Economic organisation of block-building | Barnabé Monnot (February 2023)

hi everyone it's great to be here again at a another Meetup and I hope we get to have them more often uh I'm Bali mono I'm a researcher at the robust incentives group it's a research team at the ethereum foundation that works mostly on mechanism design and things that have to do with the economics of the protocol and Beyond so the talk economic organization of blog building I've given thoughts of it previously so if you've got some of them maybe I will say the same thing again but I hope that you will take some maybe new elements tonight first hi uh just to try and engage who's in the audience who knows about the merge show of hands Notch yeah okay okay yeah okay who knows about Mev yeah okay who knows about math boost right nice who knows about PBS a bit less okay okay cool cool so yeah so there's like a general understanding of what I'm going to talk about but I will try to give a bit more detail and maybe give you a clearer picture of how all of these things sort of fit together foreign I'm going to start with a bit of definition [Music] something I've been talking about a lot lately is the idea of principle agent problem it's a very common pattern that you see in economic situations I took the definition from Wikipedia just read from it you have a principle agent problem anytime you have a conflict in interests and priorities that arises when one person or entity the agent takes actions on behalf of another person or entity the principle the problem is worse when there's a greater discrepancy of interest and information between the principal and the agent as well as when the principal lacks the means to punish the agent so incentives information punishment this feels crypto economic but we'll get to that path the deviation from the principle's interest by the agent is called agency costs so that's one definition it's a formal definition I got a second one recently at a workshop that we organized at Columbia University which is that decentralization is a b word it's really hard because of all these principal agent problems whenever you decentralize you have this situation that yeah you have to trust someone else to do something on your behalf and so today I want to talk about some of these problems why we appear and how they kind of fit together in a greater fiction so the first one Pap number one is what I call the protocol to propose a problem it's a lot of peace so there's a thing that the ethereum protocol it's not really anything tangible it's more like a specification or a set of rules by which the participants of the effort Network are able to communicate with one another so they all run nodes the nodes have the client code that is running and they're trying to arrive at this shared single truth of the state of The Ledger uh that's let's say the protocol little the ground truth or let's say the real physical reality is that there are nodes that are running this these they are the validators but do they run the protocol honestly right the protocol is just a set of rules validators could decide to shirk some of its rules and to do something else instead and so to ensure that the validators run the protocol as a protocol designer is meant it to be run there's a system of rewards and penalties that's to ensure compliance but is it enough and so here we have our first kind of Delegation which is ethereum protocol is delegating um the the mission of running a protocol honestly to a set of validators let's talk a bit more about how to become a validator you have to lock up 32 if in the deposit contract once you do so you get activated as a validator you're asked to perform so some consensus duties on behalf of a protocol one of them is proposing blocks so every 12 second a validator is sampled and is asked okay can you make a block that contains the latest consensus information as well as an execution payload and the second duty is attestation so with attestation validators are able to come to a single view of a state of a ledger and finalize that view this is something that's new to proof of Stack that didn't exist in in proof of work yeah so if Francesco was here tonight you would have heard a bit more about that part unfortunately he paid on this so you don't get to listen to it but hopefully next month we forget to do this again all the best to Francesco a bit more about the protocol rewards and the penalties you get a reward whenever you propose a Blog that comes from for instance the transaction fees in your execution payload but also the block rewards for the consensus duties you can also get penalties so the protocol figures out that you've done something wrong for instance you've been offline for too long you get inactivity leaked we call it and if you do something that's provably malicious you get slashed and so with this system of rewards and penalties we hope that the blockchain that eventually appears as a result of these validators doing their job somehow matches the specification that the protocol was expecting or at least the outcome that the protocol was expecting that means a blockchain that's live a blockchain that doesn't have safety failures and a blockchain that has as much throughput as possible yeah this is a plot I'll come back to it over because of the talk what the efferent protocol sees basically so you have these Network validators the ethereum protocol is able to provide rewards and penalties to these validators well it doesn't entirely see everything that the validators do right so there are some rules which validators are somewhat free to apply or on which there isn't really a consensus on that vfm protocol doesn't really have a