 Having me so great to be here. All right, everybody. Hello, hello. It's so good to see everyone. I haven't been in a stage like this in about three years since DevCon Prague. Back when I was an obscure PhD student and giving a talk in a similarly sized room that no one showed up for. So very emotional to be standing in front of you all here as the steward of flashbots and looking at how much has happened in our industry in the last three years. I also am a PhD student at Cornell, but this talk is in my capacity as flashbots. So today I'm going to be taking you all on a tour through the dark forest. And this is actually the dark forest where I spend my time, where I live, right? So I know I'm taking the meme a little literally, but I'll also show you some pictures I've taken in my journey over these last three years, finding MEV and really trying to fight for user fairness. So I learned in teacher school that some people only learn well with an outline. So we're going to talk about three things today. The first one is where did MEV come from? Why are we here? Second one is where are we today? So I'm going to talk a lot about flashbots and the choices we've made as a collective, how they relate to things like the base layer of Ethereum, security and fairness for users as well. And I'm going to talk about where do we go? So what is the goal for MEV? What is flashbots doing? This is a question I get a lot from the community and we're going to start to answer it today. And how can we all do it together as an open source, as an Ethereum community to really build these systems for the next trillion? I'm not kidding about that. So this all started for me with this academic paper I wrote at Cornell called Flashboys 2.0 where we looked at the market dynamics of several DEXs that were popular at the time around 2018, 2019. And we looked at basically the game that bots were playing to arbitrage these decentralized exchanges and how they related to things like user fairness and consensus security using extensive measurement infrastructure. So if you're interested in that and you haven't read that paper yet, I highly recommend you check it out. But in that paper, we define this term called MEV. Maybe you've seen some talks on MEV at DevCon here. So we define this as saying OO fees or ordering optimization fees represent one case of a more general value we call minor extractable value or MEV. And that refers to the total amount of ether miners can extract from manipulation of transactions within a given timeframe, which may include multiple blocks worth of transactions. So there's a sort of duality to MEV that we've been exploring and pointed out basically after this paper. We're on the one hand, MEV is this financial potential energy that runs through everything we build, every system we build. And on the other hand, it has risks. So what we said in the Flashboys paper was MEV poses concrete, measurable, consensus layer, security risks. And we show that MEV poses a realistic threat to Ethereum today. That was four years ago. So today it's even more so. So how do we get to a place where we preserve the decentralization of these systems we're building? Where we make sure that this MEV is used for good and not centralization. That it doesn't entrench actors into our systems, which are not aligned with user fairness. How do we make sure these systems actually work for the people that are using them? Well, there's one simple answer to that. And this photo is actually from a hike I took after DevCon Osaka. Again, all these have like a little emotional story for me. So I'm kind of indulging myself here. But we do that by playing the long game. So this is what I spend my entire time, kind of for the last three years, obsessing about. What is the long game of MEV? And how do we get out of a world where we're all captured by Citadel? Where we're all captured by Robin Hood's new crypto wallet? Where we're all captured by MetaMask? Where we're all captured by FlashBots or a centralized builder? This is the world we need to get away from and we need to play the very, very long game here. So what am I doing about this? Well, one thing is stewarding this project called FlashBots, which we announced in this blog post called Front Running the MEV Crisis. And it has three goals. Number one, illuminate the dark forest. MEV, when we started, was an opaque, obscure, niche area that no searchers wanted to really talk about publicly. You couldn't get the kind of data in the FlashBots 2.0 paper without doing months of academic research. And that's not good, right? We need to communicate to the community, to the users, and to everyone exactly what is going on in these systems and exactly how it affects them. So that's illuminating the dark forest. The second thing we need to do is democratize MEV extraction. So if we wanna preserve the decentralization of the systems we're building, we need to make sure that our validators aren't these vertically integrated monoliths that essentially operate as hedge funds. That would be basically a worst case scenario for our network. It would be a direct capture opportunity for entities like Goldman Sachs, Citadel, Robin Hood, et cetera, who you already see moving in this direction on Ethereum today, right? So we need to make sure that MEV extraction and the benefits of MEV extraction are available to all. And the last one is to distribute MEV benefits. So it should not be the case that only the users or only the validators are benefiting from this concept. All right, let's talk about where we are today. This is a picture from West Virginia, also a very dark forest, it was cold there. So now we're kind of on the heels of the Ethereum Merch, one of the largest, most massive system upgrades we've ever seen, and FlashBots has been working towards decentralizing itself. So before what we had was a vertically integrated builder where it would plug directly into GET itself and basically build all blocks. At that point we were building