okay so like Mecca Tech is building out the Builder layer for for Cosmos right so we're trying to align the economic incentives for the decentralization of Cosmos to ensure like um that there that the that Cosmos owns actually are communities of people of of a range of people like large validators institutions to block Diamonds the figments um mid-range the you know the chorus ones these are these are excellent people as well as small validators that

uh propose more more radical progress right so when it's like interrupt delegate like internet like Valley Dallas um I guess notional is not that small but they are quite radical and I think they are important voice uh ensuring that you know the cement doesn't die on this on this community um so the the first version of that is live and it's it's just a a top of block auction for Searchers so we run searching operations to do Market making uh we had invested in infrastructure to sort of like navigate the fifo market to win latency races and it felt like a waste of engineering resources to like pay Amazon more money just to co-locate servers to be able to to run profitable

operations so we moved all of that um that infrastructure into what we're calling the Builder API which is currently live with with a few validators we're just publishing the last few touches before we sort of make it permissionless but it will be permissionless so anyone could sort of as a Searcher you'll be able to use it you know um at the integration some documentations maybe we'll put in the show notes and then as a as a as a validator you'll also be able to sort of subscribe for blocks that are subject to um to this auction mechanism and increase your overall Revenue without uh affecting latency or or health of the network okay so what what does the product look like now practically

speaking so is it something that validators can already use or is it like for another sort of category of Searchers that are just looking for the Mev opportunity so as like a validator yeah who's [Music] um you know doing a lot of things like validators they gossip transactions they have like mempools they actually order transactions and the idea is like our validators the best people at maximizing like building the optimal block like if block construction is like ordering transactions whatever are they the best people doing that or and are every is each validator equally good at it to ensure like optimization of of the overall Network I'd say no um that's not really their specialty and probably not what they want to do so instead they call out to us which is an

API service like the Builder API for mechatech and uh we give them the block now how do we construct the block well we work with Searchers so our Searchers other other people Searchers to essentially figure out who wants to pay the most for that block space what is what is the real value of that and we do that off chain and then provide hit the validator when they ask us okay that's interesting so as a validator you know you receive uh transactions uh through the mempool um I I mean I'm not so much into like the technical leads here but of running the validator but I guess the validator is doing kind of like first in first out yeah um ordering of transactions that's

probably like the default behavior of a cosmos SDK chain validator uh what you're proposing is where there's there's a some piece of software that's running on your node it is calling your API and it's saying what's the optimal uh ordering of transactions that I should Implement here and then also some extra transactions probably that will allow you to x uh extract some amount of profit from exactly that ordering so in the case of which like your mortgage is past due and you just want to get into that block at that height yeah you could just pay the validator who then pays their delegators for that service isn't that also a little bit of a like I see that also as like a centralizing force and that you are

proposing the you're you know you essentially are proposing all of the block ordering to all the validators so what what's the risk what's the risk that your API is providing faulty data that um uh that causes validators to sign faulty blocks and then get slashed well right now we always provide the um the the validator so that the actual integration for the validator is a patch that has like a fallback mechanism which ensures that the validator can always propose its own block okay talk about the Builder layer we're talking about many components and this is just like a first iteration of it that comes with certain trust

assumptions that is purely meant to reduce the negative externalities of spam right over time there will be multiple Builders so for instance like mechatech will be a builder maybe block x route will be a builder I know flashbots is a builder skip is going to be a builder the idea is that they're being many Builders and that they're competing to maximize revenue for the delegators right so there will be uh a more advanced version that will allow sort of like load balancing between different Builders so that there actually is like that sovereignty all the way to the validator layer to ensure that they could choose one what do they what do they want to sell do they want to sell top of block do they only want to sell the end do they want to sell the middle they should be able to sell whatever

they want and who do they want to sell it to um so that's that's the evolution that will be sort of like robust to any one party whether it be us or flashbots block extract or whatever sort of maybe sanctioning or just not reflecting the best interests of the network the Builder layer so our service right needs to be a member of a builder layer which needs to be competitive for the overall robustness of the system okay and um he's like what's the latency because like if you're if you're running it's like on Atmos blocks are like fairly uh frequent it's like every couple of seconds so this is very fast yeah yeah it's very fast we're only on osmosis today okay um

