🎙 AVAX's Emin Gun Sirer Says Blockchains Should be Scalable at Their Core; Layer 2s Mean Giving Up

When Emin Gun Sirer first came on The Defiant in May 2020, Avalanche was not even on mainnet yet. Since then, it has risen to become one of the leading smart contract networks, while the AVAX token has risen over 7x in the last 12 months. We talk about how the project got here and how it wants to create a path for faster and more scalable decentralized network, amid heavy competition from other Layer 1 blockchains.

Emin makes the case for Avalache’s novel consensus mechanism, paired with subnets with their own virtual machines, and he also has scathing criticism for Ethereum and other Layer 2 scaling solutions. To him, once a platform starts building Layer 2s, it means they’ve run out of ideas on how to make a good Layer 1. We also discuss the big picture goal for Avalanche; Bitcoin centers around digital gold, Ethereum strives to create the world computer, and Avalanche wants to digitize all assets. If it's on a balance sheet, Emin says, it can be on a blockchain, and we discuss what it will take for it to get there.

The podcast was led by Camila Russo, and edited by Alp Gasimov and Daniel Flynn. Transcript was edited by Owen Fernau and Brady Dale.

🎙Listen to the interview in this week’s podcast episode here:

🙏 Thanking our podcast sponsors:

• ZerionZerion is Mission Control for Web3: an intuitive DeFi portfolio manager, multichain tracking & trading and the best place to show off your NFT collection
• Matcha: Matcha makes sure you’re getting the best possible price by routing orders across all the various DeFi exchanges, now with an integrated fiat on-ramp. Start trading at matcha.xyz/defiant
• Nexo.io: Nexo is a crypto lending and exchange platform where you can buy cryptocurrencies at the touch of a button and start earning double-digit annual interest that is paid out daily.
• Unstoppable Domains: Go to unstoppabledomains.com and replace your long, complex wallet address with YourName dot crypto, dot x, dot nft or a range of other endings for as low as $5.

👀 Only paid subscribers have access to the full interview transcript below.

CR: All right here we are. I'm thrilled to have Emin Gün Sirer of Ava Labs here at The Defiant podcast for a second time. Last time you were here it was May of 2020. I can't believe it's been that long, but it's been a while lots of things have happened. So thrilled to have you here. Welcome.

EGS: Likewise Thank you so much for having me.

CR: Of course. So, when you were first on the podcast I think Avalanche had just launched a testnet. It wasn't even on mainnet, and this past year for the network was amazing. It's really an explosive year. It's now Avalanche has now become one of the largest smart contract networks. The AVAX token has gained like seven times its value or so in the past 12 months and there is a really kind of thriving ecosystem of applications being built on Avalanche itself. So a lot to go over. Can you go over the some of the most important milestones of Avalanche since?

EGS: Sure there's so much to to cover just on the price point alone. I think the the actual price from July of that year when I spoke to you, in may apparently. In July we had a sale where one AVAX was ¢50 and it is now trading it around… you know I don't I didn't check it. But last time I checked it was around $100, so it's gone up 200X. Not that I pay any attention to the price, but overall so many things have happened, Camila. I think the main thing that's happened is that we introduced to the world a new consensus protocol, one that is so much faster than everything else out there. And I kept saying this as an academic right? I kept saying to the world: “Hey look there is a new thing now and it's so different, and it behaves differently. The messaging is different the message pattern is different and it has the potential to scale. In a way that the world hasn't even seen before, and so a lot of people took it with a lot of skepticism as well they should. And academics, even academics who are supposed to be forward-looking, were were kind of reticent to believe these results, because it was just so out of the left field.

And and now we've proven to the world that this thing is real, that it is secure and it's incredibly fast. There is that… being fast or technically superior doesn't really get you… There are many coins out there that are just one-trick ponies. So. Avalanche is, I don't know how many trick ponies it is, but there are so many tricks that we added into the fabric of the system. So it has a consensus protocol that is incredibly fast. It has a new network architecture that allows anybody to spin off a new virtual machine of any kind. And so we spun off not only the Avalanche virtual machine, but we also created a duplicate of the Ethereum Virtual Machine (EVM). So there is a thriving scene on top of the EVM and there is much more to come.

We have this notion of subnets that anybody can create a new blockchain on top of us, using whatever virtual machine they like, using whatever validators they like. So, that has not really come into its own yet. There's so much more potential in that area that we are about to see in the next coming weeks and months. But the world is beginning to see that the future is multi-chain, and you can create multiple chains on top of Avalanche very very easily.

We also introduced to the world. The fastest cheapest bridge technology ever using Intel SGX, and that bridge is amazing.

I think it has more than $7B worth of assets on top of it right now. It's the most valuable application on top of Intel's SGX technology. In addition to this countless partnerships with companies from Deloitte to financial companies. Countless DeFi partnerships. There's an Avalanche Rush program which created incentives for people to come over and check out the system. A huge number of people come in. There's also another user pattern that has emerged: People move their assets from Ethereum to Avalanche, carry out their transactions on Avalanche, because they're cheaper and faster. And then they go back to Ethereum sometimes. But more often than not they get hooked and they stay.

So what else is going on? There's so much going on Camila, I don't even know, you know where to start or take us.

CR: Awesome. Okay.

EGS: The most excitement, you know, forget all this. Forget the technical prowess which I'm very proud about. Forget the new stuff that we're introducing to the space. Oh I didn't even mention we're introducing new assets like initial litigation offerings and others. That's really exciting. We're we're creating brand new assets that are available only on Avalanche, expanding what's possible with blockchains.

But the thing that we even forget, the thing that gives me the most hope, the thing that makes me the most excited is just how big our user community is. The number of active addresses on our chain has been following a hockey stick growth curve. We are doing 84% of the load that Ethereum is doing. We're doing it at a fraction of the cost because we have a high capacity chain, and we're doing it much faster than Ethereum, and we're growing.

So we're in a very good position, and we're seeing millions of addresses on the Avalanche chain, so there was I think another just another million added in the last month alone. So that's a fantastic situation to be in.

Blockchain Yoda

CR: Awesome! Well, congrats on all the progress and evolution for Avalanche. So before we go more in-depth into all of these different aspects, can you go over your background? For for those who don't know you? You're a real crypto Yoda. You've been in this space for longer than most. You are deeply kind of a key part of Ethereum history itself. Can you tell us how you got here?

