MEV in eth2: introduction & open questions

Transcript

maybe we shouldn't hide the that one all right hi guys thank you all for coming this is more people than i think we expected um cool i'm alex and this is alejo we're both from flash bots research this title is a little bit different from the title in the schedule so we're gonna talk about both mev after one five five nine and after the merge so i'm even these two we'll run through both you know we don't have a lot of time but i think these are two key events that are happening in the next year or so uh in terms of milestones for ethereum and so it's important to think about mvp in that context um structure okay i'm just gonna do this all right very basic structure

i think first we'll just introduce what mev is and give uh explanation around that i don't think well it's it's good to remind ourselves of that um and then alejo will talk about the research that he published yesterday on the topic um yeah and then i'll talk about mv after the merge and then in the end we'll talk about flashbots and what we do and i guess how that relates to him okay all right thank you alex i think you can use this okay i'm gonna try that thank you alex thanks everybody for being here i will um so maybe it's a hot topic these days

but i will not assume that all of you have heard of it so i'm gonna try to give a brief introduction to the phenomenon um so this is a sort of a fairy tale day in the ethereum blockchain where we have trades going on arbitrage liquidations and all of these are good right arbitrage are good for keeping prices sensible right the prices are the same in different indexes liquidations help with the different protocols lending protocols be have have money in them and not go bankrupt so all this is like a happy path and it would be very nice and and clean if ethereum was just that but we know it's not just that

so once we come closer to it we see that we come to the dark forest this is a metaphor that has been used recently probably most of you heard about it and this is an alternative name for the ethereum member and so what goes on in the ethereum member um so say there's this uh bot here that sees an opportunity a price difference between two and decentralized exchanges and says okay i want to buy in one and sell in the other to make a profit and so the miner comes along and like wants to include that i might be missing some okay i'm missing a little bit of the um so but then other bots come along and say okay no hey i'm also want to

to do that arbitrage perhaps because i saw the same opportunity or perhaps because they saw this first spot trying to and take that value that was just lying there and and okay so the miner comes and says okay this this bot is paying me something uh to to include their transaction this other body is paying me more so what would i do okay so a miner sees this picture and says okay i'm gonna first pick uh this one because the one that pays me more and then why not i will also include these transactions even if they will probably revert because the operator traders there are trying to get make sure that they are going to make some profit by doing this so if the opportunity is gone there's no profit so this is gas that it's burned um so this of course is is bad these are

priority gas auctions because the one that pays the most gas and gets in right but it leads to a lot of like gas burn increased gas prices right this is bad for users because it's artificially increases the gas price and it's also it also blows the chain right because we get all these reverted transactions that are like meaningless and they could have been avoided right and so on so this is one example again some people are capturing some value here but this is a very inefficient process and another example here there's a an oracle update so as i said before there's some opportunity for liquidation in lending protocols sometimes prices change and then some positions become unhealthy and so the miner

i'm missing some of the things there but these are all say paying the same money so the miner has to choose at random which one they include so perhaps they pick this one but this liquidation doesn't go through because the oracle update is still not in then they randomly pick the the update which pays the thing the same thing and then now that the update the price update is it's been included there's another liquidation that does go in because now the the price is a new one and then another liquidation that doesn't work because the position was already liquidated so all of these are included in in the block some of this revert again bloating the chain and this is called back running and again leads to a lot of unnecessary uh

negative externality externalities for the network and again sorry i'm missing some of the the things here we had some issues with the animations so thanks alex for putting these two areas like google slices um all right so um it's actually more complicated than what i described this is like still already a happy path but say like the miner comes and sees this picture where there's an arbitrage bot offering to pay a lot and then liquidation like uh also happening there with an update and so the miner might actually say okay like why why would i let all these bots take all this value if in the end it's me who decides which transactions to include in the block right so i can say okay i'm going to make the arbitrage myself i'm going to like

