Layer 1s, DeFi and Crypto Capital Markets

Dramatic growth in DeFI has given rise to a new generation on-chain decentralized finance infrastructure that is designed to scale to support global-scale capital markets activity. Notably, Project Serum, initiated by industry leader Sam Bankman-Fried of Alameda and FTX, a leading digital asset exchange that uses Circle APIs, launched recently on Solana, a fast-growing Layer 1 chain optimized for speed and scale, co-founded by Anatoly Yakovenko.

With USDC digital dollars coming to this ecosystem, we're excited to have these founders join for a conversation about Layer 1s, DeFi and Crypto Capital Markets. Indeed, this is as much about the here and now as the foundations being laid for a new global capital market and monetary infrastructure.

Jeremy: Hello, I'm Jeremy Allaire, and welcome to the Money Movement. A show where we explore the issues and ideas driving this brave new world of digital currency and blockchains. I'm very excited for this week's show, it's been a big week, all around. Yesterday, we launched USDC for Solana Center, announced as an official chain for USDC. Native USDC launched on leading exchange FTX and in emerging and very exciting DeFi projects, Serum Bitcoin cross 13k, it was exciting to watch that happen on USDC on top of the serum, Dex.

We're going to talk about that we're going to talk about why all this is so important. I think the bigger focus is going to be zooming out and thinking about a few big-picture items that are really facing our industry today. One is really just layer one blockchains, where are we? There's incredible innovation happening on-chain. How do we meet the demands of that, whether it be scaling out global consumer payments or scaling out market infrastructure. Scaling out DeFi really critical topic, we're going to drill into that. DeFi itself on some of the fundamental primitives that have been built by leading firms? What is the future of that? Composability on these different layers of blockchains.

Then ultimately, what are we building here? What do these crypto markets look like? We see a glimmer of that today on DeFi, but what is this infrastructure? What is this all about? What are we ultimately building? What is that going to do for the world at large. We're in the here in the now and many of the things that a lot of us have dreamed about for years are really becoming possible. I think we're now entering an era where this can really truly, deeply begin to change the world.

I love to have amazing people and projects on the show. I'm really blessed this week to have, I think, two of the most talented, innovative people in the crypto industry here today. Really excited to welcome, Anatoly Yakovenko, who's co-founder of Solana, and Sam Bankman-Fried, who's founder of Alameda research FTX and has helped kick off projects here, and welcome, guys.

Sam: Thanks for having us.

Anatoly: Thank you for having us.

Jeremy: There's so much to talk about, I'm going to keep this super simple just upfront. Maybe just quickly 60 seconds on what you're working on, and then we're going to get into a bunch of these topics. Maybe, Anatoly, I'll start with you.

Anatoly: I'm working on Solana, which is this hyper-fast blockchain, and my day-to-day is just trying to make block times smaller and smaller. It's around 350 milliseconds today so pretty proud of that.


Jeremy: That's a good project. All right. I like that project. That's awesome. Sam, how about you, that was [unintelligible 00:03:25]

Sam: Boy. My name is Sam. I am primarily doing two things. The first is running FTX global crypto derivatives exchange. Day-to-day, that's just a really wacky variety of things, from thinking about what new products launch, to responding to irate customers over email. With the rest of my time, I'm shepherding forward project serum, which is a decentralized ecosystem, on [unintelligible 00:03:56], and that also serve a wide variety of stuff. A lot of which is figuring out what things people want or need in order to start building on or using serum.

Jeremy: That's awesome. We're going to dive into that in a whole bunch more detail. I want to start with the first topic here, which is layer one blockchains. I talked to people about this often, and there's lots of different metaphors you can use. We're in this phase of these third-generation blockchains. Solana and very much as a third-generation blockchain, Bitcoin being first-generation, Serum being second generation and things like Solana and others, these third generations. You can help me if you think this is reasonable, but I conceptualize these third-generation chains as essentially like new internet operating system layers that are being built.

People can build on top of that have built-in primitives for executing code and doing transactions and associated data, and doing that, obviously in this open permissionless way. I think for people why is that important? It's like when we move from dial-up to broadband. In fact, in reality, a lot of the problems that we're trying to solve is about throughput. It's about speed. It's about how big are the plates, as it were, to be able to do this. I guess maybe starting with you, Anatoly, how are you thinking about these things, when you think about what you're trying to accomplish with Solana, do you look at it that way? Then maybe just make the case for Solana a little bit.

Anatoly: I got bit by the bug of layer one of blockchain, in a similar way as you described, it's a new operating system. It's a new way to run and deploy code that runs there forever, it never gets shut down. That's a cool geeky thing. The way we've been approaching it is we thought of what is a lighthouse demo or use case that we could really go after that really proves this entire space. Fundamentally, I think trading and price discovery, was like the hardest problem. How do we build a system that's open to everybody to use, where it's fair for people to trade and market make and do prices [unintelligible 00:06:23] such that there is no central party that can screw anyone else.

How do you actually get it to a point where it's transparent and fair? That's really what we've been focusing on is building this project, this operating system to support that because once we can accomplish that, really all the other use cases, which I think might even be bigger than just trading on its own. From social networks, to payments to everything would work there as well. This is like you build windows to run a spreadsheet or a browser, a man everything else is possible.