upwards of 90% of MEV laid in Ethereum blocks and we decided to democratize this further by splitting this role into several interfaces with many participants. And this was a direct kind of attack on centralized buildings, builders, sorry, to let new block builders and new entities that wanted to participate in MEV have a direct channel to connect rather than needing a high upstart kind of legally heavy world the way we see in TradFi PFOF, which I consider highly suboptimal. So what do I mean by decentralization? So this picture is actually a perfect example of it. What we need to do to decentralize is build many, many parties with many, many utility functions that work on the system together and that cooperate. And these parties need to have meaningfully economically distributed power. So that's what we're working on. After the merge, the adoption of this system has currently reached about 56% of all kind of Ethereum blocks and FlashBots dominance currently on this system is at 81%. So a number we're kind of actively looking at here. And you can see that you get a quite substantial validator payment boost from running this system. So generally up to like 300% of the fees you'd make otherwise, but often spiking as high as even 80 ETH in some blocks. Okay, so that's great Phil, but how is that illuminating? How are we doing today on these three goals? Illuminate, democratize, distribute and what are we doing here? So the first thing we're doing at FlashBots is we're aggressively committed to open data. We believe open data is the only way to escape centralization and get to a decentralized fair universe. In that regard, we've been pioneering much of the data used by a lot of the ecosystem on MEV. It's very hard to read an MEV article without seeing FlashBots data quoted therein and we provide open data on both public goods, blockchain transactions that do not flow through us as well as transactions flowing through our node. Not only are we a leader in this space ourselves, but we also have inspired many other parties working in the MEV space to start by releasing transparency dashboards, which was unthinkable three years ago. So several other companies that are doing this include Western Gate, Jump, Skip, zero MEV and much more inspired by the illumination culture of our Explorer. And we're taking this past the merge with rich public data from all the centralized infrastructure we're running, as well as dashboards where people can see how this infrastructure is kind of progressing. And we're going to continue this culture of open data and illumination. Through this talk, through proof of stake and through any future we're in. One last thing I want to say, and this is the biggest font I used in the entire slide, so this is important. You cannot illuminate without research. FlashBots is committed to doing research as a first class citizen in its activities and having a dual research and development engine. You can see that in DevCon here. You can see the representation of FlashBots researchers both in the audience and the number of talks they've provided. I recommend looking at the schedule that we've listed on our forum with several amazing talks there. And it's extremely hard to look at an MEV event or an MEV article or do MEV research without using this public goods research that we've been doing. All right, Phil. That's great. You guys are illuminating. What about democratization? So democratization is extremely important to me. That means that every user has to have equal access to the system that we are building here of MEV. And on that perspective, much like my friend, the little frog, I believe we're a long way from home. So I want to address kind of the elephant in the room. This is one of the memes I've heard a lot at the hallway track of this conference, which is OFEC. So we run a lot of centralized infrastructure in our systems today. That is not the long term plan at all, but today we do. And we do kind of comply with US regulations on OFEC, which has become a meme. So this threshold actually crossed 51% today. What does this mean? This kind of shows two things. Number one, we've been very successful in our mission to democratize MEV, almost too successful. We've almost grown too quickly to outpace and build these competitive ecosystems we're striving towards. So the next thing we need to do is ensure as much global cooperation on these systems and as much diversity of participation with different utility functions as we can. If you look at what the Ethereum blockchain looks like today, this is a picture from last night of OFEC compatible blocks. You can kind of get a scope of the situation. So the reality is today there are not many transactions that are actually under these regulations, as well as you have about half of blocks complying with them. What this means is you may generally have to wait 15 to 30 seconds longer for the inclusion of a sanctioned transaction. This is not the same as a 51% attack on ETH because it doesn't operate at the consensus layer. That being said, it is concerning, right? If this number gets closer to the full proportion of ETH block power, that would be kind of upsetting for the ability of a lot of parties to get transactions included and also for Ethereum to operate as a system and as a foundation. So how do we get out of this and make sure that doesn't happen? Well, there's one thing from me as flashbots that I'm gonna say, and then I'm gonna go into really, how do we get out of this on a technical level? So what I have to say as flashbots is we do not intend to impose our utility functions or our legal risk analyses on the Ethereum network, full stop. That stands against everything I believe in. It stands against the last seven years of my participation in this ecosystem and the last 10 years of my participation in cryptocurrency in general. I'm here for censorship resistance. Everyone who knows me knows that. And I've also researched this issue long before OFAC. So how do we