EGS: Sure, let's see, I was for 20 years a professor at Cornell University. So prior to that I got my degree from the University of Washington. My PhD degree from the University of Washington. I specialized in systems specifically on operating systems and distributed systems. And prior to that I came to the US from Istanbul to Princeton for undergraduate, and that was kind of funny. It was harder to get into a Turkish school than it was to get a scholarship and go to Princeton. So the choice was obvious then

What's happened, you know, I was always interested in large systems that have a life of their own. In large systems that work perfectly every time, so being able to give a strong guarantee and to stand behind it, to have the scientific background. To be able to say: This thing will always behave in the following manner is was always an incredibly appealing thing to me. It was probably because of the environment that we grew up in. Back in the 80s, things did not actually work. We didn't know how to build reliable systems. Those listeners who remember what a blue screen of death is, they will all know what I'm talking about, but with some probability you would go to a bank or you would go to the airport or you would go to anything and the person behind the counter would be would say: I'm sorry we can't serve you. You know we can't help you today because we're dealing with: The system is down.

It was a rough time, and that's what got me into this area of being able to build reliable trustworthy systems. And that's what I dedicated my career to.

My very first project in this area was in 2002, and there was a system called Karma which was magic internet money that was created by proof of work. It was the first such system ever built, and Satoshi didn't know about it. But a lot of academics knew about it. Didn't pursue that as a replacement for the dollar. The situation post-911 in 2002 was that it wouldn't have been a viable alternative. People were really really worried about terrorist financing.

It couldn't have been done then, but Satoshi came after that. But he also added an amazing additional feature that allowed his protocol to be much more robust than what I had planned in in Karma what I had in Karma. So that's my origin story. Then I worked on making blockchains more secure. I worked on characterizing their scalability. I worked on measuring their decentralization, and I worked on strengthening Bitcoin, making it even more decentralized with something called Bitcoin-NG. Which is actually used in a couple of currencies, by the way, a couple of coins.

I worked on Layer-2s. I built the fastest Layer-2 ever called T-chain, and it gets its speed from hardware speedups, a specialized hardware. And most recently I have been working on the Avalanche system, and I left my position at Cornell. Just, maybe, five months ago now, to dedicate myself fully to the Avalanche system.

Authoritarians: for or against?

CR: Awesome catch up. Emin is also has a part in my book, if you have read it. He was deeply involved with the DAO attack on Ethereum, Emin and a student of his. And solved the bug before it happened. But, for different reasons, they weren't able to act on it. But yeah, Emin has been very close with the Ethereum community for all this time. Vitalik often went over to Cornell and you had hackathons there. There has always been this relationship, so I think that's interesting as well.

So there's this question that I want to start adding to all of my podcasts. How am I defiant? So here it comes: How are you defiant?

EGS: Ah, that's a great question. Do I have to answer it in this context or generally? My day personality is as has been the bane of my existence really through.

CR: In general in general.

EGS: I've tried over the years to sort of calm it, temper it down, and to hide it, but you look anti-authoritarianism is something that's either bred into you or bred out of you at some point. So, you know, society puts pressures on all of us, and you either give in and give up or you just get highly highly attuned to to various forms of coercion.

“Anti-authoritarianism is something that's either bred into you or bred out of you at some point. So, you know, society puts pressures on all of us, and you either give in and give up or you just get highly highly attuned to to various forms of coercion.”

So in my case, there are a couple of things that get me going. I'm deep down a hardcore technologist, and one of the main things that gets me really upset are things that don't work, things you cannot rely upon. And so in every way I want my systems to just work. When I do a payment, I don't want somebody at the bank for some reason to put a block on that payment. When I go to to do any transaction, I want to understand the conditions under which it is going to take place.

And then it should just then take place with 100% certainty. So that's wanting to build those kinds of systems is something that I'm super passionate about, and that's one. And the second way in which I have seen and again this has to do with again your upbringing or my upbringing in this case. So, I think this is, this may or may not resonate with people in the US. But because I do think that this country provides sufficient opportunities to people, but in many places around the world your path to success is blocked by various incumbents. And you really cannot go from A to B, even if you have the best way to get from A to B. If there is somebody ahead of you who has sat on that road then you have to go through that person, and if that person says no then you cannot play. I mean you see so you see some of this.

In academic circles in the US, like there are areas where you know it's possible. It's not really possible to make progress until you pay your dues or whatever. But in many places around the world. All of business is closed off. You can't really get ahead because the incumbents are there. And they have made it their life's mission to extract as much rent from everybody else as possible and they just see this as a normal thing. They justify this. Oh yeah, you know you are a graduate student you overdre your account by ¢15, and we're going to now levy a fee on you of $25, and this is actually a service. We provided a service for you, and it cost $25. And these things are just absolutely horrible ways for people to behave and for corporations, especially for corporations, to behave. And, people who would never do this to you in person, they see it as an okay thing to do because they're hiding behind the corporate veil that gets me supremely upset and supremely defiant. And the only way out is to push these incumbents aside and say: Here's better tech. I have now made you redundant. I have now replaced you with something. That is a contract that is a platform that is a lower level system that obviates and pushes aside your specialized privileged proposition on the path of money flows. So those two things are the core things that get me going, are very much part and parcel of who I am.

Fragile consensus models

CR: Love it. OK, and then Avalanche is obviously a manifestation of that, of creating a network that just works and that can help get those incumbents out of the way. Can you go over how Avalanche actually works, and in what ways it tries to improve on other smart contract platforms and other blockchains?

EGS: Sure, there are two main ways in which it's a completely new approach to the entire space, of how to build fast blockchain how to build blockchains. Number one, it uses a different consensus protocol. All of the other systems either use proof-of-work, Satoshi’s invention, which is so brilliant because it's so robust. Or they use one of the academic papers from long ago. One of the papers that people like me read or wrote. So those academic works, what we call classical consensus protocols, are fragile. This is why you see these systems halt like, some of them, halt on a weekly basis, and people are like: Yeah, that's kind of normal. Well, it's not normal.

Satoshi came and invented what he invented because he knew that these things were fragile, and they also are not very performant because they require all-to-all communication. So these systems, you can engineer them. You can force the players to have incredibly complicated, incredibly expensive machines, etc, etc. Maybe you can increase their transactions per second. But fundamentally because every participant has to talk to every other participant, there is a limit to how fast you can drive those systems.

“All of the other systems either use proof-of-work, Satoshi’s invention, which is so brilliant because it's so robust. Or they use one of the academic papers from long ago… what we call classical consensus protocols, are fragile.”

CR: And that includes proof of stake right?