profit from the benefit from this difference in in prices in the taxes i'm gonna buy here sell here and include the arb then i will include the oracle update because i need that to liquidate and i'm gonna do the liquidation myself and i don't care about these bots right because in the end like i can do whatever i want uh i'm the miner i decide what comes in the block i can also as a miner include these other transactions and again now we'll revert because the opportunities are no longer there because i minor took them so and this more bloating and the miner taking the the most out of this that's why we come to the definition of like um maximal extracted value formerly minor extractable value because again it's the block proposers uh

typically miners in today's proof-of-work ethereum who have the uh ultimate uh decision power and what gets in the uh in the chain what gets mine right so the minor extracted value or maximum expectable value is the value that can be permissionlessly extracted from the blockchain right so anybody can extract it miners nowadays are the ones that uh because they are the ones who decide what gets in the block are the ones that can actually extract this value if they want it um right and okay this is already a fairly bad picture again because of all the inefficiencies i was telling you about bloating of the chain and and so on and inflated gas prices that can get even

worse um so say there's a situation where we have like the chain with a block with a lot of value in there again perhaps there was a huge liquidation right and for you to have an idea of the values here we're talking hundreds of fifths per hour we've seen liquidations of 108 308 i think was a record one so it's a lot of money there and so the miner comes along and the standard behavior would be to mine on that block right this is what we expect miners to do to keep extending the longest chain but they can also say okay no wait if there's all this money i would rather mine on an earlier block right and and like uh capture all that value myself and i can even leave some

money on the table uh for the next miner to come along be to be incentivized to mine on my fork of the chain as opposed to the longest one and they could even rent gpu power specifically to get this this action going right so so this is problematic of course because it could thread consensus right there could be a like a race between miners so try to rewind the chain and see who gets all this value that's up there so these are called time bandit attacks based on chain reorgs and and again this this all lead to an impending mev crisis right this if this uh water common practice will be super problematic for the network and the more mev there is the more the chance the larger the chance that this

this happens right because the larger the incentive for miners to rewind so anyway this minor exposure value has been a a prevalent phenomenon the extraction of it recently and it's something we need to pay attention to and and be very careful with if you want to keep ethereum alive right so anyway this is like a an introduction to to mev um these are some numbers and i see already some from the team cringe for me using this slide this is from our explorer all these numbers are take them with a like a grain of salt because these are uh they vary a lot and they're very hard to estimate we

we in this this is the explorer we have flashbacks you can check in in that side and it's triggy i think it gives a solid um order of magnitude of the numbers but i can take the specific numbers with a grain of salt it's very tricky to to to get these numbers right so i have to go through all the protocols and and check what actual value is being started but this is a good order finally dude this year uh almost 800 million dollars 400 000 dollars in 24 hours this is the amount of value that's being extracted by by searchers but operators and so on um good so after that like quick intro to meb we can talk a little bit about what's

the future for mev so that sort of reflects what the situation today so i'm going to talk a little about eip1559 coming in two weeks now so pretty soon and then alex will tell you about mev and the merge um okay so uh what i'm gonna tell you about uh some of what what i'm gonna tell you about and some more uh you can check in in this post uh we published yesterday from flashos research as opposed in collaboration with christophe castle from nerman who's here in the audience so i'm going to check it out it has other things i'm going to just talk a little about some of the conclusions there and so and i have to introduce very briefly what 1559 is and then the the changes to a price mechanism to a

feed mechanism in case not everyone is familiar with that so current ethereum is uh essentially the user submit transactions with uh certain gas prices after 1.559 the it gets more complicated right so users will have to set a priority fee and a max fee and the protocol itself sets a base fee the basic is set in a way that if we so to compensate for the protocols usage right so if the network is is loaded and the the blocks are more than uh target full the base fee will go up and if the blocks are under under full like to the target block size the the base fee will go down

