133 - Flashbots Saves Crypto with Stephane Gosselin

Welcome to Bankless, where we explore the frontier of internet money and internet finance. This is how to get started, how to get better, how to front run the opportunity. This is Ryan Sean Adams. I'm here with David Hoffman, and we're here to help you become more bankless. Guys, great episode for you today on the topic of flashbots, on the topic of MEV, on the topic of crypto dystopia versus crypto utopia, and how we propagate the utopia. This is an insanely important topic to cover. We're gonna talk about a few things, a few things to look for during this episode. Number one, why this thing called MEV, that is maximum extractable value, why it poses an existential risk to everything we are building in crypto. We are on this knife's edge between dystopia and utopia. Number two, what the world looks like if we ignore the MEV problem. We cannot afford to ignore it. Number three, the blockchain supply chain. What actually is that? How MEV can increase your ETH staking rewards by something like 6%?

Crazy. That's a high number.

Stefan does the map in the episode is a very high number. Number four, how all this technology is getting embedded in the Ethereum protocol long term in something called PBS, that is proposer builder separation. We get into the Ethereum roadmap a little bit. Number five, what's next for flashbots after this thing called PBS is deployed and makes them obsolete? We talk about the flashbots organization as well. David, there's so much here. How would you rate this episode in terms of like skill level or knowledge level? Is this accessible enough for the straight beginner? Does this avere towards a little bit expert level?

I think if you have not heard the term MEV before, well, first off, I remember my first time I heard MEV. And if you like rabbit holes, I'm so envious of you. Going down the MEV rabbit hole for the first time was so much fun. If you have not yet heard of MEV, this episode should be fine. It'll start fine, it'll get challenging towards the end, but overall, it'll throw you down a fantastic rabbit hole. The MEV part of this crypto industry has attracted some of just straight up the gigabrains. Like all the giga brains just get attracted to MEV because it's such a complex, nuanced problem to solve with so much at stake. There's a lot at stake if we do not solve MEV or if we solve it incorrectly. And so there's just these certain types of people that have really been captivated by the MEV problem and have really put some of the highest amounts of IQ all targeting the same problem, which is solving systemically how to solve the MEV problem. Just really, really quick the MEV problem. Maximally extractive value is this thing that we discovered back in like 2017 or 2018 as a function of smart contract blockchains where you have DeFi. And anytime that you make a trade on like Uniswap, you imbalance a pool, right? Like if you are trading USEC and ETH, if you put in USEC and you pull out ETH, that changes the pool, that changes the price. And arbitragers can come and rebalance that pool as a function of how it relates to other exchanges. So it'll balance out like the price of these things. You can also talk about MEV in the context of like liquidations on MakerDAO or Aave. It's basically when users use a smart contract blockchain, they leave arbitrage opportunities in their wake. And so arbitrage bots will come and make transactions to pull out some of that value. They'll rebalance the pool, take a little bit of arbitrage. And this is all normal and healthy. It gets more and more malicious as you get further down the MEV rabbit hole. There's like front running. If you are trading on Uniswap, you can get front run by a bot and they'll sell you what you were going to buy, but at a higher price. That starts to get a little bit malicious. And it can go down to what some people will just define as straight up theft. And it can even get even more systemic than that. It can start like disrupting actual blockchains, as in like a couple blocks will have passed, but there was an MEV opportunity that was so incredibly lucrative that MEV, like people will try to actually unwind the blockchain and go backwards for blocks in order to capture that opportunity before going forwards again. So like not only is it just about arbitrage, it's about just like, you know, bot attacks in the mempool, but also destabilizing the blockchain. And so it's critically important that we get this right. So that's like your quick TLDR rabbit hole on MEV. And Flashbots is an organization of people that have targeted this problem directly, creating this sort of sandbox, if you will, for MEV players can play in this sandbox that doesn't impact like users maliciously or disrupt our blockchains. And Stefan here is one of the co-founders of Flashbots and has a lot to say with where Flashbots currently is in this proof of work paradigm, but also where it's about to be in this proof of stake paradigm, because proof of stake definitely changes the game of MEV and the role that Flashbots has to play in this ecosystem.