but I mean so all of us the the engineers at at Mac attack all come from um distributed systems we do this a long time we could build fast systems that can do this in under a second okay that's not really the risk right now okay um yeah I'd like to talk a little about like this idea of cross chain Mev because yeah coming from this perspective that we have individual app chains that want to maintain their sovereignty of course the issue here is that we don't have composibility and we need to build systems that allow chains to interact with each other so like IBC is one way to do that but once we get into other ecosystems that perhaps have different finality uh assumptions and um then we need to construct other systems so we have like Bridges there's one like

and and of which there are several different um new design patterns uh which uh have their varying trade-offs so when looking at just at the maybe let's start with the IBC ecosystem what are the Mev opportunities and like what are some practical applications of Mev when um when moving tokens across uh different chains currently and then like in the future as we start to expand IBC with other features right so in the sort of um when we think about Mev we think about sort of the the supply chain of like people who like produce it and people consume it so the easiest sense let's like just start simple right so you're executing a transaction only on

osmosis so there's like a wallet that sort of constructs or is the originator of that transaction right and that submits to some either through to temec attack or directly to the to the gossip layer or or whatever and then eventually gets finalized uh on on honest Moses right so there's like the original originator there's probably some kind of like builder in between and then there's the proposer that actually does it and all of them need to be compensated to ensure the incentive alignment of the ecosystem that's like the single chain case in The Interchange case things get a little bit tricky so in the case of which um there's some so let's go back to Crescent and osmosis where you could see

like there's some overlapping in trading pairs between those two chains so if there's if you know osmo is is you know 125 on osmosis and then three dollars on Crescent decks you know there's some certainty that you could probably depending on the depth of the pool you know that there's money to be made there and the question is like who should be compensated for that or how so let's say in the current state of things um osmosis whether it's encrypted or not so say say osmosis adopts for Veo and provides pre uh commit privacy pre-finalization privacy so no one knows anything before um the transaction is finalized on

osmosis right but the finalization event is actually an origination of of Mev so when you finalize it tells you that that price is going to be 124 and that there's a trade there and you could use that to make you know whatever the spread is between you and Chris and Dex and who gets compensated for that Crescent Dax so this cross chain Mev case is very tricky because it's not really incentive aligned in sort of valuing the the complete supply chain between who generates the Mev and who captures that right so one of the things that we're trying to do at Mecca Tech is fix this as a product uh uh the the Builder layer will eventually evolve to support multiple chains and then provide Atomic execution on multiple chains and make sure that the delegators

on uh osmosis are renumerated for the transaction uh origination as well as the um the delegates on the validator up on Crescent decks things are get even so that's the second complex case the third complex case where things get really interested interesting is where um interchange accounts allows not just like users to be Originators on chains but the entire Zone essentially can be an originator for Mev right so in interchange accounts it essentially allows um zones that have foreign representation so you could be like the cosmos Hub could evolve into like a platform for

Capital formation you know you have a community pool and say you want to make an investment right that's going to be on Crescent decks and the community pool passes the governance proposal and says yes we are going to buy you know 100 000 interop tokens and those interop tokens just happen to be you know on on osmosis right yeah so the cosmos have validators right the entire zone is like the originator for that Mev but who's going to capture it on the other side right because it's like even with encryption or not encrypted it's going to be finalized and it's going to be it's going to be a commitment and there's going to be value there and the question is like who captures that value so right now in this particular case with interchange accounts similar to the Crescent decks and uh osmosis case it'll

be someone else then who generates that okay I'm not sure I exactly follow there so can you give me a practical example of like what this would look like because the way the way I see cross chain Mev is uh and and sort of like the the use case that I had envisioned for this is you have uh someone making a trade on let's say on Crescent right right and like there's there's a wide um there's a spread between what the price of that asset is on Crescent to what the price of the asset doesn't sound much correct and so be um because there are uh Searchers looking at the entire chain and they're all the chains and they're able to like identify what prices are um the block uh producer would introduce transactions that would in fact go do