so again users set priority and a max fee and what they ultimately pay is the minimum between the base fee plus the priority fee and the max video max fee serves as a bound in case the base fee goes up suddenly so they don't have to pay an expected price right an unexpected price um okay so this is how uh one five five nine works and to our question like how mev changes after one five five nine uh the answer is not much uh so the one five five nine thing uh will hopefully improve ux apart from the burning of fifth um but there's no new mev and i put an asterisk there because there is some there are some like very border cases like there's now an uh there will be an uh

more eviction of transactions so if transactions pay a fee that's uh very low for the current base fee they get cleaned from the mempool so they get dropped but a sniper could be there um checking those transactions and storing them for later use and if the base fee later goes down they could replay those transactions and extract them maybe from them if there was any right so but again this is like a very but generally not no new big mev opportunities but still there are something to be said about one five five nine and maybe and i will talk a little bit about that now so the main point here i guess is that um eip1559 was designed and also most analysis around it

around the the incentives in it uh were done in a pre mev world not that there was no mvv so maybe it was there alone and there was even like some capturing or extraction of that mev but not to extend that it is prevalent today right so so this this design and analysis assume an option for transaction inclusion where users in ethereum what they want is to get their transactions included and just that nowadays we know that the picture is uh very different from that because users are bidding for other things in ethereum they're in particular ordering for billing for ordering of transactions paying for privacy for instance if you don't get to get if you don't want to to get your

transaction front run you might want to use like a service like flashbacks for um obscuring your transaction to the rest of the network and so on so a picture is much more complex than than and the the naive assumption of an inclusion auction might lead us to believe right and so what does this mean so if you look for instance at tim rough gardens analysis of the just a really good paper analyzing the the incentives around one five five nine and how how the the system will work and after how team will work after after the london fork so he defines three things that would be desirable from a a perspective like short term one

block type of horizon and the first one is myopic minor incentive compatible meaning that miners are incentivized to work according to the mechanism and second is user incentive compatible and this means that users are incentivized to beat their true value for for their transactions and this is what doesn't happen today in ethereum because you have to be strategic right you have to look at the gas price you have to think okay if i want my transaction included i'm gonna i want to pay as much but not so you're calculating all the time this is what 1559 aims to solve right so you don't have to think out what everybody else is doing for submitting your transaction and finally off chain agreement proof meaning that

miners and searchers are manufacturing users sorry are not um incentivized to make like off-chain agreements and site payments for miners to to include their transactions right so what he shows in the paper is that uh eap1509 is mmic is oc approved and it's uic only outside periods of high demand right so the ux improvement we expect to have outside of periods of rapid demand hikes right and we can live with that it's okay um and it's again a i don't know a compromise we accept interestingly in the same paper he considers an alternative proposal

so this is not 1.559 but something we could have done alternatively which is the tipless mechanism so no priority fee he shows in that paper that tipless is mmic is uic and it's only oca proof only outside periods of high demand so it's a trade-off right we could have run tipless and then we would have the ux improvement always but would not have the oca proof and that's okay it's fair that we chose one five five nine but the thing is in the pre mev world we didn't have these oca agreements the software agreements now with systems like flashbacks that we're gonna talk about a little bit towards the end of the talk we do have these oca agreements all the time all the time like searchers are

right we're running an auction for searchers trying to capture mev and that they're trying to order the transaction in a specific way and they're telling miners hey if you order these transactions in this way then you'll get some some of this uh some some payment and so all this auction that we know flashbots and other systems out there and actually are ocas right so right now we're running separate auctions at the same time and and and we have oca so we might think okay perhaps we could have uh work with a tipless mechanism and preserve this uh uic this ux improvements for users that are only concerned with transaction inclusion

and have that all the time and then run a parallel auction like flashbots for uh transaction ordering right so so the takeaways here is like okay if we think of mbb perhaps the considerations for design at the base layer are somewhat different um and to wrap up this this thing right now and i talked about inclusion privacy ordering this things are being auctioned in different parts of the blockchain some in the in the protocol level right and some like flashbots or other relays or dark pools right so um this just one side point here if you look at traditional finance markets like the different instruments are uh they have their own markets if you look at options they have one market