EGS: Yes, yes, all of the proof of stake protocols use these classical protocols, and every classical protocol has this.

This limitation, where it requires all-to-all communication, Avalanche differs from this because essentially it uses a trick where each of the participants can gain assurance in a decision without having to talk to everybody else. In essence, this is how it works very, very quickly. With classical protocols, think of it like a parliament and, to pass a bill, you need 67 out of 100 people to sign off on the bill and every senator has to check this independently. So somehow you have to accrue all these signatures. And everybody has to hear from everybody else that 67 people signed off, and this works with 100 people. But if you have a thousand people, well it's a thousand people talking to a thousand other people, is a million messages, and it doesn't work.

So, in the case of Avalanche, here's how it works: Imagine now that we're going to tackle a far harder problem. Not it's not a senate of 100 people. Imagine it's a giant stadium of 10,000 people, and but here is how we're going to do it. We're not going to try to put our signatures down on a bill. That will not work at all. So, instead, I ask a random set of people in the stadium if they approve the bill, and so does everyone else at the same time. And we change the way you know whether we approve based on what our friends are telling us. And we do this repeatedly. So I pick a set of random people: 5 of them or 10 of them and say: “Do you think we should do this?” If they say yes then it's a yes for me if they say no, that's a no for me. I repeat this process, and it turns out that after a very small number of iterations I can gain essentially as much assurance as having spoken to everyone.

I spoke to a bunch of people who spoke to a bunch of people. Suddenly it exponentially goes up and we have fantastic coverage of the entire field, and, we can, with high, high assurance, admit that decision. So that's one of the main things that makes Avalanche different. That process does not require all-to-all communication.

It's very fast, very, very lightweight. So that solves the perennial scalability problem with blockchains. It was a very hard, outstanding problem.

CR: And first, what's this consensus mechanism called?

EGS: The Avalanche protocol.

“Not everyone has to communicate with everybody else explicitly or implicitly. You just communicate with a small set, but there is enough overlap in aggregate. You only commit when you know everybody else will commit.”

CR: Oh, that's the name of the consensus. And so this is the way that the Avalanche protocol has of confirming transactions in the network right? Okay, so to catch people up, while in proof-of-stake and proof-of-work and other protocols require all of the nodes in the network or all of the validators in the network to communicate with each other, and for everyone to approve the transactions.

In the Avalanche protocol, it requires validators to ask others whether they they approve transactions, and everyone is asking and validators next to them. And so in this way it doesn't require all of the validators to to actually confirm. Is that kind of how it works?

EGS: It does require all of the validators to come to consensus, but you are essentially right. That is: Not everyone has to communicate with everybody else explicitly or implicitly. You just communicate with a small set, but there is enough overlap in aggregate. You only commit when you know everybody else will commit.

CR: Got it, and that is faster than everyone talking to everyone. Can you can provide a comparison? Like how how long does this process take on Ethereum and bitcoin versus Avalanche?

EGS: Sure sure, so, for example, Avalanche today commits a single transaction in seven hundred and forty milliseconds. Even as I speak that's today's number.

So it's lightning fast. It's less than a second we are measuring our decisions in milliseconds now. Compare that to Solana, which is taking 10 to 20 seconds. Compare that to Ethereum 2, whose spec is not really clear yet etc, etc. It changes quite a bit but.

Ethereum 2, when it comes to pass if, it comes to pass, is going to be making decisions in 5 seconds or more. So it's already the fastest protocol. It's not surprising if you understand what's happening underneath the covers, but it was surprising to some people I guess.

Bring-your-own-virtual-machine

CR: Okay so that's the consensus mechanism, and then I understand there's another innovation which is this the concept of the subnet right?

EGS: Absolutely. So many other systems are monolithic, they have one coin, they have one virtual machine and they have one network. Bitcoin is the best example. Bitcoin does one thing. It does it very well, but it does one thing.

There is the Bitcoin virtual machine, which has its own transaction types and its own upcodes and instructions etc that allow you to write different transactions and then there is one Bitcoin Network. Ethereum: Also Monolithic, right? There's Ether, the coin. There is the Ethereum Virtual Machine, and then there is the network that runs the virtual machine.

In Avalanche, it's kind of like an umbrella, and underneath that umbrella you or I or anyone can easily start a new blockchain when you start that blockchain you specify what asset it's powered by. And you specify what virtual machine it should run, what the rules should be with regard to executing that asset, and then nodes come in and join your network to validate for you, to create your own blockchain.

So that's the big difference between these systems. Why is this important? Well, it is important because there are many many many people out there with valuable things that they want to put in blockchain form. But those things require special handling. What kind of special handling?

“In Avalanche, it's kind of like an umbrella, and underneath that umbrella you or I or anyone can easily start a new blockchain when you start that blockchain you specify what asset it's powered by.”

Well, you might want to build an application that is supremely high-performance and maybe you don't want to run it on a huge number of nodes. You want to run it in a smaller number of beefier nodes. You want a different network, and so you should be able to create that.

Or you have some assets that are subject to US law. And the US law could change at any moment, and you need to be able to accommodate this. nd when when there so there are such changes that happen. You need the network to be able to make modifications to the virtual machine so you don't want it running on the Ethereum network because that's a global network. Maybe you don't want it running on the Avalanche default network, which is also a global network. You might want to create a Camila US network, that is specific to the US. So these are the reasons for why people might want to create their own blockchain, in what we call a subnet.

CR: That's so interesting. So far I understand there's one of these subnets created specifically to replicate the Ethereum EVM right on Avalanche.

EGS: That's right. That's right exactly.

CR: Okay, I'd love to get into that in a bit, but I also wanted to know whether there's been other subnets? And how has this developed so far? What are people using this innovation for? Or maybe has the ecosystem not really caught on to this yet?

EGS: No, no, no, no, so it's new. People are using it, or are beginning to use it, but we haven't seen any big subnet offerings yet, so the time of subnets hasn’t come. People are now understanding the concept. I've been talking about this for three years now. Maybe more?

But it's only now that it's kind of sinking in, and people are realizing: Oh I have these assets, but, before I can put them into blockchain form, my legal department is asking me what my compliance rules are. My legal department is asking me what kind of a relationship I have with nodes in the network, because you need your legal infrastructure to go down to all of the levels. So those folks are busy building new virtual machines either for their usage or adopting existing ones specific to their usage.