that trade over on osmosis because it's more profitable for them correct and then they would capture the spread and then like give the user whatever that user thinks they should be getting because that's what the price they saw yeah so I think we're like halfway into the podcast and we haven't done the traditional thing of like explaining what Mev is you know so maybe we'll start with just a definition which I think will clarify things a little bit so sorry about that but like Mev is like the the maximum amount of value like a sequencer that that can be exploited by a sequencer of transactions yeah so like if if you are the privileged like finalize her if you're like the first person to know what's going to happen on uh on osmosis either as like a proposer or as a finalizer maybe the same person

but it's the maximum value you could extract from that so in the traditional case you would think like Oh I'm a I'm a validator right I see some transactions come in it's for some spread or whatever I will inject my own transactions or that someone will pay me to to inject those transactions maybe front run it back run it whatever and then the payment that I get from that operation is Mev right that's one but that's one form of Mev another form of Mev right is as an originator so if I finalize a transaction and I know that there's some huge displacement that's going to result I know that the cosmos hub is going to buy you know a like a hundred thousand interop tokens or whatever I have unique information

right we have this is like confidential information this is the kind of information that like Nancy Pelosi steals from her husband right so this is worth a lot yeah right and even if you're not sequencing that that block even if you're not finalizing that block or exploiting it there the fact that you can sell it to someone else as order flow right is very very valuable so in The Interchange case right A lot of them the the cross train Mev is going to be the incentive for validators to sell information right to essentially become Originators of of transactions right for other chains okay can you give any an example so come come back to this governance uh this

governance proposal example um yeah can you maybe run right here's another example here's another example with uh with um nfts right so when nfts get minted yeah it usually creates a huge auction for the floor price yeah it's like this huge race so right now when does it when does it when does an nft mint happen well it depends you know they're like scheduled and usually there's some Insider information about it and all the Bots are are set up or and they're competing for that but what if you have interchange accounts where like the Zone who's like initiating the mint is not the one that's executing the mint so like what is what does it dimit it's just you know a smart contract call yeah and with interchange accounts that could come from anywhere so the cosmos Hub

could mint an nft on Stargate of I don't know you or us together yeah right yeah and then the question is like who knows first so it's not going to be stargaze Zone anymore it's going to be Cosmos hot okay so what you're saying is by using intertrain accounts the the cosmos Hub um can initiate a transaction yes okay maybe maybe my understanding of entertaining counseling is not is not um perfectly correct So within the cosmos hubcod initiative transaction on another chain correct and then uh benefit from from the Mev but so I mean for me interchange accounts is simply the ability for

um a user on one account on one chain to control an account on another chain correct but not the chain itself it could be anything it could be the the entire it's just wrapping a message right so how that message is generated how's the how's the cosmos Hub or how does the cosmos chain initiate a transaction as a whole I mean it's just like it's so interchange accounts envelope yeah right and inside is like a message that's going to go to a state machine yeah so how that message is is integrated is like you know it could be a user it could be a Zone it could be anything it could be a dow you know on a Zone it could be anything just a message yeah right so who generates that message is not really the point it's where it's generated and where is it where is it

guaranteed to happen so the point is this like if you generate an uh uh an nft mint on a chain it's going to happen the question is when and who benefits from it so if it's going over IBC you need like confirmations and all that stuff but it's going to happen it's going to be asynchronous and the question is like who's gonna who's gonna um get the floor price for in the nft case okay I think I think I understand yeah no okay but yeah I think maybe it's just my my understanding of entertaining accounts needs to be broadened a little bit because I I I saw this like very kind of narrow use case of energy and accounts as like I have some tokens on chain a and

um and I want to control those tokens from chain B so for instance like I want to be able to um tell these tokens to say like move over to another account or I want to like use another account another chain to vote or something like that so in general like Cosmos is just like this General message based system yeah right which or the cosmos SDK split into modules with process messages messages can do lots of things one of the things they do is move tokens another thing they do is like process evidence they slash users they mint nfts like it's a very customizable system that fits a whole set of of applications right to fulfill this sort of like application specific blockchain now usually different applications do