purview to impose these rules one of them is block building so there are many ways that block can be made and we will see over the course of the talk that this has been exploited very heavily in in recent years before I get to that I want to talk a bit more about why the blocks are made in the first place so maybe what's the president of the ethereum protocol and to me it's really the protocol is trying to supply network resources it's trying to provide blog space so that users can run dabs and so that transactions can be executed so this is a very basic uh slide amount of gas in ethereum you have gas to meter how much resources are being used by usual transactions with transactions we can compute consume several types of resources one of them for instance is compute when you run a node and you have a transaction that's incoming you have to I don't know do maps sometimes when the transaction is adding money to a contract or multiplying something you have storage consumption if you need to store in the short-term memory State accesses or historical data and function calls and you also have bandwidth for instance the size of the block these are all resources that the network has to expand to to run the protocol payloads are just blocks of transactions they consume gas which is a single measure of resource use so even though we have all these different resources we only have one real measure to to compute it which is gas and we have a gas limit which is 30 million and so now I want to think about the validators not just as protocol agents that are trying to come to consensus but also as block producers so think of it this way maybe it's a bit of a shift in perspective the protocol is basically providing validators with some gas it says okay at that slot the other person is going to make the block on my behalf here's a bit of gas for you you're free to get user transactions and to spend that gas by making a Blog okay and so the validator is really some kind of marketplace that is helping the supply of gas provided diver protocol to meet some kind of user demand for that gas and what results from that interaction is a block of transaction so that interaction is not completely unconstrained first as I mentioned there is the gas limit so you can't borrow as many resources as you want from the protocol the protocol is trying to rate Limited but the new constraint we have since now almost two years is eip1559 so not only do you have a guess limit but also now the protocol mandates the transactions have to pay at least a reserve price that is called the base fee that reserve price we we've actually done a talk two years ago at a very Meetup to to go a bit more into these details but that reserve price is dynamically updated based on the demand and so if we see that the demand for Block space is high we know that we have to raise the price to price out more of a user so that we go back to the Target resource use and if the demand is too low we decrease the price so we try to let in more users to to use the gas so you can think of a reserve price like a congestion price a bit like uber when there's too many people clicking on the app to get an Uber they say okay now the price is rising and you have to decide if you still want to make that transaction or not and so that's EMP 1559 more recently you might have heard something about the multi-dimensional eip1559 so as I was mentioning earlier the gas is sort of a numerator it's like this one stick against which you measure everything else and so that means in particular that you have fixed relative prices between different resources it's for instance what we're saying today is that okay free units of execution of like three additions is kind of the same in terms of resource load as one byte of data that is kept by all the nodes forever it's very hard to make these kinds of of ratios so they were fixed a long time ago in the yellow paper with the gas costs of a different op codes but it's very inefficient in economics we kind of call that Central planning and we know that it leads to inefficient education where sometimes you have to trade off different units that might be not complementary that you might be able to substitute but you might be able to use at the same time so the idea of moving to a multi-dimensional gas so having more Dimensions to this Market is that now you can have floating relative prices and in particular you can move from this fixed budget constraint when you have to trade off between resource a like compute and resource B like storage to a more optimal allocation when you can break out that constraint and provide more of both the first instantiation of this multi-dimensional eip1559 is going to be the two-dimensional eip1559 something that we call the data I guess and something that will be rolled out with eib4844 show of hand who has heard of eap424 yeah you might know it under the term Proto Deng shouting this ID that okay now we're moving to this roll-up Centric roadmap so we have these l2s that are able to secure themselves from the ethereum protocol what these l2s need to consume as a resource is something that's called Data availability like we need to ensure that at some point in time there's common knowledge of the data that the rollup has been posting but the data has been made available for everyone to challenge it or to integrate it into their storage and so that data availability is actually not so difficult in physical terms to provide as a node you might have to download quite a bit of data but you don't really need even to store it permanently so we could for instance increase the amount of data availability that is provided without shortly the execution that the