and if you look at the i don't know stocks they have different market because the instruments have different properties so in order to get a a an efficient market for each you need different designs for each one so um so this says okay we have all these uh auctions that are auctioning different goods and we this is a very hap hazard process which how we build different auctions so one big problem at flash roads where we're trying to do research on is to what extent we can have like a universal auction to express all these complex things that someone would eventually want from a blockchain right and this again is one of our main research uh problems and if anybody is interested welcome to come and work with us on this on this uh

very hard problems right so anyway and the takeaway here is that now that mev is commonplace we need earlier involvement with the aip process because some of these things are super relevant at the base layer and perhaps could be better solved at that layers in particular with ethereum two now we have bonds with the validators and so on so that's very has a lot of impact for mav so perhaps some of these things can be solved at the best layer and now alex will tell you about that and how mv interplays with ethereum too wait wait it's not done yet all right thank you that was great also i can see how like my face

cool all right so just like alejo is talking about some of his research this is some of the research we did on eth2 uh there's a post on hackmd and there's also a repository for some of the modeling we did um that's relevant that i'm going to talk about so first i'll talk briefly about the merge uh i assume that you guys are familiar with it with the ethereum community conference but but still right so the merge essentially is merging the beacon chain with the current ethereum one chain and ethereum moving from a proof of work to proof of stake consensus protocol um and so why this has already started here right well yes why this has already started here because the beacon chain was launched in december 2020 and has already accrued

validators that are depositing security deposits of 32 if and so what essentially happens at the merge which will happen sometime in the future is you're changing the consensus engine of ethereum from the proof of work engine to a proof of stake engine um kind of mid-air right which is why that image was very fitting um not as risky as you would imagine just because the beacon chain is already launched and as it accrues more validators uh the security of beacon chain uh supported consensus is just it's just better um and so by by saying this what essentially happens with what eth2 ends up looking like is this kind of two two chains in a way that advanced in lockstep so you still have the

existing ethereum chain that's stripped of its consensus power and then you have the beacon chain and so the beacon chain has beacon blocks and the ethereum chain still has eth1 blocks right and so in each beacon block you have a few objects in it some of them are consensus objects like votes from uh people that are part of the beacon chain so from validators the testers or the block beacon blog proposals but you also have the eth1 block in there that's added and so that's how the the eth1 block with the transactions as you know them today would be uh validated by the consensus the proof of stake consensus and so because again this this dual chain setup uh the the eth2 client that might be run by

um validators or generally the ethereum client looks like this kind of com like this dual component client um where one part of the client is the beacon client that validates beacon blocks and forms them and then does a bunch of other consensus roles and then the eth1 client um that assembles the eth1 blocks um and so why that's important why i put this here so this is this is not necessarily it but this this was shared a lot in ethereum community um is because you the person running these clients i mean the there could be two different entities running these clients so the person running the eth1 client can be different from the person running the beacon client this is going to be relevant to

what we'll talk about after um and so let's go briefly through transaction execution kind of or submission before the merge and after the merge so today users send their transaction their transaction is propagated through the eth1 p2p layer skin shows up in different client mempools so miners will maintain an eth1 client each client has its own mempool and so as they receive their transactions propagated through the prp network they have that in their mempool and they pick transactions from it the eth1 client forms the block and the block is added to the ethernet chain if the proof of work lottery is one and then that's how the blockchain moves forward after the merge the user still sends a transaction the transaction is still