different things and the question is like how do they communicate usually they communicate over IBC which is a general standard which is not application specific so like token transfer is just message it's just message passing so the message that could be passed between them through interchange accounts can be like uh it could be a foreign Zone specific message so instead of me talking to you in the general IBC you know application Level like erc20 oh sorry ics20 is the token transfer standard for for for for IBC and that facilitates exactly as you say sort of moving transactions or moving moving tokens between between two different chains with industry accounts or ICA right it's generalized it's not just moving tokens it's literally any

message that is interpreted by the foreign State machine which could be take out a loan it could be uh mint and nft all the application specific utility becomes composable over IBC with interchange accounts so it's actually going to be kind of a a big deal um when it's released do we have a a timeline and one entertaining accounts but goes line I think this year yeah I think before shared security I think it's uh entertaining accounts and then share Security in uh early next year yeah um yeah I mean there's like lots of things also here that I wanted to talk about but we lost a little bit of a a a bit of the um the flow yeah sorry about that no

no but when you work in Mev you become like a little bit isolated because it's at the intersection of like economics and and Game Theory and consensus design and you you really have to to make progress in that in that intellectual space you have to go quite deep which I think is one of the the issue the education issues that we're having across the ecosystem is when we think about it we pose these very simplistic questions like is it good or bad right which is a very reasonable question to to ask like am I going to get hurt by this like as a user are my users going to get hurt I'm protocol designer application designer trying to go to market all my application all my users get going to be impacted by this and the answer is a very complicated yes right for a whole bunch of sophisticated

reasons that we're still sort of exploring you know we could take sort of the lessons that we've learned from from traditional finance and know that like what's most important is that the value chain is compensated we could minimize the um the the the gossip by not randomizing we've learned that from from ethereum but the truth is that there's still an enormous amount to learn you know so it's it's not it's not the easiest conversation to to to to to sort of navigate without with the context so one of the things I was thinking about uh when preparing for this is you know like mechatech is say one um Market uh for for Mev or would you call it a market

uh but yeah it is a market yeah okay so it's like an auction market for maybe and like you propose uh block ordering that optimizes Mev for specific miners so like as a minor I should be incentivized I should have an incentive uh to to use Mac attack in order to uh generate as much um value from from the from my block space uh one other uh way to look at this is say we have a group of like um say more altruistically uh instead of or um inclined minors that would say like okay this this Mev that we're extracting we're actually going to funnel it back into the ecosystem and maybe put that money in a dow uh have that Dao um fund public goods for instance or

like go to some other cause I don't know maybe like doing some like charity to Fight Hunger or whatever like uh is this is this something that uh exists and do you think that it's possible that some valid errors would um would participate in this sort of activity and like is it you know maybe coming back to some of the things we talked about earlier maybe the the answer to that is that like it's not viable just because uh there's an incentive to like stay in The Ballot or set and sort of like if other writers are extracting value and compounding that interest like you'll never remain in the dollar side unless you're uh you know attracting continuous amounts of delegations well yeah I think there is a way to make this happen I

think Cosmos already has like a really good governance uh system for this and it's just like the delegation system delegates choose their validator that reflects their values that votes their way for those reasons so one thing that could attract delegators is just Revenue right you could have like you don't have to think only about my miners in in the way that we did in sort of ethereum eip1559 research but instead can think about it at the governance layer as like what are you selling to your delegates it's your job as a validator to sort of represent their interest what are they interested in so I I think that there will be validators that have different sort of Local Economic policies for what they do with their their revenue and they already do like right now there is like a uh an

um a free to choose commission rate for validators on most networks you know it's a pretty contentious um conversation actually yeah to to say who decides how much Revenue validators should share and and with who right so one idea is that it should be done at the protocol layer there be should be some minimum for this idea that it'll increase decentralization I think that validators my personal opinion is that like validators should charge enough to be sustainable because the idea that they could not make money means that they have like a low cost of capital which again is like some source of centralization like you probably want like revenue streams to be quite um uh transparent to ensure that these

