Flash Talk by Ben Weintraub - Creator, Pinto

What's up everyone? Can people hear me? Okay, wonderful. All right, so I am here to talk about how DeFi will finally out compete tradfi. Uh let me see.

Is this working? Uh hold on. Yes. Uh this is something I think ultimately we're all trying to figure out how's it actually going to happen. stable coins have started to show some sort of product market fit, but at the end of the day, we're still a far cry from having the adoption or really the type of productive economy that can out compete tradfi.

So, what I'm going to talk about is one of the key properties that we need in order to get there. And that property that I want to highlight is expressivity. uh when you zoom out and think about what is DeFi, uh markets are really the best way to express and profit from your views uh on the truth when it comes to prediction markets or information markets uh on mispricings in general and DeFi has started to find its home uh when it comes to different market types. But at the end of the day, uh there are a variety of different problems that have to date uh held back DeFi from out competing trad. And I want to highlight three in particular.

First is a lack of privacy. Uh at the end of the day, there's a real limit to how much people are willing to express themselves in an environment that's totally public. Uh and privacy is a whole topic in and of itself, which I've decided not to focus on today in my talk. Uh but I'm going to focus on two other uh major problems that are currently limiting people's expressivity when they're using DeFi. Uh first is capital inefficiency.

Uh right now in crypto and in DeFi in particular, uh I have to choose whether to express A or B. Uh what this looks like in practice is I can put my capital in this protocol or that protocol or I can uh add value to this order book or that order book or this AMM or that AMM. And this choice of A or B means that even if I have perspectives about a variety of different markets, a variety of different data points, I really have to fragment my capital across all of these positions. uh and that obviously results in a limit in how much I can express myself. The second where where exclusivity is really limited is when it comes to developers.

Uh and the reality is that if you have built anything on chain one, you know, there's an incredibly high friction to do so. Uh but more than that, it results in uh a very limited set of protocols that give you a very specific set of actions that you can do. We are far from a totally composable uh highly expressive developer environment where everyone can uh even if they have one idea for one function to improve things uh can actually manifest that. So the two things that I want to highlight uh and speak about today uh are capital and efficiency uh and high development friction. So before I keep going and I don't want to talk too much about this because usually I hate when people talk about themselves in these talks.

Uh I just want to get across I've tried extensively as you see all these protocols that I've worked on and been one of the founders of are all in the EVM. Uh and what I've learned the hard way respectfully is that the EVM sucks and it's not going to cut it. And so if ultimately our goal is to have a DeFi environment that can't out compete Tradfi, we need to do better. So what I'm here to talk about today uh is a new primitive in the blockchain environment which has the ability to radically improve the expressivity of capital and the expressivity of developers to express their ideas and that is a new type of blockchain state. Uh it's called reactive state.

Now in the traditional model whether it's the EVM or I have a lot of experience or the SVM or even the move VM all of the blockchain state is static and what that means in practice is that in order to trigger a state transition the way that it works is you have to read the state perform some computation offchain calculate your transaction submit that transaction to the network have that transaction get ingested by the network and then once it gets sequenced it gets processed. There is latency at every one of these steps and this latency results in uh a lot of capital inefficiency which limits your expressivity and as we get into it you'll see it also has an effect on the types of experiences that developers can create. So over the past couple of years, uh, one of my co-founders and I have been working on this new VM we're calling the Nebula VM that introduces reactive state that inverts the way that the VM works or that state is stored in crypto where instead of in the traditional smart contract model, you take value and you put it into some smart contract where the logic is in that smart contract and typically it works with invariance. In the Nebula VM, this is totally inverted such that you continue to hold all of the value in your account. And now within your account, you can attach arbitrary sets of functions to your account that each individually if the conditions associated with the functions are valid uh can trigger a state transition.

So the result, the punchline is that because I can now attach multiple sets of independent functions to the same value, you go from this uh like system where you either have to put the capital here or here to now I can actually deploy my capital here and here simultaneously. And this has quite radical implications when it comes to DeFi on both the capital efficiency perspective and the developer perspective. So I want to talk a little bit about each of those. So let's start with capital efficiency. Uh and there's two types of capital efficiency I really want to talk about.

One is for orders and the other is for loans. So when it comes to orders, uh this reactive state for the first time allows me to quote multiple order books with the same capital. Uh the result is if you look in the longer tale of assets uh or even more exotic uh value like prediction markets, right now you have a tremendously high competition for liquidity. And again, it goes back to liquidity providers can either provide capital here or here. And so they have to pick and choose which markets to provide liquidity to.