propagated through the same peter p layer which is the fp2p the transaction then shows up in eth1 client mempool just as well the eth1 client still forms the block the only difference as a kind of you saw in the previous slide right is that the block is past the beacon client and the block is included in beacon block and then as the beacon block gets voted on by validators testers in that case it kind of gets reflected back to the ethernet chain [Music] all right so why am i kind of summing up all of this is just to kind of make the point that transaction or transaction ordering as we know it today still exists in eth2 so the mev that alejo talked about the one related to arbitrages and liquidations will stay very much exist

in that world right nothing changes at that level of how not only yeah the p2p layer which transactions are propagated but also the software that orders these transactions so software like like gef or like back i mean open ethereum i guess before um so let's talk about mev and how mev is affected by that i'm only gonna cover two broad points and then because i've given a version of this talk before i want to focus more on some of the open questions and things that might be interesting to to you guys if you've followed any talk before um if not yeah um all right so technical change

one thing that happens in eth2 there's a strong decrease in the rate of issuance of ethereum and proof-of-stake uh ethereum um so what that means concretely i mean the reason for that is there's no more block reward or anchor reward so you don't have that two if block award for every blog for minors and you don't have uncle rewards as well because you don't really have uncles at least in the same way and so why that's particularly relevant is because from the previous slide what i mentioned that mev as it exists today will still exist in eth2 and will have the same size which like you know alejo mentioned that 700 million figure um and so relative to the rewards that a validator or block proposal would get mev will be a larger share of that than

it is in each one um yeah and this is from a post from a testant sorry uh great you should check out the post um and sorry maybe i'll if i want to comment on this these are different phases of issuance of ethereum so this is ethereum two phase zero which is when the beacon chain was launched so there's more issuance there right because now ethereum issues rewards to validators on the beacon chain as well so it's slightly more uh eip155 is put here and i guess there's not a lot of impact to issuance at that point but that grows with transaction volume and then the end of mining here this is really like the merging boom you know that that happens um and so it's it's not

that's not exactly it because actually issuance in east 2 grows proportionately to the number of validators but it's still a lot less than in proof of work that's one of the the key advantages of of um proof of stake um and so why that's interesting because the share that mev represents is going to be higher and we didn't really cover this a lot but mev is also a reward that the blog proposer gets and that's something that flashbots kind of enables uh that alejo talked about as the ultimate actor that includes those transactions can it either front run the people that submit those transactions or submit them themselves and so what is meant by this what i'm trying to get at with all of this is just that mvp has a bigger deal

in eth2 because it represents a larger share of the reward that someone who's part of the consensus protocol is going to get so we ran some numbers on this and at the current level of validators that's changed now but right now it's about 195 000 uh so i put two to 200k uh mev can increase validate rewards relative to the base rewards that you get from the protocol by 667 percent which is a considerable amount um and so i guess what's important about this is mev will be i mean the extraction will be much more desirable in eth2 and people will seek it a lot more um and they'll have a much bigger reason to do so just because of

of its impact that it has on the revenue of the block broad proposers or consensus actors in that case the the validators um that's the first point that i think is important to remember so when you think of mav today and how i guess aggressive or how competitive the landscape is when you move to eth2 things are going to get i think a lot more intense all right second point uh a political change that impacts me is that miners are replaced by validators so i'm not i'm not a you know you guys are all aware of this but who are going to be the biggest validators they're going to be the largest heath holders and who are the largest heath holders

exchanges so you can see here already this is taken from beacon cha dot i n and you can see already kraken is in there i was like a nice 19 percent chunk and then binance is in there you can't see coinbase but coinbase is definitely in there as well um and so that's that's kind of interesting that exchanges are going to have that much power relative to to miners and mining pools because uh exchanges have other businesses if you compare it to mining pools mining pool's sole business is usually um you know to to host this this mining services for forever miners and distribute rewards and then maybe they have some kind of like staking service on top or they do some some trading on the side but