At the moment we have one thing that's about to be unveiled, the Spaces VM, very soon and it's an example virtual machine that shows what you can do. It's essentially a key value store in which you can store anything you like, and it can be used for any purpose. It can be used like a name service, but it can also be used to store the data for NFTs. NFT data today is just on centralized websites mostly, and is subject to disappearance at any moment for an essential reason or for reasons outside the control of the creator.

So I cannot wait to show the world how useful these things can be. The Spaces VM is one of them. There are many other VMs that we have worked on in the past, and there are other private VM installations that we have made many many times, but, as far as the public is concerned, subnets are just about to take off. We're prior to the subnet explosion at the moment. The Spaces VM code is about to go public in the next two weeks

CR: And you mentioned earlier a partnership with Deloitte. I'd imagine something like that would be like a subnet would be useful for them. Like is that what they're looking at?

EGS: Absolutely absolutely absolutely. That's exactly what we're working on. You are getting me to leak alpha, but yes, so okay.

CR:That's the purpose of this podcast.

Okay, so what you're describing with subnets reminds me a bit of Polkadot’s parachains. Like, how do they compare?

EGS: Yeah, good question. So parachains and Polkadot, or zones and Cosmos, are kind of similar concepts and that's it though. So of the eight-thousand coins out there, only three of them have caught on to the need for not being monolithic, to be able to create these specialized blockchains for specialized use cases. So in the case of Polkadot, they have only so many slots so they are doing this funny thing where they say: Okay well we're going to be running some parachains, but, because of various limitations, we're going to limit the number of parachains that can exist. So I think it's like a hundred or so? And they're auctioning the parachains. I think that allows them to raise money? But to me it just conveys a huge limitation in the underlying technology.

Why should you have a limitation like this? In Avalanche, you can have infinitely many subnets.

CR: Oh that's interesting and and then Cosmos is then more similar though, but Cosmos doesn't have like a main chain. It's just like an aggregation of all these different separate chains adopting like the Cosmos SDK but there's not like one Cosmos chain.

EGS: Right.

Ethereum and the rest

CR: Yeah, okay I want to ask you now about Ethereum, because I think what we we started to see last year is this separation between EVM chains and non-EVM chains. And there seems to be like two camps right?

It's like layer1’s are either forced to develop their own virtual machines or they decided to adopt Ethereum. Where do you stand on those two camps, and what's your view on these rivaling systems? Do you think the blockchain ecosystem will overcome the EVM, and that will be kind of like old news? Or will everyone have to adapt to play with with Ethereum, which is kind of the the lead right now, in the smart contract platforms.

EGS: Great question. I think we've been called the Switzerland of blockchains. We are completely agnostic. So you are so right that there is this one one camp that says the EVM has won this race. It will forever be the means by which we execute smart contracts.

So that's one camp, and then there's the other camp which is typically like: the EVM is a bunch limitations and design misfeatures, which is not wrong. And so therefore I will invent a new virtual machine with its own design misfeatures. So you end up getting these two camps.

So let me in some academic form mention a couple of the issues with both. So the EVM, if I were to redesign it, I would make so many decisions differently all the way down to the word size, which is just badly chosen. That is, it's not chosen by people who understand how to make things go fast.

That design really needs a performance-minded person to go through it, and redo it. But, you know, that's infeasible at this point at least for the EVM ecosystem. On the flipside, it's very appealing to say: “Hey, I'll invent a new thing that doesn't have the known problems.” But then you end up introducing some unknown problems as well as incurring some some issues with trying to start an ecosystem where there is none.

Tezos did this very very very smart system. It's appealing to functional programmers, but you can see that there aren't that many people who understand the Michelson virtual machine, not that many people who are writing code for the Tezos smart contract platform. There is a growing amount, but it's very very small in comparison to Solidity programmers out there. Solana is trying to do something similar with Rust. That's a slightly bigger community than than the the functional programmers the oCaml types who are going towards Tezos, but again much smaller if there is one regular rust programmer, there are 10X many of other programmers out there. So that number already tells you that this thing cannot have full penetration into the developer community.

Where do we sit? Well we have our own virtual machine. Plus we support the EVM, and if tomorrow it turns out that you know Rust is a very popular platform? It's so easy for us to spin up a VM that supports Rust. In fact, we have such a thing in the background, we just haven't made it public yet because it's not quite ready for production and also that community hasn't hasn't actually emerged yet. There aren't that many interesting smart contracts written in non-solidity programming languages yet. So we could spin it up at any time. We are agnostic ourselves at the moment.

The EVM is has been exactly the right play. Those people who've chosen that ecosystem to grow have been exactly right. All of the exciting latest contracts have been written in Solidity as far as I can tell and it's been a fascinating run. So I am in the you know EVM in the short term and be agnostic in the long term. Be-ready-to–adopt-technology-wherever-it may-be camp.

CR: Okay, that makes sense so are most of the the DeFi and other applications on Avalanche being deployed on this Ethereum like VM.

EGS: Yes, almost yes I think there is like 10X more activity on the c-chain than there is on any other x- or p-chains that we have.

CR: What are X and P chains?

EGS: We have the p-chain, which is the platform chain. Think of it as the meta level chain. People do their basic activities like staking, and also the creation of new subnets. So if you want to go and interact with a subnet, it's from the p-chain that you pick up: where is the subnet who's part of it and how do I communicate with it?

In addition, we have the x-chain and that's a different chain. It's very limited in scope. It's very limited in what you can do in the transaction. So that limitation allows the x-chain to be very fast and underneath it uses (it's going to get a little technical), but it uses a blockchain that's structured like a graph a directed acyclic graph, as opposed to a totally ordered linear chain. So if you have a payments infrastructure then the x-chain is the fastest cheapest thing you could be using, and so that's a nice feature for people who want to build something really really fast. They don't need smart contract features. So if you want smart contracts then we have the EVM, and if you have any other virtual machines we could also add them to the system as well.

CR: Do subnets have any limitations like to scaling or are are they as scalable as the main, the p-chain itself?

EGS: Yeah, the p-chain itself think of it as a subnet. So everything in the subnet for us, the p-chain is just a designated special subnet. The x-chain is also one that existed at genesis. So it's also very well known, and then c-chain also existed at genesis.

So the X P and C, the platform, the Asset X, and the contract C chains are there for everybody. They've existed for all time as far as the chain is concerned, and at any time you can create another subnet and all of these at an equal level.

Maybe except the p-chain. The p-chain is very special because that's how you discover the other chains but all the other chains, the X versus Camila's new chain, they're just chains you have yours I have you know exactly the same capabilities on all of them.