get out of here? We believe one key tool is open source. So open sourcing as much of our infrastructure as possible is the only way we get out of this centralized builder morass, right? We need to make sure that people can easily build this technology, that the barriers to entry are low and almost non-existent. If we wanna have a diversity of participants, a diversity of competition and meaningfully economically decentralized power in the systems we're building. So this is something we're actively working towards here. So this first graph is an adoption graph. You can see on the bottom basically the adoption in purple of the MEV boost system on ETH blocks. And it's steadily increasing in adoption as we reach 50%. But if you look at who's actually building these blocks, where that bright purple bar is flashbots here, you can see a number of new industry players are coming in to building blocks and as well as relaying them. We have several relays as I showed you on a previous slide. And this is to us key to democratization. So in our opinion, the relay is not a piece of software that will exist in two or three years. It will be replaced by PBS or more sophisticated kind of interfaces. On the other hand, the builder is extremely concerning to avoid centralization in because of the entrenched incentives in that level of the stack. So how do we avoid centralization? Open sourcing is one way. Supporting our competition is another. And a third way is making sure we lobby aggressively the entire ecosystem to avoid entrenching centralized choices like private or exclusive order flow into the way they do business because these will destroy the system that we love as a whole. Okay Phil, so how do we get out of here? Well, like any forest, when you're in the middle of a forest and it's dark, the only way out is through. So come with me on a journey and let me take you through where we are today. So the first thing we're doing is on the left. We're supporting not only Paradigm as our partners but many others who are asking these hard social and legal questions about the base layer, about validation, about how this impacts everything we're building. We're supporting them as much as we can with discussions with engagement and in many other ways. The other thing is on the top right, you can see a part of a tweet storm of mine where I talk about how to actually achieve censorship resistance and neutrality through decentralization, through geographic distributions of the preferences of different utility functions. I don't think you can pressure anyone validator to be decentralized and I think that will ultimately make you more centralized. If you knock on Coinbase's door and you say guys, you have to shut down now or you have to like mine transactions that will get your public company sanctioned by its relevant authorities, that's not how you build a maximally robust system. That's how you push your system to the edges. This is MEV for the next trillion. We're trying to go to the whole world, the entire financial value of the entire world here. So we are absolutely trying to go beyond the preferences and utility functions of any single player. The other thing we're doing is these are research posts you can read from Vitalik and our team that we've been engaging on for a while on how do we bring back the status quo of proof of work style censorship resistance where the validators have agency on choosing only builders that are not censoring while maintaining the revenue that they can see from MEV optimization. So we don't wanna make validators choose MEV optimization or censoring. We want them to be on whatever point on that trade off point they wanna be at and we wanna incubate as many points on this graph as we can. And how do we do that? We do that through research. Again, research is absolutely key and we cannot get out without it. Alrighty, so much like Tree Roots, the last step is to distribute this MEV to our entire community organism in this room. So here is actually the part we've done the least work on but the part we're really gonna focus on next. So we do have this product called Flashbots Protect which you can plug into your MetaMask and essentially get the benefits of private MEV auctions and extend these to all normal retail kind of users of blockchain so that you don't need to be a sophisticated party to access the benefits of what we're building at Flashbots. But we have a much, much more richer set of features to come on this front and I'm gonna talk about that next. The bottom line with distributed is we've been too slow. Like I'm not gonna lie with you, we must do more here. And I think the other two have kind of taken priority but it's time to kind of make sure we really give users the benefits of this MEV. So my guess is in a year, people are gonna be happy to be sandwiched because they're gonna be saving so much money on their transactions. So don't call it a sandwich attack. Call it a sandwich and let's make sure it doesn't lead to a centralized builder world. All right, let's talk more about walking out of this forest over here. So where are we going? So this is how I see the MEV marketplace as a user validator supply demand optimization problem where the user wants to minimize the fees plus MEV, basically the rent it's offering to the system and the validator wants to maximize this. Basically the commodity it's offering to resolve this potential financial energy of our universe into a single collapsed discreet state. So what do we need to do in this market? We need to make it more private. Privacy is the key here. You can't have freedom without privacy. We need it to be more decentralized, economically meaningfully decentralized. We need it to be more competitive geographically so one actor cannot dominate the preferences of the system. We need more user tooling so users can be sophisticated on their side of the curve. We need a market that's more robust, efficient and geographically diverse. To me, this is self-explanatory. If