Is Ethereum Realtime Ready?

Transcript

[applause] >> Good afternoon. I'm Kevin, founder of EthGas. Um where do I start? I guess for a background I have a TradFi background, but I also have like a tech background. And so, TradFi, I started off as a vol trader. Um I started off as an FX option trader, then got into interest rate options, uh structured credit. Uh I ended up leading Morgan Stanley's uh structured credit group for a number of years pre- 08, uh Lehman days, and then after eventually got into like regulatory compliance, cuz that's what happens when the world blows up in credit, and I didn't really enjoy that. So, I I got into um uh built a a tech company

separately. So, got into tech, and here I am now at the core of Ethereum infrastructure. And so, today we'll be discussing uh Ethereum, and you know, how do we make it much faster, much more performant, much more fair, so we can really, you know, benefit all the users and get the broader uh ecosystem on chain. And so, this is just a brief kind of recap of where we are from two perspectives. One is uh AI, and the other perspective is yield. So, uh as it is now, you guys have probably heard some presentations on agents, and indeed, uh this is probably outdated even as of uh last week. Uh but we're seeing a significant amount of flow

being uh agent-driven. There's no like way to definitively test if it's an agent or not yet, right? Whether it's a bot or an agent, it's, you know, um debatable. But, um for the for the most part, basically everyone's using agents, but the point is that it's streaming down towards like the everyday person with, you know, cloud code and codex, and um everyone being being able to access um you know, higher velocity um, micro payments, really. Um, X402 I'll skip through, uh, but basically from a yield perspective as well, we're seeing this surgence, I guess, in vaults and yields.

And so, even last week there was a uh, a yield summit at the New York Stock Exchange, uh, with Morpho running that. And and it seems like almost every institution or uh, regulated entity I talk to, one of their first forays or points of access is within kind of that vault space. And so, uh, we're seeing a lot more different types of curation or strategy, and it's not called asset management, it's called curation. But, um, indeed, it's a very exciting space, uh, growing. The problem, however, is that well, Ethereum was built, you know, 10 years ago, a little bit more, and it was built for humans or users, and, you know, like, was it built for, you know, machines to

be trading every like, uh, millisecond or two? And so, Ethereum is very slow. It's very uncertain, you know, when you try to land a trade on chain, you think it might take 12 seconds, but it might take 24 or maybe 36. There's a lot of this uncertainty around, can I lay can I land a trade? When is it going to happen? And this becomes extremely difficult as it relates to, um, multi-agent coordination. So, if I have an agent that's trying to land a trade on chain, you know, it's like probabilistic. It's not like certain that this is actually going to happen. And how do you how do you manage? How do you work around this?

How do you build fallbacks? And so, this is a frustrating process for a human as it is an agent. But, when you have multiple agents running different strategies or, uh, having different functions, be it like reporting or compliance or something else, it also gets difficult. So, um, you know, how do we get more certainty? Uh, ETH APRs, I'll skip past that. But, on the KelpDAO and Aave part, you know, I will also touch on this a little bit later. Um but the point is that with Um uh what's what's the new Mythos name?

Just came out? Is it? Bit What's it? Fable. Is it? Yeah. All right. Um so, with that coming out, you know, AI will undoubtedly break certain things, but it's important that these break before we get trillions of dollars on chain. And so, you know, uh it it'll be great to kind of get through these. But the point is that there will be multiple uh unfortunately um minor attacks in the near term before we can get kind of this high-velocity, high- high-volume business.

Um finally, one point here is that a lot of agents, as they are currently, are rather predictable, just as our strategies for vaults. And if anything is predictable, then you can be front run. And so, that's one of the issues also is that um you know, the more predictable, more autonomous things are, uh you need to kind of keep them unpredictable, but uh I'll touch on this a little bit later. The solution to many of these things, as it says at the bottom, is block space. And I want to walk you through this experience of how does block space and how does the reordering and restructuring of block space change this?