Absolutely, guys. You find there's some strange God mode type abilities in the transaction ordering of our blocks, and that is the source of MEV's power where a certain party has the ability to order blocks in a certain way. Anyway, we're gonna get into all of this. It's a fantastic episode and how we solve it, how it relates to the Ethereum roadmap. And by the way, guys, if you want to watch bankless podcasts on video, you could do that on YouTube, but you can also do it now on Spotify. So all bankless podcasts are now published on Spotify with video. So you can go ahead and check that out. Let's get right into the conversation with Stefan from Flashbots. Hey Bankless Nation, we're super excited to introduce you to our next guest on a very important topic today. This is Stefan Gosslin. He is the founder of Flashbots. What is Flashbots? It's an RD organization that's research and development that's trying to fix what we think is one of the most pernicious problems in crypto, maybe the most pernicious, the problem of MEV. That stands for maximum extractable value. And while David and I are massively optimistic about crypto, and we're not really worried about too much in crypto. I mean, we think this thing is going to be pretty much inevitable. We are kind of worried about MEV. It's the one thing. It could really be crypto's Achilles heel. So we're going to talk about it today with somebody who is trying to solve this problem and actually whose work in Flashbots is getting ready to become important enough to actually get embedded in the core Ethereum protocol. So we'll talk about that as well. Stefan, welcome to Bankless. How are you doing today?

Hey, I'm doing great. Thank you so much for having me. This is exciting. Always a pleasure to

talk with more people about, yes, the very scary problem of MEV.

Well, can't promise you about the smart part, but um hopefully you bring the intelligence for David and I, because uh in fact, you know what, we want to kind of start at the less smart level of this MEV problem. So I called it the most pernicious problem in crypto. Um, do you think we're like blowing this out of proportion? Can you tell us why MEV is important at a high level and what it is for somebody who is just now hearing about it?

Yes. So, no, I don't think you're blowing it out of proportion, though I am very biased in saying that. We did literally get every single person that we could figure out that agreed with this point of view together sometime in 2020

and start this organization together.

That's sole mission is to figure out how to solve the problem of MEV.

So I am definitely in my little MEV bubble where I spend all my day and night thinking about it and worrying about it. So, why did I surround myself with a bunch of people who worry about this?

Well, MEV is sort of this emergent phenomenon that we've observed on smart contract blockchains when people start to actually use them.

We realize is that there's a lot of different powers that miners have.

That they can use to extract sort of extra value out of the protocol in ways that the protocol wasn't necessarily designed for. So when the initial sort of paper by Phil Diane came out and defined the term minor extractable value,

it looked at things like reordering, censorship, and insertion of transactions within blocks, looking at basically arbitrage opportunities on decentralized exchanges

and how

when you have miners that start to target these, they might do things like.

Doing re-org attacks against the chain and some other sort of nasty behavior.

At the time, it was mostly like a theoretical paper that looked at okay, well, here's what would happen if

these DEXs would get a lot of traction.

And lo and behold, a couple years later they did.

And around you know, uh summer 2020, during DeFi summer, we started seeing a lot of these

bots becoming very active to the point where it started having sort of a negative impact on the rest of the system.

The two years following that, so between summer 2020 and where we are today, has just been sort of an explosion in sophistication

of bot activity on Ethereum, but also in every single other smart contract chain that has any sort of meaningful economic activity. And we've seen sort of the space go from very simple sort of arbitrage strategies

to evolving into like very complex generalized front running liquidations, cross domain arbitrage.

All kinds of NFT sniping and uh strategies that have just created a ton of demand and activity for these chains.

You know, the role that Flashbots plays in all of this

is we see MEV as being probably the biggest centralization risk

for crypto.

The way to think about it is

if chains aren't actively thinking about

MEV in their design, they are guaranteed to make some fatal mistake that will

have some negative externalities on their users.

Or cause them to become very heavily centralized.

So we sort of take a research first approach of saying, okay,

we know that this is a risk. Let's try to understand it better.

And then let's develop solutions that help blockchains deal with this risk. So we launched on Ethereum and we have this solution, MEV Geth,

on Ethereum that's running with all the miners. And then we're preparing, as everyone is, for the transition to proof of state.