nodes are doing and so if we move to this two-dimensional pricing system we can let the market figure out how much of the allocation can be done and break and and have a more optimal provision of these resources and so now that's the same plot that was very a couple of slides ago where the ethereum protocol gives you not only gas but it gives you both exact gas with which you can do the regular execution as you've done it so far and the data gas that the roll ups can consume and so now you have a new class of users on your network which is the Roll-Ups and the Roll-Ups are coming to validators and they're saying I want to consume a bit of data I guess to secure myself and again you have a block that comes out of this interaction so that was a bit on the background so yeah why would you visit the first place why we care to run a blockchain why we care to run a protocol we really want to meet the demand of the user that can be other infrastructures like Roll-Ups and provide these resources for them as a network and so now I want to move to the second principle agent problem which is that in the cool service operation so in the course of validators making their blogs to meeting to meet the demand from users with the supply that is given to them by the protocol we observed that validators delegate this call function to other parties other entities that you may know as Builders and so now I want to go a bit more into why that happened it's a way of economic reasons why we saw this separation take place in the first place and and kind of where we're heading with that right so the ethereum protocol interest the execution its execution in the hands of validators but validators Outsource some of these functions to the builders so let's see why this is more of a historical view so how was block Construction done in proof of work in proof of work people would let's say miners will be the ones that are making the blocks the big transactions from the mempool and they make their blog that's kind of a naive view a less naive view is to say the miners pick the transactions from the mempool and they organize it in decreasing order of gas price that for instance incentivized people who really want to get to the top of the block to pay more the phenomenon that you might know as priority gas auctions why this happen is the top of the block is is kind of a privileged position for a person that wants to transact like you're the first to act upon the new state of the of a blockchain and so if you want to do things like Arbitrage or back running uh it's very profitable for you to have this first slot in the block and so once my nails figured out that oh there's actually a real demand for for this slot and the priority gas auctions are a fairly inefficient way at realizing that expression of preferences once we figure that out what happened is an auction system between block proposal and what we call Searchers so Searchers are specialized entities that find very valuable transactions that would be very valuable if they were executed first and and try to tell the miner look I have this bundle of transaction if you put it at the top I will pay you one if because I will make 1.1 if for instance and so you have sales like find these bundles that try to convince the miner to to take that bender and to include it at the top against payment for the for the magnet so what we see is the system it was known as the flashbots auction that was the first system that was deployed it wasn't the only one that was also Eden before the switch to proof of stake so we see that there is this auction between the Searchers and the Niners and behind you have a user still making the transaction in proof of stake we had to move to a different system the auction between the Searchers and the miners really worked because there was a fairly strong trust assumption between the miners and researchers so the Niners were mostly organized under mining pools the mining pools you could identify them they were reputable entities and so you could kind of pre-create a semi permission network over which the auction was running in proof of stake you have a long tail of solo validators who are people who are fairly interested so like people running a validator in their living room and so now you can't really trust them into the semi-permission network and so the mother that was found is to change from a model where you're only telling the miner what to put at the top of their block to just making the whole block on behalf of the block producer so the block producer and proof of stake is validator and so now in that system you see this supply chain of transaction or of blocks that is getting a little deeper so yeah the users making the transactions The Searchers coupling the transaction in profitable bundles the bundles being shared with the builders and then the builders auctioning off their whole block to the validators we are what we call the Builder API the economic idea behind this is that if there is a lot of value for instance in the bundle if you find a transaction that has a juicy Arbitrage or things that are bad like sandwiching you want to try and get that value as far Downstream as possible so you you'll be of course better if the user was getting to to recoup it that's another topic that we can also discuss later but here the ID is as long as you have a lot of competition at the Searcher layer another Builder layer most of the value that these agents are making from doing these profitable arbitrages has to be sent Downstream to World block proposal okay because researchers are competing to get into the Builder blocks and because the builders are competing to get their block chosen by the validator