Avalanche DeFi and NFTs

CR: They have the same capabilities. Okay so what kind of things can you do on the Avalanche ecosystem now? Like what applications have you seen taking off?

EGS: Well, there are a couple of different areas I think it's very difficult to even name them and because if I start naming a few and I leave out some others and I get angry messages afterwards. But there's a very very vibrant DeFi scene. Our DeFi has seen an enormous amount of growth. TVL is on the order of, depending on how you count, somewhere between 14 to $21B. So, a very large value locked and very very vibrant community. Most of our transactions on the contract chain relate to DeFi at this moment. I would say maybe 60% to to slightly more would be DeFi.

Our second area of growth is NFTs, and so there are a lot of people who like to issue NFTs, and they want to do it on a chain that is not environmentally destructive and they understand that proof of work chains like Ethereum 1 are just terrible for the environment. Some artists will not touch it.

“Our DeFi has seen an enormous amount of growth. TVL is on the order of, depending on how you count, somewhere between 14 to $21B.”

If you mentioned the topic the first thing they say is oh all that destroys the environment and you have to say no no, no this is not Ethereum this is different. And it's cheap, much cheaper. So our NFTs offerings are a lot of fun. Very exciting things are happening there. And then there's the whole gamut. I mentioned ILOs. I mentioned alternative assets. Overall Camila, we have a different vision than everyone else. Most other people are trying to cannibalize each other. They're going after the same use cases. They're just inward looking and playing a zero-sum game.

My goal in life Is not to try to compete with Bitcoin and replace the dollar. My goal in life is not to compete with Ethereum. Ethereum is just a computer in the sky and that's a great thing. It's a wonderful place for people to experiment. My goal in life is to build the best platform for digitizing all of the world's assets. Everything that lives on Wall Street, that's on a balance sheet, should be on a blockchain. It should be traded around the world easily without any boundaries. That's the vision that we've been building and there's been a lot of really exciting things happening on our C-Chain, especially in the DeFi space.

“My goal in life is to build the best platform for digitizing all of the world's assets. Everything that lives on Wall Street, that's on a balance sheet, should be on a blockchain.”

The latest that just happened three days ago was Platypus, which is like a Curve plus-plus or something like that. It's like Curve. It allows you to do swaps but it allows you to do swaps with minimal slippage. So I'm really excited. The community is very excited.

I think within a few hours, it ended up collecting almost as much TVL as Curve itself. So I'm really thrilled about what's happening on it and you know if you want some examples of things that's the latest one. So we have our own ecosystem where people are experimenting and deploying natively first on Avalanche.

CR:So in DeFi, that’s the main use case and within DeFi, things like exchanges and swaps are I guess that’s wehre most of the activity is concentrated.

EGS: Definitely I think asset transfers of all kinds, meeting of the minds of all kinds. So there's Platypus that I just mentioned. But there's another one called Dexalot and that's a central limit order book. It's like dXdY but we can run it on-chain. I mean it's so expensive to use on Ethereum. It's not feasible for them. But we can do this because we have a high-capacity chain.

And there are many others. There are Uniswap clones, there are OHM clones like Wonderland and TIME and so forth. So there's a lot of things going on underneath the covers. I tried not to name names. So we're going to get angry messages.

New Kinds of Assets

CR: It's better with concrete examples. Okay so as with many other Layer 1s, tons of Ethereum DeFi clones. But you mentioned before that you were excited about assets that are being created on Avalanche that are different kinds of assets that that couldn't be created before, could you get a little bit into that?

EGS: Sure let me give you a concrete example. So a while ago a couple of the people from Ava Labs were talking to a couple of lawyer friends of theirs and they came up with a new asset type that's perfect for blockchains and so it's called an initial litigation offering token. ILO token, and the idea is this.

There are many people out there, especially in the US Justice system who have a grievance. They've been wronged somehow. Somebody did something to them and they need to sue that other person to get back what is rightfully theirs but they lack the means to do so. They just don't have the money to sue and or or they have a budget but the other side has a deeper budget, has deeper pockets. So typically when you're in such a situation the other side will resort to legal tactics to slowly drain you of your money.

At some point they come to you and try to settle and you probably give up part of your claim just to just to just be done with this thing and not spend so much money. So what you can do with an ILO, is you can get funding from the global community for your litigation. That funding goes directly to the lawyers. It's used for legal expenses only and if there are returns then the returns go back to the people who funded the lawsuit.

The very first one of its kind just happened last December. It was oversubscribed and sold out very very quickly. It was an interesting case of a sheriff in California destroying a hemp crop because you know he didn't like the way the plants looked. They looked like marijuana plants to him and now there is a lawsuit to be had around this. The amount of crop that was destroyed is the largest of its kind. About a billion dollars worth of crop. Biggest in US history.

So the person who started this needed financing and that financing was located with the help of the Avalanche C-Chain and the Avalanche community. I'm thrilled about this. This is just a new thing. I mean litigation financing is a very lucrative area but you and I could not participate in this. Suppose you know this now and suppose you want to put your money towards a cause. How would you do it? You can't. I can't.

“...litigation financing is a very lucrative area but you and I could not participate in this. Suppose you know this now and suppose you want to put your money towards a cause. How would you do it? You can't. I can't.”

CR: How much money was raised?

EGS: I think it was $250,000 and I forgot the numbers off the top of my head but it's hundreds of people who donated. And it's happening against a very interesting backdrop. There's national legalization going on everywhere and this is becoming an issue that's becoming normalized and there is this hold out sheriff who doesn't like this and he goes and destroys a bunch of crops without due process and so that's the issue.

But at a higher level, the fact that you can very quickly find funding, the fact that you can very quickly find your community. These are things that the internet brought us. Before the internet there were a lot of fringe communities where they couldn't come together and the internet allowed like-minded people to come together from all over the globe and now they can not only come together from all over the globe but they can put their money towards the same causes. So that's a wonderful situation to be in and it's a very big market.

Spiking Transaction Fees on Avalanche

CR: The value layer for the internet is coming alive and we're just here watching it happen. It's amazing. So all this activity on Avalanche caused transaction fees to spike a couple of months ago. We covered it on The Defiant. So can you go over what are the steps that can be taken to mitigate this?

EGS: We took those steps. So that's how fees went down. So let me tell you what to expect and let me tell the audience, they can check what I say for the scientific validity of what I'm saying, but let me give them the envelopes of reasonable expectations.