In the Nebula VM, in this new reactive state model, you can now quote multiple order books. And in the result is you entirely remove liquidity fragmentation from the market. And by removing this opportunity cost, you remove the competition between these other assets. Uh and so the result is we expect particularly in the longer tale of assets uh in those more exotic markets like prediction markets and information markets this is going to result in a dramatically deeper liquidity uh tighter spreads and generally more efficient markets. Uh the way I would put it with orders uh or in the context of orders is you can now express more positions with the same value.

Now, it's worth saying and going back to the idea of competing with Tradfi, certain centralized exchanges already support this. So, particularly for market makers on centralized exchanges, they often have this feature called debitless orders where the market makers can quote multiple order books with the same capital. This is the first time that that will be possible in a decentralized trading environment. uh and I think it's quite a big deal honestly because now it's also permissionless whereas in most centralized contexts you have to be permissioned to get access to this this is entirely permissionless now in addition to just having tighter spreads and deeper liquidity for assets uh which ultimately is a major feature when it comes to markets and what they're competing on lending as we just heard from the previous speaker uh is another major uh phenomenon in crypto and the way that the debitless orders or reactive state I should say uh affects lending is similarly profound. So there's a couple elements to how it relates to lending.

Uh the first which isn't written here is that I can now quote uh if I have value that I'm willing to lend out. I can now offer a a an infinite variety of terms under which I am willing to lend my value out. So this is really like something in between orders and lending where you sort of place an order to lend out value. You can now say okay if this is the collateral you post these are the interest rates. If you want a fixed rate this is what I want.

If you want a variable rate this is what I want. Uh and you can really customize it infinitely. And because you can now attach multiple sets of conditions under which to facilitate the exchange of the loan. You can now as a lender get super creative. And what this means is now when I'm a borrower and I come to the market, I'm going to be given a a significant variety of options under which I can borrow capital.

So when we think about again the the point which is for DeFi to out compete trad, this is the first time where if I'm a borrower in a permissionless setting, I'm going to be able to borrow stuff with a lot more flexibility than what exists in Tradfi. And it's worth saying right now in Tradfi, these types of more custom or complex loans are typically limited to large institutions or rich people that can go to Goldman Sachs for example and have them customize some sort of loan package for them. Whereas this will facilitate super complex liquid order books for loans uh in a permissionless context. So then when we sort of zoom out and think okay well what else can we do with lending in this reactive state model this also facilitates rehypothecation and your first reaction should be a little bit like uh-oh yikes rehypotheation I thought that was dangerous and it is uh where the idea is that I can now uh post some value and lend it out on something like a uh and then I can take that outstanding loan and I can post that as collateral on a on compound on another protocol and I can use that to borrow against. So I have some outstanding borrow and now I'm borrowing against that value that I've already lent out which is rehypothecation.

And [snorts] the difference between and when we go back to DeFi out competing tradi the difference is whereas in tradi rehypothecation exists and there's all these different pockets of risk in the system that are opaque where if you have insolveny you may not even know it in an onchain lending environment you can actually get a proof of solveny. So this is a really powerful thing where this reactive state for the first time facilitates an onchain lending environment that is just as efficient as traditional financial uh loans and that entire banking system but without the risk where you have this opacity and you can't see into it. So on these on top of this type of reactive state we expect highly efficient lending systems where you can actually have proof of solvency. Uh so just to summarize some of the big things you get on the capital efficiency side, you get debitless orders which means you can quote multiple order books whether it's for value, for information or for loans at the same time. And then when it comes to lending markets, uh you can rehypothecate.

Now the big vision here when it comes to capital efficiency is for the first time you have like a unified account at the blockchain level not at the individual protocol level at the entire blockchain level. So any value that is deployed on a chain implementing the NVM the Nebula VM will be able to use all of the value against itself. I was just talking to someone who I think this it clicked for them and they were like wow so for prediction markets for example where uh particularly for longer dated markets uh participants don't want to lock up value until the market is resolved for a year or two years or longer time horizons. Now you can actually take that position and use that position as value throughout the system. And so it's not just capital efficiency in the moment.

its capital efficiency across time and through this expansion of how value can be used is how I expect DeFi to ultimately compete trady when it comes to raw capital efficiency. Now I also want to take a moment to talk about how this affects the developer experience. Now uh again if we go back to it uh I've been through the ringer trying to create really cool stuff in the EVM and ultimately there's a lot of friction there. So, one is just let's say I have a good idea uh for a new lending protocol. In the current environment, if I want to out compete the state-of-the-art, I really need to innovate at least a full order of magnitude, a 10x improvement or greater on whatever the state-of-the-art is.