they have to give away as much of a value as possible so what you see is that the value of these transactions something that you might know is Mev flows Downstream all the way down to the to the blog proposal so I think this is a nice mental model and so what does block Building look like today in proof of stake so in proof of Stack you have two options you can of course do local block building that means you're running your own mempool uh you're listening to the transactions you're making the blocks yourself when it's your turn to make them but you can also summon these external blog building Network that is fairly complex and that yeah summons this external network of builders that connect to you this network is organized along several layers so between the builders and the validators you have this extra layer of what we call relays which are basically middlemen in this market so they are Brokers they take the bids from the builders and we validate that the bids are correct and if a bid is correct and if a proposal has made a commitment to using the bid the block is then revealed so think of a relays as Brokers of that transaction there needs to be some kind of trusted Escrow in the middle and that's the role of a really a bit of data this is a plot that I got from mavboost.pix which tracks the the state of that market immediately after the merge already something like 10 percent of a proposal we are signed up to get their blocks done by the external network of builders and since then I think today we are very close to 90 to 95 percent of the blogs made by external Builder that's I think quite impressive or significant in the sense that a core functional validator which is something that we could naively think okay validators make the block is actually not really the case to the most of validators are getting their blogs done by other entities the builders at the very beginning flashbots which is the research and development entity that mostly developed the map Boost network was very dominant over that network from their relay and from their Builder but the share I guess has dropped down a little bit with the operation of new relays such as ultrasound block native or agnostic the Builder Market has also had interesting Dynamics at the beginning we had mostly the vanilla Builders so people who are making good blocks themselves for people who were not using for people who are using math boost most of them would have their blogs made by the flashbots Builder but since then there's been quite a few more builders that are now very competitive and that are contesting this Market against one another like build the ox 69 Beaver build or multi Builder your names are always funny to say especially when you're giving like a serious conference okay but whatever risks so why do validators do this in the first page right they expect to get more money because you have the specialized entities the builders that are supposed to know how to make blocks better they are supposed to return to you more value than you could make as if you were making the blogs by yourself but that also means that you're at risk right whenever you delegate things to someone else they might not do the thing that you want them to do and in particular one risk and something that was very discussed uh in the in the recent months is the idea of censorship resistance so by definition vf1 protocol wants you to include the transaction as long as it pays the fee especially with vip1559 if your transaction is willing to pay at least the base fee and if there's room in the block you should be included that's kind of let's say rational to to have a neutral or creatively neutral protocol so the protocol proposal program told you that you know individual proposal might sometimes censor transactions let's say I have personal preferences I don't want to include Cash pass transaction for some reason but because validator set is expected to be very decentralized Casper knows that even if his transaction is not included by me he might get in in the next proposals block because you have many many different types of proposals in your validator set the ones proposals Outsource the production of their blog to entities like Builders which are much more centralized if you remember the plot I was showing you you could count on two hands how many big Builders you have in the market so once you have these Outsourcing you might have a problem right some Builders might decide well actually I don't want to include your transaction so details or relays they might be for instance registered entities they might have concerns legal concerns around the inclusion of sales and transactions and then if most of your network is sourcing blocks from these Builders or these relays you might find that some transactions are just not going through in your network so in ethereum well it was always the case that transactions would eventually be getting in but that means for instance if your transaction hurts someone's concern you might have to wait a little longer to to get in this is really an agency cost in the sense of this principle agent problem I pay a cost as someone who delegates from the agent not seeing eye to eye with my own preferences and so the solution is for instance the proposal could give a hint to the Builder could tell the Builder look I really want this transaction to go through so I will put that transaction in your blog or you can't build the blog for me this is something that's known as inclusion list I'm not going to talk about it too much but it's something that's been discussed possibly for instance for inclusion into the protocol this is a recent plot on this censorship so someone was tracking