It is impossible to build a blockchain that guarantees low payments. Not possible. Why is it not possible? Because if you create a valuable thing, if it has value, then there will be times when everybody wants to transact at the same time. And when that happens you have essentially two choices either, you arbitrarily choose who wins, by dropping some of the transactions and some of the load, you have such a capacity, and you have load over that capacity. What do you do? That's the question. So there are some chains that will say we have free transactions. If you have free transactions that tells me that you have an unused chain. It's so worthless that people are not even attacking you because I could attack you with 0 cost and just consume all your blockspace and now you don't have any more space to put transactions in.

They can't be free forever. That's one. Two, we have fixed costs per transaction. If you have a fixed cost then under load, you will have to drop some transactions. Which ones? Well I don't know and you don't know either and if you're going to do this randomly then some people will be very upset. Some people who place high value in some transactions will be displaced by people who didn’t care as much. So at a time like this, the only sensible thing to do is to auction the space. That's why the fees go up and down on every chain that's real.

So you have to allow the fees to spike when there is a rush to the chain and then you should come up with mechanisms for the fees to go back down. So far so good. There are lots of different mechanisms for that auctioning process.

“...you have to allow the fees to spike when there is a rush to the chain and then you should come up with mechanisms for the fees to go back down.”

My students were one of the very first people to examine this in the context of Bitcoin and also later in the context of Ethereum. And many common schemes used to manage these fees are broken. They typically cause ratcheting, that is fees go up very very fast, but they don't come down.

So in our case, what happened a few weeks ago was again one of these cases where people were doing drops of a token. They're dropping a token for the first time and the very first person to buy it is going to pay a tiny amount per token. And the later you are, the more you pay for the token. So everybody was trying to be that very first user, the very first person to get it and that created this huge load spike and when that happens of course the fees have to go up. They just have to.

Then what happened is we looked at what was happening and we looked at the load in question and we realized that we had parameterized our chain for a very low level of load. That is we were not anticipating that much of a rush and we had parameterized it for a fraction of what the network was capable of doing. So we re-parameterized and we admitted many more transactions per second than we were previously accepting without raising the fees.

Now I think ten days or two weeks after that load event, fees went back down. And they have been consistently far lower than they were at the time of that event or before that event also. So they went from a plateau that's fairly high to a consistently low plateau that's much lower. Do we have spikes still? Of course we’ll have spikes.

When you have rushes to the chain you have to have spikes or else you have to drop transactions randomly which is not an acceptable state of affairs. Then you get angry users saying why is my transaction not working. Why is your chain not working? Instead what we do is say well when there's a rush happening, you have to pay more. That's our load shedding mechanism. That's the only sensible load shedding mechanism.

“When you have rushes to the chain you have to have spikes or else you have to drop transactions randomly which is not an acceptable state of affairs. Then you get angry users saying why is my transaction not working. Why is your chain not working?”

CR: But now the load threshold has been raised on Avalanche. I mean could it happen with more and more demand and activity then that threshold is met more frequently and you have consistently high fees?

EGS: Absolutely. That can happen. So I can tell you the many different things in my back pocket that we are ready to do to address it when it happens again. So you can lower the fees as you introduce new optimizations and we have many optimizations in store. So we've been working on them diligently and consistently ever since the load event but instead of naming them, let me just say this, the team that brought you the biggest thing in distributed systems in the last fifteen years and last ten years ever since the bitcoin paper the team that saw the load event and responded within ten days to lower the fees. Is very much the same team that's going to be addressing the same kinds of problems in the future. So rest assured that we have the capability. This is one of the things I'm really proud of. I do have the most innovative team in blockchains. I also think I have the most extensive team in blockchains.

I've heard of chains where there are only 6 or 7 people working on the core platform. I have more than 50 people working on Avalanche. So we're in a very good situation with where we are and we have distinguished ourselves by using the best of science. We're not just going by gut feeling.

I think everybody saw the EIP-1559 situation where something was introduced that was supposed to lower the fees on Ethereum and it didn't work, just did not work, and did not manage to lower the fees. Maybe it was a mistaken communication. It was a mistaken expectation, but regardless it just failed to do what it was built to be doing. But that has not been the case for us. So all I'm going to say is yeah sure fees can spike. They will probably spike in the future. We will introduce optimizations one by one by one to lower the fees and keep them at an acceptable level and we have a good track record of having done this so far.

“...everybody saw the EIP-1559 situation where something was introduced that was supposed to lower the fees on Ethereum and it didn't work, just did not work, and did not manage to lower the fees. Maybe it was a mistaken communication.”

CR: Got it. Just a note on EIP-1559, I think the plan was always to make fees more predictable not necessarily lower.

EGS: The plan always changes in retrospect though. So it wasn't so much the plan in future respect, but in retrospect it became the plan so I was expecting the fees to go down, I was told to expect the fees to go down. And my whole team was watching intently to see what it would do.

A Vision for Avalanche

CR: Okay, okay. The long-term vision for Avalanche. So Bitcoin has the digital gold vision. Ethereum has the world computer and you mentioned the big vision for Avalanche is to digitize all assets and put them on the blockchain. If there's an asset on a balance sheet, it should be on a blockchain. So if this came true and all the equity, all the stocks and all the bonds and currencies and derivatives in the world came and were dumped on Avalanche, can Avalanche withstand that load?

EGS: Absolutely and if the current network cannot handle it, we will take some of the assets off to a Subnet. So even now for example, NFTs could get their own EVM Subnet if the aggregate load was too much for us. So that is definitely something in our arsenal.

So the fact that we have Subnets means that we can handle many many many parallel flows. So it's a wonderful situation to be in. What people are trying to do with sharding and other complicated technical things, we can easily handle with Subnets.

Layer 2s as a Dead End

CR: Interesting. And that brings me to Layer 2. So there's a meme going around that 2022 is the year of L2, so L222, do you believe that?

Right now what we’ve seen so far with Layer 2s on Ethereum and for listeners as you know, Layer 2s are scaling mechanisms for Ethereum and other blockchains that take some of the transactions off the main chain. So some of the main ones are optimistic rollups, side chains like Polygon, zero-knowledge rollups, all of those. What are your thoughts so far? Do you find them promising? Do you see challenges? What's your take?

EGS: Let me answer it initially sort of as somebody who's heading Ava Labs. At Ava Labs We are completely agnostic and open to Layer 2s. Layer 2s, if they work, then they're great and they would work immediately on top of us.