Safan, before we go into uh Flashbots and MEV Geth and some of the solutions to this problem, I'd really just like to find the scope of the problem a little bit more. At the very basic level, MEV is just basic arbitrage opportunities. There's a difference in the ETH price between Uniswap and Balancer, an MEV bot can arbitrage those differences. And so, like in theory, the idea is that some bot will come in and balance out those two pools, but then the whoever's actually mining the block will see that bot come in and they will front run that bot. They will balance those two pools instead and they will be able to pull out that arbitrage. They're just able to arbitrage faster. At this level, it the things aren't all that crazy. Things aren't all that evil or so. And this is kind of actually considered healthy in the ecosystem, in the DeFi ecosystem. You know, liquidity pools stay balanced faster, liquidations in Ave or compound happen faster, keeping those systems more robust. So there's like a lot of good MEV that's out there that just helps stabilize the DeFi ecosystem. And like it's almost synonymous with just general arbitrage. And this part of MEV we don't really have a problem with. Can you illustrate the more dark side of MEV, the evil side of MEV, down to the point of uh things that are considered ethically poor or ethically dubious, but also down to the point of like destabilizing a blockchain? Can you tackle those two things as to like the bad side of MEV?

Yeah, you know, this is where even a good thing has its drawbacks. In many ways, MEV is what makes a lot of DeFi systems work. It's what makes AMM possible. Like AMMs, you would not be able to be a passive liquidity provider if there wasn't someone doing arbitrage between

Uniswap and Binance, right? It just would not work as a strategy and like no one would use a system. But because there's someone doing that arbitrage, the prices on Uniswap are able to stay in sync.

So it has some value there.

Similarly for liquidations, right? If you run a liquidations protocol that's using some kind of Oracle system.

You need some bot operator to come in and actually execute the liquidation to keep the collateral ratio of the lending platform in balance. And someone literally just needs to like send the transaction that triggers this function. And that is MEV. It's rewarded in some way.

So good MEV takes place when a protocol or like a DAP designer.

Thinks about okay, what are the permissionless functions in the system that we want to incentivize third parties to take? And then how much money do we need to pay them to take these? Where it can starts to become a little bit more questionable and nefarious is when protocol developers and dab developers

do not think about the MEV that's exposed through just regular user interactions.

So when their users start to interact with their protocol, they might actually be exposing some MEV as well.

And that might have some negative sort of impact on their experience. So on Uniswap, for example, the case that happens quite frequently, right, is you set your slippage limit to high and then send a transaction for a swap into the transaction pool.

Some bot operator out there sees this, they sandwich your transaction and provide you the worst execution that you accepted using your slippage limit.

Uniswap by like allowing users to set high slippage limits and like allowing them to send it to the transaction pool might in many ways be giving them a worse price execution.

And so those kinds of behavior end up being sort of less beneficial to the end users because there isn't really any reason for the system to have these kind of value exposure.

Got it. So like what you just described is maybe in a bad MEV scenario, the user gets ripped off a little bit, right? There's some additional costs that they're paying to these kind of MEV rent collectors. But can you take that a little bit further? So, like, what is the worst case scenario here? I'm almost like maybe talking more from a system perspective rather than the kind of the user. So the user experiences this, they're getting ripped off in small ways here and there. Maybe they notice, maybe they don't. It would have to get really, really bad for them to maybe like fully notice. But I think there are some more pernicious uh like system level attacks. And I want to ask the question of like, okay, so this is a problem, but what happens if we just don't solve it? Like, what happens if we just like leave it? How will the system evolve in that condition?

Yes. So I sort of mentioned briefly negative externalities. I think from a system designer perspective, from the perspective of Ethereum core developers, Solana core developers, whoever, right, who's looking at how do I build a blockchain?

You have to see okay, what incentives am I creating by these like different mechanisms I'm implementing?

Three ways in which like MEV tends to be extracted is either a spam war,

a latency war,

or an auction. This is like different ways that you allocate opportunities or you design a system of block space, and how competition for that block space ends up being targeted is using one of these three systems. And each of these have these negative externalities,