how many blocks were called offic compliant so how many blocks or how many Builders and relays we are declaring themselves to be compliant with the offic regulation it's been going down in recent time so I got that today it seems like 45 percent of the builders are enforcing of a compliance which means that on average if you're sending a transaction that is violating this list you might have to wait an extra block to to get in so in terms of waiting time it's not so bad but you do have half of a network that is enforcing that okay I affect me it's safe so I just wanted to end a bit on it's a widening the frame how do all these things fit into where the protocol is headed what other principle agent problems exist that we might be curious to take a look at so one of them is okay PBS is today expressed out of protocols so we have this map Boost network and I told you the relays we are trusted parties we validate the goods sold which is the block they act as Brokers but it's this enough right as a protocol we see this happen and this is all happening outside of the view of the efl protocol all this network is not something that the protocol knows anything about but shoot it right if if this is so systemic to the network is there a point at which we decide okay we need to do something about it and they maybe move some of that infrastructure into the protocol and so the idea is really the idea of boundaries so given that now we have this new network of Builders to which validators are delegating the function of making the block well should we metaphorically extend the boundary of a protocol so that we have some kind of control over the Builder maybe even some kind of system of rewards and penalties or enshrining some of the Dynamics of this Market into the core ethereum protocol that's something that has been quite discussed uh recently another question is there are much more principal agent problems out there so you have a one up here on the right that's the network of builders that Network itself is becoming more complicated we might have heard about Suave from a flashbots there are many other proposals to do things like payment for order flow or wallet side auctions such things it's getting much more deeper on that side so that's one of the principal agent problem another type of principal agent program is for instance liquid staking so you have State girls that don't stake at home don't stay directly themselves they give their tokens to a pool and the pool gives them back a liquid staking token and so then you have when it's some part of a network of validators that's not really controlled by the validators who are running your network that are that are controlled rather by the token holders of these liquids taking derivative that might induce some of these delegation problems once again a new type of Delegation problem that yeah if you are on Twitter you it's been also quite discussed they released their white paper a couple of days ago is eigen layer so the idea that stakers can restake their their Capital their if to to perform additional services and so they sign up to this eigen layer system they allow themselves to be slashed by conditions that the protocol the ethereum protocol is not enforcing but other agents are enforcing and they might get extraordinary ones but there's also extra risk so now you might have a misalignment between what the ethereum protocol expects of its validators and what eigen layer expects of the same set of people are popping up and I think it'd be very interesting to to have a more coherent view of that scene I wanted to I I made that slide very just like a couple of hours ago after reading the base announcement from coinbase another type of infrastructure that existed that wasn't pictured here is the Roll-Ups the Roll-Ups are L2 chains they pay their data availability fees to validate us who provide the data gas for them in exchange the validators are securing them so overall apps are in a sense delegating their security to the two validators and they are also deeply embedded in that web of of infrastructure so some of the rollups might get their blocks or patches built by the builders eigen layer has plans to release something called eigen da which would provide data availability to the rollups I couldn't make the arrow because when the plot would be too busy but yeah what I'm saying here is that the economic landscape of ethereum is becoming much more complex and I think it's worth taking time to understand how these relationships work how they evolve as well and what's really the role of the ethereum protocol in that in that system and I know I sound a bit like it's always sunny meme but yeah I I really think this is important and and it's happening very quickly and I and I want us all to to pay attention to to that thank you for listening I have more notes on the post on sub stack and you can also find me on Twitter [Applause] thank you for the talk it was very nice um at Devcon I felt like it was a bit of a elephant in the room um do we feel now like it's okay what is okay okay so let's see oh there's no time on this I think Defcon was like here I don't know I remember somebody posting that chart but it was much on a shorter time frame and it it was just like all increasing the red bones and so I yeah at that point I guess people were like okay this is something terrible is happening and it doesn't seem like it's stopping anytime soon it is true that it got better I don't think we can say okay this Victory like we've defeated that um so that's why people are talking about in protocol PBS that's why people are talking about inclusively so there is still a lot of food that's given to okay how do we