So every Layer 2 that works on Ethereum will work on Avalanche C-Chain as well and that seems like a wonderful outcome. If you ask for my personal view on Layer 2s that will trigger a rant that touches up on a couple of things.

I've mentioned these before so let me try to organize my thoughts. Very simply I think Layer 2s are the wrong way to approach scalability. When you bring Layer 2s in is when you've given up on being able to write a better system. Blockchains are appealing because they are flat democratic and open to all. If you build the right protocol for your Layer 1 then you have those properties and everybody gets those properties and that's what everybody deals with.

“...I think Layer 2s are the wrong way to approach scalability. When you bring Layer 2s in is when you've given up on being able to write a better system.”

So the moment you run out of space and you run out of ideas for how to grow your Layer 1 that's when you start shedding off load into Layer 2s. Think of Layer 2s as Joe's shack for executing on the side. Every Layer 2 is a shack. It's a weird shack. Each and every one of them has a weird set of restrictions that comes with it. It takes seven days to get into it or it takes immediate no time to get into it but takes seven days to get out of it which should immediately be a huge red flag.

If you're in the shack, then who comes in and out of the shack is controlled by a guy, Joe at the gate. That's weird. Why am I giving up all MEVto some other guy? Why is it that I had this beautiful network and now I decided that instead of executing on that network directly I am now allowing Joe to be in a central position? What do you think Joe's gonna do? Joe is no different than the people on Wall Street were here, holding down every vertical and charging you rent for it. He is going to do the same exact things.

How is Joe not going to turn into Citadel? How is Citadel not going to buy that Joe? So all of those systems if they were to be successful, they would be terrible from a techie perspective and they would be terrible from a retail user perspective. Why is it terrible from a techie perspective?

Because these things don't compose. If you're in Joe's shack and I'm in Jane's shack then I can't just invoke you. Things are happening on different timelines and the mechanisms for communication have been split up. If there is money in Joe's shack that's distinct and separate from the money in Jane's shack. That liquidity cannot flow.

So these are all terrible outcomes. You don't want any of these. If you're a developer what you develop on is what you see when you look below you. And so normally you look below you and you see a nice elegant EVM. And instead now with the Layer 2 approach to scaling, well you know if I'm on Joe's shack I see Joe's weird shack and all the crud in it and then there's the EVM underneath that and if somebody else is in Jane’s shack then it's a different infrastructure.

All of these just limit the ways in which we can interact with these smart contracts. They take us away from the whole vision of blockchains where these things are a flat open democratic world. So I think it's completely the wrongheaded move. It's a move out of desperation. If I was desperate and I ran out of things to do on Layer 1 then I would start pushing people into Layer 2.

But the right way to build systems is to build fast, flat, simple systems open to all, not defer the task of scaling to other people and allow them to reap huge financial gains. They could censor whoever is running the execution engines. These things can cause censorship. And I know at least I understand the situation for Ethereum more or less, I don't really know what happens in the mining pools nobody does, but at least I have some empirical evidence that there isn't much censorship going on. But if I jump into you know Joe's shack or whatever, some layer 2 or other, then I have no idea what's happening.

“But the right way to build systems is to build fast, flat, simple systems open to all, not defer the task of scaling to other people and allow them to reap huge financial gains. They could censor whoever is running the execution engines.”

And when it comes to sidechains and other fancy terms. All of those are essentially Layer 2s. So what's a sidechain? Nobody has a proper definition for it. It's just a Layer 2. And final thing I was going to say about this, all of these Layer 2 projects, they are not trying to expand what is possible with blockchains. They are just trying to carve out value from Ethereum. They're not bringing much in, in terms of business. They typically don't even have business development units. They're just like hey we're running the EVM on the side in a centralized setting.

I don't know what the world wants me to do. I'm not going to go up to them and say whoop de do you get an award. You have reinvented exactly what we had before. I will never say that. I think it's the wrong vision. I don't think we should be heading down the Layer 2 path. I've said this multiple times and this is one of the reasons why I'm building Avalanche and not following along with the Ethereum vision.

The Ethereum vision, Ethereum 2.0, I think is a mistaken direction and it's not going to be that great when it's done, if it's done, and the Layer 2 vision for scaling is I think the wrong vision as well. Just from the elegance perspective alone, if I'm a developer I will not want to develop on it.

Problems with Layer 2

CR: Wow what a strong view. I mean do you think it's possible to create a decentralized Layer 2. I mean right now zero-knowledge rollups haven't been implemented yet. But I think there is some promise there of a greater decentralization and at least there's this, I believe faster way of communicating with mainnet. Maybe there's improvement on bridges. Is there any way to save these or to you they're all doomed?

ESG: So I don't want to categorically say that the entire Layer 2 approach is bad. I haven’t seen a Layer 2 that I am proud of and would stand behind. It's possible that zero-knowledge-based ones are maybe better than the optimistic rollups. Optimistic rollups have a data availability problem and that problem remains unsolved.

“I don't want to categorically say that the entire Layer 2 approach is bad. I haven’t seen a Layer 2 that I am proud of and would stand behind.”

So zero-knowledge ones are better. But then there's a proof time problem. It takes a lot of time to generate the final proof to settle back on-chain. So settlement times can be as long as days long, maybe even longer. And maybe you might have to write your smart contract in a language other than Solidity for a virtual machine other than the EVM. You need to write it for the zero-knowledge prover so that typically requires compiling down to a different bytecode

So all of these are distractions. I think there is a big battle to wage. There are so many assets out there. We need to grow the blockchain space. Those assets are itching to go on to a blockchain and we know how to build fast systems. Why would we want to go and explore these esoteric complicated things?

The Ethereum community has tried to explore every single esoteric complicated crypto invention. Every single one. There was a time when Ethereum 2.0 was going to have verifiable delay functions. There was going to be a battle to see who could build the fastest VDF hardware. Do you remember this? I don't know if you remember this. There was a time when VRFs were going to be a part of Ethereum 2.0. So there are a lot of components that are being thrown into this future vision Camila.

I don't have a problem with this. So if we want to wrap it up, I would simply say look if there's a Layer 2 that works, then it'll work on Avalanche and I'll be super happy. That's wonderful. I don't want to see a world where liquidity is fundamentally broken up into pieces. I don't want to see a world where there is a bottom layer that everybody uses and knows it's there, but nobody really interacts with it because it's too expensive and there's some other layers on top. It's just a messy messy worldview.

“...if there's a Layer 2 that works, then it'll work on Avalanche and I'll be super happy. That's wonderful. I don't want to see a world where liquidity is fundamentally broken up into pieces.”