And so, you know, for the last 10 years, arguably, we look at Ethereum's blocks as just being built every 12 seconds. And so, you have the previous era, you know, proof of work, and now we're in proof of stake. But for the most part, we just we just kind of take it for granted, all right? Now, uh last year, we did something pretty cool. We basically said, "Let's get access to this block space, and let's do something with it." Um so, we then uh we got access to this block space, and now you can buy, sell, and trade it just like you would um um, and gas or crude, for example.

And so, um, yeah, you can basically speculate on oil and gas. Secondly to that, uh, we've also managed to, like, refine block space for higher performance and for other purposes. And so, I'll get into that right now. Ethereum Ethereum's core business is in the production and the sale of block space. That's all it does. Um, it happens to be a very high-grade, high-quality type of block space, um, because there is unlimited block space on other chains, but Ethereum's is scarce, uh, and it's very high-grade.

Um, so, as I mentioned, you know, we create this ability to trade it just like a commodity. You know, you think you think 10 year 10 20 years into the future, global finance is running on Ethereum. We have trillions of dollars. And here here is the pipeline. Here is the fabric that actually controls the network. And so, are there going to be various actors trying to control or manipulate that? You know, we need an ability to trade, manage, uh, risk manage this type of instrument. And so, that's why we created the ability to to trade these instruments.

Now, I'll talk a little bit about how blocks are made. So, uh, we we effectively have, you know, a a few blocks here. Uh, and the blocks are built every 12 seconds. 12 seconds is super slow, super boring. Now, what happens is that trades are they are propagated through the broader ecosystem. You know, broadly speaking, some trades go to the public mempool where everyone can see what's there. Other trades go to, like, a specific builder or another builder. They just kind of like waft around, like, Ethereum.

And then, slowly over time, then, at the end, boom, they are put in here. Now, what happens at this point in time from the previous slide to here is um, is effectively you have these block builders who will reorganize the trades to maximize the amount of uh extraction effectively uh in the block. And so it you you liken it almost to uh like a nightclub where it's like, "Hey, like you're good-looking, you're good-looking. You guys are going to come in first. You guys at the back, it's going to take a while. It's very crowded." But it it's kind of like that.

You basically take like the most profitable the best trades and you reorder them in such a way. And so it's is that fair? No, right? Is it extractive? Yes. Uh and so again, you know, we need more certainty, we need more speed. We don't want people waiting at the front door to this club. We want to let them in like right away. So um what we've done is we've created these fast and exciting blocks. And so how that works is whenever trades appear they immediately get inserted into the next block. Okay? They immediately get inserted into the next block and every 15 milliseconds we effectively broadcast the state of the block. So if someone used Uniswap or Curve and it shifted the

um it shifted the prices a little bit, then every 50 milliseconds you're going to get a new price update as this happens. And so effectively, you can trade multiple times across this 12-second interval. And not just once. And so I think I get into this a bit later. But effectively, when you can trade multiple times in that same slot, then you're going to be paying, you know, more fees because potentially you can make more money. Um you know, the validators are going to earn more, the protocols are going to earn more, the entire state of Ethereum is richer simply because it's much more usable now in this environment.

And so we call these pre-confirmations or real-time pre-confirmations. Um currently, we can subsequence a block in about 50 milliseconds. We can get it down a little bit further, but for the most part, it's a pretty good start. Um and finally, of course, it brings, you know, much faster speed and much greater certainty around execution. Now, you know, I I alluded to this a little bit, but you know, what does it mean for for DeFi? And so, as I said, you know, when you can trade multiple times. So, we make the the experience over 100 times faster. It doesn't mean we have 100 times more trades.

Practically, maybe we have three or four times as many trades, but when you bring all that together, then, you know, effectively on Uniswap alone, the LPs would make two to three billion dollars a year. And this is simply just sex decks trading in between Binance and centralized exchanges and like just that venue itself. So, we ran this study earlier in the year. It is uh based on empirical data. It's also based on theoretical data, and both of the two align. So, um effectively, yeah. If you were to make Ethereum real-time, Uniswap alone, the LPs would make somewhere between two to three billion dollars.