people actually are representing the interests of of their delegates like if they have like a zero percent commission they're probably selling order flow right they're probably doing an off-chain agreement with someone that you don't know and that's how they pay the the servers yeah um I don't think it's that competitive right now I know that so if you're like a validator in uh like the sub 100 on osmosis right you're probably making like 50 bucks a day it's like not a it's not a place you want to be so the people that are there or it's not a place you want to be economically let me say that it's like not enough even though servers you can still be profitable with 50 bucks a day on osmosis I mean like like you're you're yeah I mean like you're not gonna live a life

you're not going to be in but like you know what is what is a like hessner Bearer marital server cost like 50 bucks a month right like so um yeah yeah you could you can do it um but I what what I'm saying is that valid it's important that we have like a lot of heterogeneity uh in the validated reset to ensure like the ecosystem continues to evolve right and that means that even low validators need some kind of sustainable rate to participate not just to buy servers but to spend the time reading governance proposals right so there's like you know computer time is is not expensive right yeah no one no one cares about that people care about people time and about like what it takes to be an active member of this community what does it take to like keep up to

date on Mev to listen to this podcast to do the research necessary to make very difficult decisions about which way the ecosystem could go and I don't think 50 a day is enough I think that it should it should be more it should really 75. 75. no like you know one one of the things that that we're thinking about uh with interop and and some other some other friends that um you know shooting this idea back and forth and uh this is sort of like Alpha here I mean I've been talking about it all week but I have never mentioned it here on the podcast is that we we want to we want to create a validator that um takes a higher commission so instead of it being like five percent the commission would be something around the range of 30 the validator um commission withdrawal address would

be a Dao right so like we would uh withdraw all the commissions to a dow that Dao um would pay us some amount of money to run the infrastructure um and then the rest would um would be up for governance so the this would be like the Holy Grail here would be for uh the delegators to have Dow tokens in proportion to their stake now the the challenge here is that that stake you know is constantly moving so like would you have like a rebase mechanism or what like that's that's a bit complex but essentially the idea so that the delegators get to decide what the money goes to right and it's like why does this not exist I I know like oh I know really like so we thought of this

and we're like we're just a bunch of idiots you know we thought of this like why does this not exist and and um you know I think there's there's some precedent for it if you look at fmos like there's a top five Ballot or that takes 100 commission to fund public goods and to fund like uh I think her name is Loretta and she she does this all this open source development and there's another validator that takes something like 20 uh also I think to fund some kind of like education or public good and so we we we're looking at this opportunity we're saying like hey like this is this is a cool thing that allows us to fund probably good and set some precedent around like how delegators um think about like what their stake means like really this money that I'm going to be staking

what is it what is it doing to help the ecosystem not just like help me as a delegator you're still making some amount of commission but you're also helping grow the ecosystem by funding this public good and then you know that they kind of um end State here and this you know this this is a lot more complex and you know is in order to uh to help even grow the the stake on on some of these chains is to set this up such that the Dow is actually like a registered charity or like a registered non-profit that uh like is tax deductible so then you can you can go see investors and say hey like buy it that buy a million dollars of Adam stake that Adam you get to participate as a delegator you get to

you know enjoy sort of like the long-term uh you know if you're along you get to be long on Adam and you're also getting a tax break because you're funding public goods infrastructure it's like yeah it's a long long road to get there but that's kind of like the idea though that we want to try to explain so is the implementation just like changing the the payment address and like a website or like well uh that that we're not entirely sure so I know that when you so in a cosmos chain uh you can specify the delegator the the commission withdrawal address but you know of course like we you know we could change that at any time so that might not be sufficient um I was talking to guys here uh Obel they're doing do you know them so they're they're building this kind of