Harden the network against these types of things how do we realign the incentives between the Builder and the proposals and the protocol um yeah so in my opinion we are getting stronger not weaker and not just stronger in the sense of okay the market of map boost is diversifying becoming more healthier becoming more contested which is good but also stronger in the research that we have and the kind of tools that we know are available to us and also maybe the traders that these tools have so there's still a lot to go but I would say yeah overall I'm fairly confident that we can get something nice uh so you talked about expanding the protocol to the basically maybe prescribing how blocks should be built can you give some examples on how that should be designed or would be designed so right so I guess I I don't know that we'll expand the protocol all the way to how the blocks should be built there are protocols out there that are much more prescriptive about this for instance you might have heard about this idea of fair ordering or receive time ordering uh so protocol like ekitas femis we try to come to consensus over when did the video transaction appear they're basically trying to do first come first serve but you know in a decentralized setting where where you don't have like these weird incentives to to game the queue let's say I'm not sure we'll ever see that in ethereum I think we care for a bit more generality it does seem like it works well for certain designs but for instance these protocols also seem like they could induce latency walls and over optimization of yeah basically latency and so in my opinion the local the ring will remain fairly unspecified at the protocol level what we could see is in protocol PBS which is basically getting rid of that layer of relays and getting rid of map boost and moving parts of that into the protocol so running the auction of a builder and proposals but where the auctioneer is no longer this out of protocol system but the protocol itself which could Harden the finger a bit more so yeah something else that was maybe I'm just ripping on this but something that was discussed was the idea of um encryption at the base layer so shutterized Beacon chain which closes chain is exploring very actively it was actually a proposal from uh well shutter but yeah yeah so I I'm not sure that we will see that at the base area I will do there is a lot of research on the idea of encrypted mempools and trying to get the encrypted transaction to execute eventually but I'm not sure what it will look like in the future just one question you mentioned shortly there I think AP 4844 I always get the number wrong um well if roll ups are like a thing to stay and then maybe some people are like some users have a choice where basically executed in the selection so doesn't that mean that most problems that you talked now there will be obsolete yeah that's a good question but I think Roblox will have the exact same problems a they're already building their sequences so the people who are making the blocks on the Roll-Ups to be decentralized and that means you sequencer could be a builder for instance if I want to execute a transaction that does across chain Arbitrage between let's say optimism and ethereum L1 or optimism and Cosmos or something and if I know that there's a builder which is making blocks on both sides it could be advantageous for that Builder to yeah to to sit in the middle and make blocks for both domains and so all of these problems of Delegation of block construction we could see them appear also at the roll-up level but could not affirm this site themselves like how much data availability they allocate to certain Roll-Ups if certain Roll-Ups are like more fair and others or maybe more deprivable I don't know how to say that I think this would be very hard again with the idea of let's say fuzzy idea of a credible neutrality unless you had like a credible signal that told you okay visual app is not playing fair I think you'll be very first it will be very difficult to get that signal but even if you did get that signal I'm not sure you would see that in protocol like some kind of rate limiting of people we don't like okay so when AAP 4844 when suing TM I actually don't know I haven't been following the development so much yeah going by set summer of 2023 is of course they are making the Rodney for us now hey um well first of all thank you so much for the talk it was amazing uh I have one question in if you had to pick one what would you say would be the biggest initiative that we could take to systemically improve Security in the ethereum ecosystem so that's a tough question so it's just like um so basically I went to law school and now blockchain is my life and uh I'm just really interested in cyber security because I um like I tasted Every Flavor of the blockchain so to speak from artists to real estate agents yeah and um I figured the way to stay in touch with all of them is to protect the infrastructure that they're operating on and well ethereum is obviously the best blockchain so I was wondering how how we could do that how we could keep the city safe so yeah I guess security is like very multi-dimensional so people say for instance okay no no count you you want to increase the node count because if you have more full nodes your blockchain is more secure I think that's part of it another part is we could be maybe a bit more explicit about the kind of failure modes that we as a community would say okay this is a hard no for us so for instance with proof of stake you have the ability to Fork people who don't play fast if you have like something like a 50 attack or 67 percent attacking improve of stake you