we're not building towards these three goals, we're going to fail. And we don't wanna have externalities on the Ethereum network, corruption of the validator set, changes of incentives and proof of work or any of these other kind of externalities of MEV. And we wanna maintain the success of flashbots in democratizing MEV to the searcher community already and prevent first party deals between searchers, builders, relays and other actors in the system. These are best done when you communicate on a technical, not a political API and that's what we're aiming to build. So WTF do we do next at flashbots? Well, we're gonna decentralize the crap out of this thing and that's the answer. So we are here to avoid entrenching centralization at every step. This is my commitment to you as long as I am here in this community I will be working whatever the status quo is to further decentralize power in MEV in an iterative way. And I expect every single person in this room to hold me to account. Okay, now you can clap, sorry. So we're not gonna end up owned by Citadel. No, no, I'm not done. Sorry, sorry, just a little applause. So what are we doing at flashbots? This is the question every community member asked me and today I'm gonna answer it for you. And please kind of follow up with me and start the conversation. We're building a project called codename SWAV. It's something we've been working on for kind of the last almost a year now and that we're gonna be releasing iteratively to the community and building towards further. It is not an end state but an initial state to start the conversation about how we decentralize this existing ecosystem that flashbots has built to all of our friends, all of our competitors and all the other projects in the space. So what is SWAV? SWAV is an MEV aware privacy first encrypted mem pool for users and wallets. So SWAV will make the maximum leverage of a user's economic value to force the system to provide them fairness essentially and vice versa as a negotiation tactic in this market. It features progressive decentralization. So along with the iterative steps on our centralized infrastructure that we're gonna take towards SWAV, we're also gonna have a very clear decentralized roadmap that we expect you again to keep us to account on. And for our friends at rollups, L2s, any other whatever you wanna call them, side chains, multi-chain, we're gonna give those actors a turnkey decentralized block builder for both their validators to maximize their revenue and for their users to maximize the output of their economic transactions without entrenching further centralization into the rollup or into the other chain. So this is very, very important. Cooperation without centralization is what we're doing here. I'll give you some more keywords about what we're building and spend a little bit of time kind of ranting about each one of these. So first is a fully decentralized block builder. We stand against centralization in the block building market and I know that may seem inconsistent with our previous views, but it is my belief that the only way to achieve decentralization is to front-run the MEV crisis and credibly decentralize every modular component of it. This is gonna be a fully open source system that's gonna be developed in the open. So starting in about a week, we'll start releasing information, specs, possibly even prototypes and code. We'll see two weeks TM, right? Like for users to play with and we want you to come discuss with us. That's what decentralized means. If you consider yourself a competitor of us, we want you to come to the table and we wanna build this together because we really believe truly that the best outcome for the space is if we're all doing this together rather than fighting each other in zero-sum games. It's gonna be ETH-native, so that means aligned with the ETH validator set and intended not to corrupt the validator set of Ethereum or impose things like undue token incentives or other centralization risks on the EVM layer and it's gonna be fully EVM compatible to kind of bring the benefits of all these EVM chains and their interconnectivity to MEV. We're gonna provide optimal user execution that harnesses MEV. So we're gonna use MEV as this decentralized engine in the long-term to make sure the users get really the best execution on their trades in a way that's really gonna make TradFi look embarrassing very shortly in my guess. We're gonna offer full compatibility with FlashBots today, so many of you may search on FlashBots, may relay blocks to FlashBots, may do whatever. And we wanna have full compatibility with that system going forward. We want cross slash multi-chain support. So this one is kind of self-explanatory. I mean, everyone asks us, are you going cross-chain? Are you going multi-chain? Come on, what do you guys think of course? Like, yes, but also ETH-native and ETH-aligned. We think this is the key combination for these systems in the future, so they serve the whole world, not just one area or one set of clients or people. Last thing is we're gonna maximize competition in geographic diversity. So this is the most important point that I kind of want everyone to leave today with, is we don't get out of the dark forest unless we have competition and geographic diversity, and unless both of those are really meaningful. So we can't build systems, Solana, that for example, privilege certain economic zones, or require you to have low latency to an existing set of infrastructures because then we lose out on the benefit of these global utility functions and the amazing cooperation that these utility functions enable. So every step we're gonna take is gonna be intended to maximize this competition. We're gonna enable open order flow for Ethereum's future. What is open order flow? Open order flow means no first-party deals between wallets, builders, and or validators, because we believe that these first-party deals in TradFi have led users to have absolutely horrible