their main business is this kind of mining and exchange provides of course exchange services in the case of like cracking or binance they also provide usually custody services and that type of stuff and so what's interesting that might happen as a like dynamic is that the largest exchange can provide this kind of like full stack mev service so they own transaction flow from the users they have the capacity to extract that mev themselves and they also own a large chunk of transaction ordering because here when when when this says like cracking 20 percent this means that in 20 of the blocks reckon has the order it's not exactly that but um that's kind of worrying in some sense because right now the roles are separate and i

think that there's some equilibrium in in that market and when you have players that have that much power it can really encroach their dominion over over the the space and i think it's something that we all i don't know if you want to avoid but at least we want to be mindful of before before we kind of set those dynamics and then it's very hard to change um second thing related to that is the other largest actors are going to be meta pools so pools of pools like lido let's mention here although that chunk is definitely wrong uh but it doesn't matter so lighto is a pool that connects to other pools and offers on top of the the kind of staking service they offer they offer also staking

derivatives one thing that's interesting is ethereum is currently locked so when you stake it their withdrawals haven't been activated yet right but so what what lighter offers as one of the product their office they offer is liquid staked eath and so you're you're able to use the eve that staked and that's still locked um in in in the rest of the i guess defy and ethereum ecosystem and so now if you imagine that mev rewards kind of flow back to that token uh you you now have potentially some kind of defy ecosystem that springs up of mev based uh d5 products especially uh thinking about like future cash flows related to mev and so that can be really interesting and again it's something that's much more advanced than what we've seen today but it sounds

or at least it seems like this is a very natural direction and something that people have been talking about there's someone who posted something on twitter yesterday i think julian traversa along those lines i think people have been thinking more and more about mev cash flows as well so that's interesting i talked more about all this in the post and a lot more about our numbers as well i barely covered that oh yeah this is the liquid liquid thing from lido um all right i don't cool all right um so i want to more focus on like the open questions and for research real real quick because these are the the cool the cool things in my eyes um so one thing that i didn't talk about is that mev can increase the inequality

between validators so one of the good things with eth2 at least initially is that relative to east one about seven seven eight percent of the rewards that you would get as a uh a validator so compare validators to minors is just for good behavior of like submitting at the stations on time uh whereas on eth1 you would get most of the reward by being a block proposer so whenever you propose a block you get that huge two ef chunk but so what happens with mev is because mev mev's distribution differs per per block whoever proposes a particular block might get a lot of mev reward from it and so them might

amplify inequalities between validators because whoever gets whoever is lucky enough to propose a particular block that has a lot of mov is going to get a lot of rewards so if you think back to the to the pie charts that i mentioned before uh maybe some of the largest players are going to have outsized luck compared compared to others uh and so they're going to be able to even spin up more validators and you know stake even more 32 if uh security deposits and then that you get into those kind of rich get richer dynamics or this this matthew matthew effect um so it's something we want to research more because it's not as as obvious uh as as the intuition that i just gave and so this is says frp15 here because it's one of i guess our research proposals that we're working on

and um several people here are are working on i think alon might be in the room somewhere there you go um allen from from blocks taking and then uh nabe and casper from the ethereum foundation from the robust incentives group and while we already did some modeling of this that's the first link here and found some interesting results actually that are counter-intuitive i won't go into them now but that's one kind of open question that i think what we're worried about is that mev needs to because it's a bigger deal from what i mentioned before actually creates strong inequalities that that are unwanted another thing that we're looking into and that's part of the same research kind of track is the vectors of attacks that are spawned by movie and eth2 so alejo mentioned you know block reorg is a time-banned attack so

we're trying to understand how does that look um he needs two what are other attacks that someone might do one thing that could be mentioned for example is because in eth2 you know who proposes the the block in advance so one epoch in advance you could for example you know find that valid they're in particular and there's more kind of room for collusion given that you know them in advance and so there's stuff there the third kind of open research question um is on the blog proposal builder market separation so if you remember early on i talked about how uh the eth2 client that kind of dual client uh it can be two entities running them or two separate entities running them and so vitalik actually wrote uh an eth