Um okay. Now, secondly is there's less MEV. There's less extraction. And so, currently, there is about uh 250, 300 million dollars a year of extraction. And uh this is bad, right? Uh Do I get into it next? No. Uh Where this Where this leads is like uh as I said earlier, if you're running strategies in a vault, then every time you land a trade, you're going to get front run. You're going to get exploited. So, if your return on the vault is 6%, it should be 7% without MEV. And so, that that's what we want to create on that side there.

Um So, yeah, there's just less MEV, it's more fair, less extraction. Uh, secondly is uh, event risk. And so, I define event risk as you want to um, be able to mitigate, I guess, large movements in price, right? Uh, for the most part in financial theory, you assume prices move with some, you know, continuous time behavior, some some diffusion, but effectively when there's a large price jump or gap risk, you can't hedge and you can't do anything. And so, right now with uh, liquidations, you have this 12-second interval where Binance has dropped, you know, 10% and the price on Uniswap is still for ETH is still 2,000. And so, in that period, you

basically have like a massive liquidation. It poses a risk to the to the borrowers, it poses risk to the protocol as well. And as we've seen through Kelpp and that situation, it poses systemic risks across the entire ecosystem. With much faster block times, you would the the price jump is much much less. So, you're not going to the price isn't going to jump 10% over that 50 milliseconds, maybe it jumps 2%. So, the point is that event risk is reduced substantially. That's good for the individual users, it's good for the protocols, and it's good for systemic for the for the entire ecosystem as a whole. So, um, so that's where it happens.

Um, finally, there's uh, this initiative we call institutional block space. And so, you think about it, a very small portion of global finance is actually on Ethereum today. It's it's down that way, you know, on Wall Street. Uh, with this uh, how can I say? With with this framework, what we've created is an ability to almost like white list the validators, the block building, everything end-to-end. Uh, I won't get too deeply into this just given time. Uh, but effectively the way it works is that we have conventionally, anyone can land a trade on Ethereum.

Um you will have block building built by a variety of builders. These builders will do um MEV. And you'll have a validator, which could be anyone. But on this side of the equation, let's say you are a BlockRock or Fidelity or Franklin Templeton, you want to buy like 10 million of some asset uh on chain. You don't want to buy 10 million of an asset and be front run. You'll be very happy to pay 100 bucks, 500 bucks, whatever it is, just to not be front run.

You collect the universe of people in TradFi who don't want to be front run. Um and there's an incredible amount of money to be paid there to to prevent that. And so currently within the traditional financial system, there's about 10 to 15,000 dollars of fees paid to intermediaries every 12 seconds. That is kind of the like the the TAM or the upper scope in terms of how much could actually come onto Ethereum. And so currently it's $50 for a block, but can we get higher paying uh more compliance-first types of players on stream? And so with this institutional block space, you have a validator at the bottom who is uh known and uh has an address. You have a block

building process, which is not extractive. It's first come, first serve. And then a set of policies around it. Maybe for these regulated entities, you can't touch Tornado Cash. So um in this example, imagine TradFi saying like, "Hey, now I can come on chain. There's no front running. Uh no one's going to touch Tor- Tornado Cash. Uh it's a validator I know." And so all the controls, procedures that they're accustomed to within a traditional financial world, they can now apply on Ethereum. You can access all the liquidity of Ethereum. They don't need to set up another L2 or another L1.

Yesterday, we had this session. We We basically talked about this, but we had about 70% of Ethereum validators broadly aligned to to to move forward with this in some direction. So, expect more from us soon on this account. So, where are we currently? There is We have about 5% of the network, but this will be changing very quickly. Um Towards the end of this year, I thought we would have maybe 20, 30%, but I think we'll have a lot more coming through us. And the key point here is really that a lot of people are scaling Ethereum.

It's like us and a whole bunch of other folks, right? You have the size of the block going last year from 30 million gas to 200 million this year. As we expand like the size of this pipe, we can't have MEV and these blockers at the front just like stopping people from people from using the chain. Like, how do we use the block space that's there and maximally? So, yeah, that's where we are today. Um I'm Kevin, founder of Eth Gas, and thank you for having me.