execution and have entrenched extremely predatory rent extractors, rent extractors if you will. And the last thing is, we're gonna have programmable privacy. So how do you achieve all this? How do you give users back their bargaining power? They need to have control over what information they're giving, and that only comes through privacy. It's one tool that we need to get out of the situation we're in, although it's not quite enough by itself for reasons that will become clear over time. So this to me is the genesis of a new phase of MEV, of new cooperative positive sum phase of MEV, to get us out of this zero-sum negative centralizing ecosystem that we're headed towards. Even though we have avoided vertical integration, we now need to avoid builder centralization. So that will be the next job of FlashBalls. I want you all to help. Please, please, please, please, please. Please, every person in this room, let's, you know, all these photos are from my hikes, and when you're on a hike, the best thing you can do is get to the top, because then it's all downhill from there. It's all nice and easy. The cool breeze is flowing on your back. So I want to hike with you all. We're gonna have to climb the mountain first. I'm sorry. It's not gonna be easy, but we're gonna do it. We're here for years and years to do this together. Especially if you're a user, a wallet, or a rollup, we want to be talking to you now about centralization in your system and the rest of the stack. We're hosting a roundtable on this later today, so please reach out with an application if you're interested in that. Lastly, please check out our forum where we make all these discussions and come troll the crap out of us about OFAC, centralization, competition, unfairness. Just hammer us, you know, please, please. So we really want you to participate in this because you can't meaningfully diversify power without diversifying participation. So we look forward to talking to you all and hopefully this makes you feel a little better about us not being like the evil Goldman Sachs Borg that censors us all. Thank you. That was awesome. We have time for one or two questions. Make them spicy. Yeah, there is a question here. One and two. Hey, thanks Phil, great talk. So I think talking with the community recently, one sort of trust issue that people have with Flashbots is you do all this great open source work and it's not quite clear what's like the end game, what is the way you make this into, you know, something sustainable, profitable for Flashbots. Obviously, there's a research collective, there's also a company or foundation or some sort of entity there. So can you talk to people about like what is Flashbots, the company or organization trying to build and how do you see yourself integrating with that ecosystem in a way where like others can also thrive? The truthful answer, and this doesn't satisfy people internally either is we don't know yet. We think it's too early in this market and there are too many kind of variables and competitors to make a call without entrenching something that further provides centralization in the long term. So we're intentionally avoiding that. People ask like, what do you pitch the VCs? Why do they invest in you? Well, we pitch them like, look guys, you own ETH, we're gonna solve this problem for you and like do you want your ETH bags that are worth 100X more than us to go to zero or not? And yes, we also pitch them that we do wanna be sustainable because as a mission-driven organization, if we're not reasonable and sustainable, we won't be able to continue operating. So the more specific answer is I think there's three things we still need to do at Flashbots and they were all on that slide. So one of them is to decentralize ourselves completely and meaningfully and economically. One of them is to enable benefit to be redistributed to users much more efficiently than it is today. And the third one is to enable support for any zone or trust domain in the world that wants to use Flashbots without harming the centralization of Ethereum. Those goals are still open research questions. And once we've answered that, we'll look into being sustainable and reasonable and monetizing. I realize it's not a satisfying answer but personally I would rather die right now than entrench something shitty. So it's kind of what we're gonna be doing. We're also gonna announce much more about this in the next month or two because I understand this is a question people have. It's a question people have internally as well. So totally reasonable and totally understood. Hi, so I was wondering, how do you think can small builders become competitive with the big ones? Because it does seem that order flow is a very strong centralizing force, right? And if you just have more orders, you can build more profitable blocks. So this is what decentralizing the builder role is about. In the future, I don't think there will be such a thing as a monolithic builder that builds the whole block. That's kind of the future we're working away from. What we wanna build in the future is much closer to a network of searchers that eventually chaotically kind of produced this block hash output. And in that case, any economically meaningful information that you have in any possible strategy in any way can be expressed into the system without kind of ruining its decentralization properties. So I agree with you, open order flow, or sorry, closed order flow does make this whole thing like much more centralizing. That's why we wanna get rid of it and that's why we wanna move into systems where any builder, no matter how big or small, can access the order flow of the biggest parties in the space, MetaMask, Coinbase, whatever else. Like that would be how you get out of that equilibrium. You make the order flow democratic as well as the MEV extraction API. And that is what we're working on, so. Okay, thank you very much, Philip. It was an amazing and awesome talk. Thank you so much for having me.