And this is a really high bar to get over. For example, if you consider Morpho, right? Morpho just came out with V2. You're like, okay, they're they're missing one or two features that I like. In the current environment, the only way to improve on that is you have to create a competitor that in order to really convince people to use, you have to have an expectation.

It's going to be a full step ahead, which is really hard. It's a high bar for people to get over uh in this environment because again, people are now just attaching logic to their account. You no longer have to, if you're a developer, you no longer have to create something so good that you can convince people to withdraw value from Morpho and deposit it in your new protocol. The value proposition is now just, hey, if you also attach my software to your account, you get additional features. So, it's a dramatically lower friction for developers to get their technology adopted.

And the result is there's a lower hurdle for developers to say okay I have one idea and I want to implement it and I have a reasonable expectation that people will actually use it. So as a result uh we expect that this will significantly accelerate the rate of DeFi adoption. This could be the types of loans that are available the terms under those loans uh the types of markets that are being traded. uh if you zoom out and think about like in an abstract way even the shape of different tokens that are being traded can be changed uh and now you don't have any sort of competition where people have to choose between A or B they can now provide liquidity to protocols A and B at the same time now there's a bonus here uh that I want to talk about briefly so one of the biggest risks in crypto is when it comes to smart contract upgradability where most smart contracts are upgraded able and the result is that if you're using the protocol in practice, you're trusting whoever owns or controls that smart contract not to steal your funds. Uh and there's some friction here because of course once values in the system uh it's really hard in the present context to have people upgrade the system uh without of course uh compromising immutability.

So I want to give a an example here of just how powerful this new reactive state is and this uh what you might call function level composability is. So uh back when we were building Beantock uh we had used unis swap for all of our liquidity and the unis swap pools uh the protocol requires an onchain data oracle. uh Uniswap's pools had a data oracle for onchain data that was MEV resistant up until the merge happened because after the merge for the first time you had the potential for interblock MEV. Uh once you had the merge you had the potential where a sequencer would know not just that they would have control over one block but they would know in advance that they would have control over multiple sequential blocks. And the unis swap oracle relied on the assumption that there it would it would basically be uh like from a capital perspective impractical for you to manipulate the oracle at the end of each block when it would track the data because if you left a price mismatch or a a crazy price in the pool at the end of the block at the top of the next block the expectation would be that someone would arbitrage you.

And so practically that meant that the oracle was safe from mev or some sort of manipulation. Now after the merge and you had the introduction of multiblockme for the first time now uh you could arbitrarily manipulate a pool between and across blocks and the result is that the unis swap oracle was no longer useful. It was highly uh manipulable. So we of course went down a rabbit hole to try to design an oracle for manipulation resistant data. uh even though you could arbitrarily manipulate the underlying data which was quite a doozy in and of itself but once we had the oracle constructed we faced a problem which is in order to get that oracle adopted we would have had to go to unis swap and convince them to adopt our oracle and even if they had adopted our oracle they would have had to make it such that their contracts were no longer immutable in order to upgrade the existing contracts to support our oracle in each of these stages there's an incredible amount of friction to having again just a light change in practice it's just the oracle of the pool it doesn't even change the underlying math of the pool to have that adopted so what should have happened right what should have been able to happen is we should have been able to take our oracle and attach it to the existing pools with some logic that says anytime someone trades in this pool update the oracle and charge the protocol for the gas so not the trader has to pay the gas we'll pay the gas but anytime someone trades is update our oracles so we can continue to use the existing pools.

That of course was not possible in a reactive state environment. What you can now do is exactly that. You can take existing software and you can just attach one more function to what exists already such that when anyone trades through a state uh it could actually update the oracle. So what this means is for developers you can now have the best of both worlds. You can have immutability and in a way you can have upgradability.

The upgradability follows sort of like a soft fork model where it has to preserve backwards compatibility in what you can do because you can add you can't really remove functionality but this is also a very meaningful change for developers. So that's what I came here to talk about. Uh ultimately this has been totally in stealth. You guys are actually the first to hear about this new reactive state. Uh we're very proud of it.

It's been years in the making. Uh it's going to come to mainet uh early next year. So look out for Nebula and the Nebula VM. We think it's going to be a real uh state shift for uh for DeFi. Thank you very much for the time.

[applause]