research post about that a few weeks ago after we kind of brainstormed with him and it's actually applicable to other things where he envisioned a market where the ethon node runner assembles the best block that they can and they submit that to the beacon client uh and that also has applications for how how shards are submitted to the to the to the beacon block uh sorry shard blocks are submitted to the beacon chain um that's very interesting that's more of like a cryptoeconomic open question of how you would design this in the case of eth2 is how do you design against you know malicious [Music] validators are there kind of any opt-in slashing mechanisms can you piggyback off security deposits et cetera et cetera that's also very interesting and that's frp

that's the kind of 14. and then the last thing is thinking about alternative designs for mev mechanisms giving the new constraints of this system and that's something that tomas from from nethermine and and flashbots and euler uh is going to talk about later today uh not in this room but i think in the main room um and he's gonna present kind of alternative designs for uh what a system like flashbots might look like uh and so again thinking about what i mentioned before about block proposers being known in advance uh there's interesting things that you can do with that and so that's one of the designs that we have and then we also participate in this hackathon with lido and nethermine and we kind of built a

proof of concept of this kind of dual climb model but that's enabled for for mav so that's that's it in terms of a few open questions and for your research all of those uh all of those are public on on github and anyone can participate in research they're already kind of on their way a little bit so right now might not be the right time but generally we always have our open research tracks there and we have contributors that are working with us and then we write together and um kind of work together it's nice um oh yeah so i'll pass it back to alejo to just talk about flashbacks a little bit and then we can take questions i think we'll have a bunch of time for that i don't think so oh you know four more minutes moderator

oh can we do uh questions offline and and complete this four minutes okay thanks yeah um so yeah i want to say a couple of words out flashbots the organization and because we haven't mentioned we haven't mentioned that in passing but perhaps it's uh unclear so flashbots is a sorry it's a research and development organization working on mediating the negative externalities of movie extraction so our goals are up there illuminate the dark forest democratize me extraction distribute benefits so what we want is essentially to this so we believe mav is essentially unavoidable in stateful blockchains so the idea is to minimize

the negative externalities that its extraction has for the rest of the participants in the network right and we have several areas one is the auction this uh thing made more efficient for for extraction so that again we don't have this races i was mentioning before we have a data part where we collect some measures on mev which i showed before also and the research side alex talked a little bit about the frps we have grants so if you're interested in this please come talk to us or you can check our flashbots.net github i'm going to say just a couple of words on flashbots ethics so right now 85 percent of the miners are running our software and we have a as part of the auction a fork of get

called mvvgs um which essentially is a new shelling point right for the network because we could push an update to miners that they could passively adopt and collude by just passively adopting a software update right so as an organization we have to uh think about this and and and think about the consequences of this because again we want we don't want to harm the network we want to be super careful and transparent about this so ethics is a very important point for us so this question i was saying before can we help miners collude and to what extent for instance one hour question we tackle in the post i mentioned is can flashbots defeat the purpose of 1.559 we did some data analysis here uh

can we propose a mechanism or why we could keep the low the base fee low and it's unclear whether this will be good for miners in the in the midterm anyway i encourage you to check the post for more on this and and finally i wanted to touch upon one question that was uh making the runs recently where we would support reorg so we make mev extraction more efficient with these bundles where users can beat for transaction ordering order sets of transactions essentially so it was a question will we support this time-banded thing where we make it easier for miners to rewind the chain upon request and payment so the answer is no and i encourage you to read that post we

we posted last week on this we believe rewards are negative for the network again because what i talk about threading consensus and making finality impossible and so we are not building a client that does this we think it's a it's a it's a thing that's important we research and we have to pay attention uh but it's something we ultimately don't really want in the network so i wanted to clarify this point um so yeah i think with this yes we are done so thank you so much sorry about not leaving time for questions i think but we are here alex and me so if anybody has questions we're happy to take them offline you