I would say that most importantly, I think I want to build a world that is open, free and democratic and I don't see Layer 2 s as being any of those things. I think at least the existing ones. Can someone invent something better? Maybe, but if somebody is selling you something that doesn't yet exist. I call that the Waiting for Godot narrative you know, stay with me stay with me very soon Godot will arrive and he'll make our day so awesome. You know, don't go away. Godot is coming and anytime he comes here the world's going to be so beautiful.

I'm paraphrasing from the Beckett play, Godot never arrives. I don't know if you know the play but every day there's a character called Lucky who shows up and then he says Godot is not arriving today and he gives to some gibberish. Nobody understands why Godot is not arriving that day. Every day they wait and every day they're disappointed. So sure we could wait for Godot and wait for some Layer 2 that's magically fixing every single problem and I don't know how they're going to fix the splintered liquidity problem that's fundamental to Layer 2s.

Maybe that's possible. I'm scientifically open-minded and I'm definitely not against L2s. I'm just calling out observational facts that any techie should be able to see with the layer 2 approach. I don't see it as being a winning technological direction. I'm not putting any effort towards that direction at Ava Labs. But as I said if it does happen to work and that's wonderful and we'll have it work on Avalanche.

Belief in a Multi-Chain Future

CR: So if you don't believe in a future of splintered liquidity does that mean that you believe in a winner-take-most-world for blockchains? Or do you think there can be a cross-chain future where multiple Layer 1s connect the bridges?

EGS: I totally believe in the multi-chain future. So the future is multi-chain. These chains never die. So there's no such thing as an Avalanche killer right. We're pretty darn good, at least at the moment there isn't one. But even if you could come up with something, these chains never die. They just don't die. Even the most broken ones that should die. Don't die.

So it is what it is. These chains are not going anywhere and in fact we're going to see a world with many many many different chains. Who will emerge triumphant from this are the chains that can absorb the new growth. Forget the world as it is now. It's fine. It's great and a trillion dollar valuation for Bitcoin is wonderful. but there is 700 trillion worth of assets out there. All of it should be on a blockchain.

“...a trillion dollar valuation for Bitcoin is wonderful. but there is 700 trillion worth of assets out there. All of it should be on a blockchain.”

So as we do a 700x from where we are today, some chains will be able to absorb that growth, with Subnets Avalanches and prime position to do it. Some chains will be able to provide the necessary performance within each Subnet, within each chain. So Avalanche is in prime position with its consensus protocol. We want to absorb that growth and the chains as we know them are not going to go down in value I don't think. I don't think they're going to die. It's just a game for that growth is what we're playing around with.

Crypto Crossing the Chasm to Undeniability

CR: Final question to wrap up on crypto more broadly this year, the macro environment has turned against crypto it seems with the Fed raising rates and monetary policy becoming more restrictive, risk assets go down. For someone building a layer 1, how does a potential down market affect you? And first do you agree that this will be a down market for crypto?

EGS: I'm not sure. I just don't know. I'm a techie, I'm not a price predictor. So I can't give price predictions etc. I can talk about the general macro directionion and how it might impact blockchains as a whole.

This part I think I want to bring some color to. We ended up seeing the price levels in blockchains go down. Why is it going down? That's an interesting question. I don't know a single regular person, a regular trader, person-to-person, somebody I know. I don't know a single one of those who sold during that downtime. What I do believe is that institutions had come into blockchains en masse and the huge rise we got was from that institutional influx and when the tightening noises started coming out of the Fed then we started seeing institutions selling their risky assets to compensate for their leverage requirements for the equities market. So we ended up having crypto assets connected to the equities market because the same actors were invested in both so that when one of them is about to go down, the other one starts selling off the margin requirements that are necessary.

“I don't know a single regular person, a regular trader, person-to-person, somebody I know. I don't know a single one of those who sold during that downtime.”

So that's what I think we saw. Am I worried that we're going to have a giant down market? We might. It might happen. But I'm not too worried about it. More than likely I suspect we're going to have perhaps a sideways market in the months ahead but blockchains are big and a lot of this downtalk, we had a microbear market I think you'll remember, maybe five weeks ago everybody was convinced. There was like a series of three days when everyone was convinced that this was the beginning of a huge bear market.

And the loudest voices were the seasoned old-timers. People who got into crypto fairly early on. They had seen many many many bear markets and they are really really skittish people. They're insanely skittish. So I think those days are over. Days of low liquidity, lack of institutions, those days are gone. I think this area has made it. It's so big that it's a force to reckon with. Even regulations cannot come in to hurt the area because they would end up making a huge number of people upset if they were to be done in a thoughtless manner.

“I think those days are over. Days of low liquidity, lack of institutions, those days are gone. I think this area has made it. It's so big that it's a force to reckon with.”

So I think we're in a very very good situation as a space. We might see a sideways market. We might see fluctuations up and down. We will see fluctuations up and down. These are all part and parcel of the space. But overall I'm super optimistic. There is so much more money waiting to get in than anything else.

And I know this because I talk to these people. I talk to really big funds. There are retirement funds that haven't been able to buy in because the legal infrastructure isn't there, but they're waiting and they've seen what's possible that this is now a legitimate asset class that everybody is trading. Why would they not buy? When that happens you're going to see multiples. So overall I'm super bullish on the space and where we're headed.

“I know this because I talk to these people. I talk to really big funds. There are retirement funds that haven't been able to buy in because the legal infrastructure isn't there, but they're waiting and they've seen what's possible…”

In the short-term, yeah fluctuations are fine, some shakeout is fine. Supremely bullish markets are not all that healthy. They create a lot of froth and you can't tell the true value from meme coins because whatever you buy it goes up. So people start buying stupid things that really have no reason to exist. So I think I like where we are and I am super bullish and I'm super excited about where we're about to go.

CR: Awesome! Great note to end on, very optimistic. I am as well. I think you're right. Crypto has come to a place where it's just undeniable. I think the world has accepted that crypto is here to stay, is a new asset class and there's so much growth to go from here. So it is being impacted in the short-term by monetary policy and so forth.

But in the long-term for builders, that's just a blip. And I like you say maybe even a good thing to take away from all the distractions of the bull market which, the hype was just getting a bit too much I think last year. Anyways Emin, this has been awesome. Pleasure chatting as always. Super interesting and congrats again with all the amazing progress. We'll be covering and watching as always The Defiant. So thanks again for joining.

EGS: Thank you very much for a great discussion Camila. Thank you.