See all Lex Clips transcripts on Youtube

Speaker 1

00:03

Some people describe MEV, minor extractable value, as an existential risk to Ethereum. What is MEV? How important is it to solve MEV? If it's important, what ideas do you have?

Speaker 2

00:15

Sure. How about after this 1, we'll also talk about sharding because it's amazing. Yes, we'll

Speaker 1

00:20

return back to sharding, which is, we'll return to the big picture of the scaling problem, as you mentioned.

Speaker 2

00:26

I love this conversation, you know, depth first search instead of breadth first. So basically, okay, EBV, minor extractable value, it is not different in proof of work and proof of stake, right? So like, if you want to call it, you know, block proposer extractable value, like it sounds less sexy, but you know, we can call it BPEV instead of MAV, who cares?

Speaker 1

00:46

But the- So this is a problem in both proof of work and proof of stake.

Speaker 2

00:50

Yes. So the basic idea is that if you have the ability to choose which transactions go into a block and in what order, then you have the ability to take advantage of that position for economic gain in a lot more ways than just collecting transaction fees. For example, there's decentralized exchanges on chain like Uniswap. Let's say the price of ETH versus USDC was 2, 700 the previous block, but then there was a bit of a market drop and now it's $2, 680, where you can go on Uniswap and you can just gobble up the entire part of the automated order book that's between $2, 700 and

Speaker 1

01:30

$2, 680.

Speaker 2

01:32

And then at the same time, you run a bot and you buy some ETH back at 2680, and you've just made about $10 of profit. Or well, $10 times, whatever the depth is. So there's lots of little things like that.

Speaker 2

01:46

There's also things that involve front-running other people's transactions. So 1 example of this would be that if someone sends a transaction that says buy me 5 ETH for whatever price that you can get, but with a maximum of, let's say,

Speaker 1

02:08

$15, 000,

Speaker 2

02:09

then you can go and you can put a transaction right in front of that transaction and you can buy up that ETH first, and then you resell it to him at $15, 000 minus 1. So there's- And then you

Speaker 1

02:20

get to make a little bit of money that way.

Speaker 2

02:22

Exactly. So there's a lot of these different like arbitrage, front running, back running, these different tricks that allow block proposers to-

Speaker 1

02:29

To get some percentage on top, like overhead.

Speaker 2

02:32

Exactly. And the reason why this is a challenge is because it's, like, first of all, it's sometimes, it sometimes degrades user experience because users get in a less favorable trades, but there are sometimes ways to mitigate that for applications. Sometimes it's not that bad. But the bigger risk that I think some people consider more existential is that there's just much more economies of scale in figuring out how to extract all this revenue.

Speaker 2

03:01

Because if you're just collecting transaction fees, there aren't really economies of scale. There aren't really benefits to centralizing, right? Because it's a very simple formula. You just like grab up the transactions that pay you the most.

Speaker 2

03:11

But with MEV, there's all these sophisticated algorithms. And if you have lots of money, then you can hire really smart people to make amazing algorithms. And then you can use the other half of your money to get a lot of mining power, a lot of stake, and you get a lot of opportunities to use your even better algorithms. So there's this risk that as a result of this, mining is basically, or even validating proof of stake is going to centralize.

Speaker 2

03:39

So I think the ecosystem's best reply to this sort of risk, and it's the direction where projects like Flashbots are going already, is if you can't eliminate the centralization, then you try to firewall it. And the way that you firewall it is you basically say, we're going to try to deliberately create a marketplace where people can just do the complicated work of creating what are called bundles, like bundles of transactions that are very profitable. And then at the other side of the market, you just have block proposes reminders that are just dumb notes. And they go and ask the what are called searchers, the bundle creators, and they just ask, like, hey, how much can you give me if I put in your bundle?

Speaker 2

04:24

And then they just take the highest offer. Right. So you sort of separate out the task and you know, you have the easy part and then you have the hard part and you have like this special class of actor called a searcher that does the hard part. And then the easy part, the people doing the easy part, which is just miners and validators, they kind of just talk to all the different people doing the searching and they just, you know, accept the highest bidder.

Speaker 2

04:45

Right. So, I mean, this is also just an interesting example of economic design philosophy. Sometimes you can't just make centralization go away. Sometimes it's inevitable.

Speaker 2

04:58

But at least you can try to contain it, you can direct it, or you can even sort of firewall it away from core consensus, the parts that really do need to be decentralized.

Speaker 1

05:10

But you don't see it as an existential risk, it's just a bit of a problem that has to be constantly dealt with.

Speaker 2

05:16

It's a risk. Like, there's obviously a risk that, you know, it's a very severe problem and that even this flashbots approach has some fatal flaw or whatever. But I'm definitely, we're definitely approaching it with the mindset of,

Speaker 1

05:45

you