um I mean it's mostly for for ethereum validators right now but it uses BLS signatures to split up the validator key so that multiple infrastructure providers are validating on the same key like core clocks yeah like horcrux but where um you can build clusters essentially of like different companies uh and it's like a service right so it's like I don't know if can you do this with artifacts can you I think you just like it's just threshold decryption for the keys so yeah split it up so I think like they're building this sort of like as a let's say there's like a gossip layer that allows like different infrastructure so you could have like a chorus a pigment and whatever like all validating under the same key so here the idea

would be that the the Dow would actually own uh the validator key right so like the the um the the Dow token holders would effectively be like the owners of the validator key and not like the infrastructure company so yeah there's lots of we need to do a lot a lot of technical research on but why why would you want to aggregate a valid it is behind a several key versus not just have multiple validators um well yeah I mean in in this case uh in this case like you know a lot of the value that would exist sort of like in the stack would is is the validator key that is that has a policy to give 30 of its commissions to

open like open source software um you're saying why not just have multiple validators doing the same thing yeah I mean like other validators can do the same thing yeah yeah because I think we want the idea I think the idea for us is for like we want to figure out how a dow can own a validator yeah yeah like all the way up and down the stack so like we as a company would just get paid to run soft servers but we wouldn't actually own the key it's just like we're an infrastructure provider to this Dow no I think I think it's a great idea and like to this so I think what we want in in Cosmos is to have like really high competition between validators to

do great work and usually at the human scale so not just that like the computer scale as in you know validating voting on blocks about the human scales and like participating in the process of making the network better and I think that uh one one good side of what you're proposing as I see it is that um actually like actually developing public goods is like super expensive at the human scale and it's like something you need to do for like a long time and you need to be quite reliable it's not an operation that you want to like spin up and tear down in in in six months and right now the current you know we have these institutions that kind of do this and kind of delegate to other people um to do it but I I think in in Cosmos the direction that we we want to take is

that everyone just participates in the governance system that we've built which actually works pretty well in terms of Delegation and that for it should just be validators doing doing core research and um it should if we could fix the economics of it it should be enough to sustain teams um competing on that level so I think I think there is a real potential here to sort of flesh out this idea it's pretty cool yeah I hope so maybe we can talk about some more of um off air um but this goes into like so one of the Innovations of of PBS is like right now we went over like the set of responsibilities that a validator does like gossiping transactions ordering transactions being sanctionable you know

like all these things and I think at the beginning of Cosmos it was absolutely essential for for um for validators to do all these things right but now I I think the kind of stuff that you're talking about is a more valuable use of validated resources right so they've just been they should really focus on serving their delegates and not on optimizing blocks right or running Mev shops I don't know if like I don't think like setting up an internal nav operation is like a good use of that commission that you take from the network via inflation or or block Rewards or whatever so I think we are sort of setting this direction but there is a technical component of it so if we want this like political objective that

we want a politically competitive validator set that actually takes like very very meaningful role in the development of the network it also means that we need to sort of streamline a lot of the stuff that is like core to their business which yeah you're talking about specialization talking about specialization yeah yeah so yeah validators need to well I mean this is this is something that you know I was talking to with I was talking with um like Daniel from from state fish and uh Teddy we were in in um in Korea and we had several conversations about like the the political role of of validators and you know they did a validator Commons workshop at nebula Summit the uh you know was focused a lot about around

um encouraging validators to like have sort of political opinions about like where the networks that they're delegating in uh or taking in uh should be heading in like yeah I think it does take a lot of time like even even just like me running one validator like you know I've missed some governance proposals you know like well you could see it in minscan it's it's all it's all apparent so there's like part of it that's already quite legible right you could tell per validator what they voted it or not but the interpretation of it like it varies a lot you know different networks validators like do you need to vote on every automated like uh like inflation adjustment to the different pools in osmosis like I don't know I think that's important and also there's a lot to be

considered behind a vote so even when they're when you do vote on something the question is why did you vote on it because you're high conviction or are you voting because you see the political momentum sort of swaying in that and you want to be on the right side of it yeah I think I think there's a lot of that I think the the there's a lot of that and we need to think about what what that does to our our ecosystem and well the the big threat here is that large validators have so much at stake that is it reasonable to expect them to not do that like when you think about like like a like a like a block table I don't mean to be calling out people yeah but these are these are these are these are good people that are bringing a lot of value to the ecosystem