have the ability to take the attackers out and reboot your network from there um it's been discussed as okay there is the possibility to recover from his attacks if they ever happen they don't depend only on the Node count they depend really on some kind of strategic planning to some extent by the community to say okay if we have a set of validators that are constantly censoring what do we do if we have a set of validators that are throttling the chain so that they can extract Monopoly pricing from the users what do we do all of these types of questions I think I actually know that some people are thinking about them like from a legal perspective and I could share some things with you if you're interested um but yeah they do get to be a sense of what other kind of properties you can guarantee let's say economically or cryptographically from your protocol and what kind of guarantees do you have to just say yeah when it gets to that point the certain layer has to kick in some kind of social layer or some kind of community layer and when it does how do we recover from the attack like how do we reboot the chain and what does it look like so maybe preparing a bit better for these contingencies could be something that hardens the security and maybe defining the cases where we think this is yeah important to do so I don't know if that answers your question but it it does but um how would you say that to someone who just installed their meta mask you know what I mean uh hey you mean like that so to them I would say don't share your seed phrase yeah so okay from the user side there's a couple of things like more right clients would be nice for instance so having the ability to not be trapped into the metamask default RPC that's that's one thing another thing that's more like in the economic realm I was discussing okay Mev flows from the user to the proposal basically but some of that Mev especially when it comes to things like sandwiching I don't know I mean the user shouldn't be losing it in the first place and so if you find protocols or mechanisms that allow the user to protect themselves against unfair extraction of the value that they provide to the network I think that will go a long way to making the yeah the network a bit more fair and secure from the user perspective well thank you sure all right um so for the idea of inclusion lists we talk about like forcibly you you force people to you force Builders to include your transactions in a block um and this is a preference that you as a validator may have do you foresee there being a market for this from the perspective it was like there is a value at which I wish to express that like I would be prepared to give up a fraction of each for this but if it's going to cost me um 0.1 of an eighth uh then that that might be something I'm more interested in and if that is the case like if we do see these happening um does this not bring back like being able to bribe your ways back into censorship Again by repeatedly paying for these things to occur yeah good question um so I can answer it from two perspective the first one is what you're describing is kind of where in math boost already there's this new Option when you're running MacBooks that says basically the mean beat that you have to be paid so you're saying I'm willing to get a Blog made on maybe halfway Builder if the blog returns me at least this value so something that people running May boost really want is okay sometimes you have like extremely juicy Mev blogs and you just don't want to lose out on them but these are fairly rare like the Mev distribution is heavily long tailed so most of the blogs don't return to you a lot of Mev compared to the default blog that you're making yourself and so you can see already that yes some proposals are making that trade-off between okay there's a toggle let's say between efficiency so how much value we're getting and censorship so yeah that's one way of seeing that people would be willing to make that trade-off a second thing is you don't necessarily there's a model of inclusion list where you don't have to make that choice yourself the model having released we have a protocol enforces it where you make a list not for yourself but for the next guy and so yeah now you for you it's cost less to to make the list I think we need to think a bit more about this because you'll be the first time in the protocol that you can as a proposal enforce constraints on the on the next proposal so these constraints are supposed to be free but my intuition is that there may be some attack vectors here but yeah this is more of a foot at this point cool thanks you may be answering one last question sure what's the biggest threat to ethereum oh these are good questions wow um well Meb is pretty threatening I mean it's threatening in the sense that it got so big so quickly and it made so many people basically profit their lips like oh there's always dollar amount that I can be making as social as a builder as a relay as a proposal um I think if we don't find good ways to contain it that could pollute the protocol fairly heavily so the idea that yeah users feel like when they are sending a transaction on ethereum they are not being executed fairly if it's a swap or if it's something else the idea that proposals are okay you put up some stake and then you're just getting all these extra rewards from basically doing nothing just running math boost so yeah that to me feels like it could make people feel like transacting and securing ethereum is yeah it's like the game is isn't just not fair that's yeah that to me is quite freaking I hope that answered your question all right let's have a short break of 10 minutes thanks again