in terms of bringing in institutional money that is going to secure the network right that's very very good but the question are they going to put that money on the line to move the ecosystem forward are they going to take that risk or even large large investors yeah you know are large investors going to step in when the ecosystem needs dramatic change to rejuvenate is worth potential right do you think that the one one solution would be to have like private voting for for validators I think the solution is to have small validators have more of a weight in voting uh implemented at the protocol level I think it could be the protocol level or even just uh at the at the feasibility level you know like if

you're a small validator right you deserve sort of a voice not just like on chain but probably probably off chain and there's probably going to be new sets of Institutions that are going to facilitate that um I think that um instead of there being sort of a a central body that is sort of um a statutory State body like you know like a foundation like a non-profit Foundation that sort of oversees developments of um of of sort of core infrastructure or of the network itself or even a development team I think it should it'd be interesting to see what kind of Institutions you can form whether just like institutions of validators right so like if you're like a core development team I think osmosis did some really

great innovation in the token economics and ensuring that the core team today doesn't have to be the core team tomorrow because who knows like they should have to compete for that role there should be an open market for that role um I think the ICF is also doing a lot of interesting work trying to diversify it's it's bored bring in more members bringing new ideas right but that's a statutory institution and swiss Swiss foundations are very slow they're meant that's the the one thing this is blockchain it moves very quickly yeah you know so is that the right institution I would argue no and not because they're not good people trying out good ideas but just because the structure of it is just doesn't fit with the Cadence necessary for evolution in in crypto so my hope is that you know um this is probably stealing an idea

from Jay that um you know most Cosmos chains will operate like republics and that the validators will be will be senators yeah I I think that's not an unlikely scenario yeah I yeah I think like validators I think as as more token holders come into the ecosystem like there will be less and less um and and also as token holders are abstracted away from the token right like with like liquidity token uh derivatives and like liquidity mining tokens that less and less people will participate in governance and

validators will basically be um making all making all the decisions effectively perhaps not like practically speaking anybody will be able to uh to vote but like you know there is this like issue with voter participation in in crypto but also like in our general public yeah there's nothing new yeah which is I said like how do you develop a political economy that's like very competitive amongst validators is is is the question how you create institutions that facilitate that competition yeah that I think cosmos had its at its rate of development has been a stage of like oh we're all friends we all get along until we don't and then it's like a big shock but actually like the whole point of Cosmos is for people to be Sovereign and to and to disagree but in a ways that is that is productive yeah and I

think there's like some interesting developments around sub-dials and things of that nature but yeah Dow Jones developing a lot of great tools yeah for this and we're seeing those protocols like some of the a which is essentially uh a bunch of companies sort of collaborating on like a bunch of dials it is a composition of inclusion and Strange Love and and a lot of other projects sort of contributing to that to that one protocol yeah yeah that's very cool well um we're we're going a little along here we're an hour and 12 minutes so maybe to wrap things up um what's uh what's on the roadmap for Mecca and working people found out more and maybe like if you have any research that you publish or oh right yes uh so you could read our open documentation we've kind of like opened it up I think we're going

to put it in the show notes where you could see sort of how to use the API and how to integrate it where can we get that because I was looking for that I was looking for docs but I couldn't really there's a no that we have a new website coming okay there's a new question before medie so um I know this is live so there's like not a lot of a chance to to capture the attention but before Medellin we will have a website that will have the open documentation but we could provide a link probably in the show notes or we'll we'll post it on our Twitter which of course is uh Mecca underscore t Pre-K you can follow us uh on Twitter and of course you can come see us um in in Medellin for the for the panel with uh sunny and Skip okay cool thanks

for doing this thanks for having me yeah what are you gonna do later what's your what Your merge night plans uh I'm gonna go sleep okay and feel like a human being again tomorrow because I got to go back to Lisbon um to uh to work with the team all Mecca Tech is coming in for uh Retreat cool well enjoy the rest of your week in Berlin then yeah cool thanks again thanks