Jeremy: Having these immutable data structures and programmability and scale, absolutely. I know Jack Dorsey and team are thinking about their own protocols for being able to do decentralized Twitter, and so on, theoretically, you could do that on something like Solana.

Anatoly: What I see now that's really exciting is that I think we've solved this the puzzle of getting information that's replicated on these systems in a very quick and short timeframe. To where you can post the trade and you see it notified and being in the order book almost immediately. That looks and feels like a central system, users are not even going to know that they're up using a decentralized service versus a centralized one. I think this is where innovative teams, like you guys can start actually breaking out into the space of competing with the Visas and Mastercards, the established companies, to where they don't have any advantages anymore. To me, you guys don't now have all the advantages so this has been like my dream, the last few years working on this, is getting it to this point.

Jeremy: It's amazing. This has been the dream, this idea that you could have representations of values, the values, well, that could just transmit at the speed of the internet, and with the interoperability and openness and cost efficiency of all that. Effectively, moving value becomes a commodity-free service on the internet. We're [unintelligible 00:08:31] here. I heard you talk about and you said in your intro this obsession with speed. I heard you 350 milliseconds, that's pretty good for a decentralized. That's amazing, for a decentralized, [inaudible 00:08:49] that's computing and processing transactions.

I heard you talk about wanting to get to effectively, the speed of light on fiber 80 milliseconds, and people who are into electronic markets that's meaningful, because you think about information moving and then how do you respond to that, is that in reach?

Anatoly: This is where I think that this Lighthouse making Serum as good as the New York Stock Exchange, this is what we've been obsessed with internally. If you have a system that can propagate state transitions as fast as the news, it's going to be as good as the New York Stock Exchange with price discovery. You're still going to have these local high frequency trading markets, but all they're doing is taking orders from their customers and then doing statistical arbitrage against a global price point.

Sam will probably have a bunch of nuances of [unintelligible 00:09:43] but this is my understanding of it. We can get to that point, we can actually go faster than fiber because you can build neutrinos that go through the center of the earth at 40 milliseconds. I fundamentally believe that if one of these networks- if Solana is the financial like infras for the entire globe, it's worthwhile to go do that R&D and do that to cut those 40 milliseconds.

Jeremy: We're going to come back to this in a little bit, because I want to talk about crypto markets, capital markets, this concept of markets much much more writ large. It's a huge concept. It has huge implications, but tokenization of information and the ability to exchange value around that. Other things become possible in this realm. You see experiments with that information markets today, but you can imagine that exploding. Being much, much larger.

Anatoly: I mean, you look at like advertisement? That's a tradeable market where people buy and sell pixels, very small numbers. Of being exchanged at constantly. I think that's one of those use cases where if we succeed, we can transform the web from being parasitic to stealing your data, to actually being symbiotic with you, and your environment. How do you consume news? To me, all this stuff is good. It feels like a really good mission to work on this.

Jeremy: Being able to move a fraction of ascent of value for a significant fraction of that. People go, "Why would I need that?" Well, because you can revolutionize the business model of the web. You can do some pretty amazing things with that. Sam, I want to talk about layer two, and your choice in launching Serum and working with the Solana team, and looking at what's happening out there, and saying you obviously have a deep understanding of what markets need to do, and what the innovation needs to be on top of this. I'd love to hear your thoughts and then I'll follow-up.

Sam: I think to echo some of what you guys were pointing at, one place where I definitely think about choosing chains differently, certainly chains broadly understood as including both layer ones or layer twos or whatever. Is rather than thinking what's the biggest pressure point with the thing I trying to do right now, which I do think is important. I think also like trying take a step back and say, "Where are we actually trying to go? What does that mean?" I think the big difference between that is when you ask the first question, when you ask the question of "What's the biggest pressure point? Most people will say, "It's scaling. It is throughput, it is speed."

[unintelligible 00:12:47] isn't scaling enough, and we need to do something new. There's a lot of suggestions, whether it's some other layer, one, any layer, one, whether it's a layer two with some roll-up, whether it's waiting for each 2.0. A lot of those can solve the problem of making you most of the current Ethereum apps, more cost-effective. That's not the end goal. The end goal is you're thinking, "You want a billion people to use this." The end goal is what happens if instead of saying, "Can we do a trade?" You say, "Can we do most of our trading on this chain?" What happens if as you guys are saying, "Could Twitter be on chain?"

It could like, there's no reason it can't, and to expand from this set of narrow things we're looking at now to almost a blacklist and itself, a white list of, what things can't we get there and can we get everything else? All of a sudden, you're not saying, "Oh, shit, we need an extra factor of 10. You're saying, "We need as much as we can get, we're trying to go from a thousand people to a billion." The first thing that I would say is that everyone knows that some house, we need to have something faster. I think it's not just like a binary of faster or not faster. It's how much faster. I think how much faster really matters.

Faster also means two different things, although they're correlated with each other, one of which is throughput. How much stuff can you have on-chain, which is a lot of what this is, but another is also latency. Latency, there are lots of ways to define it. For now, maybe let's just say, latency is like, "How long does it take for you to be like 99.99% sure what a chain did?" For all intents and purposes, you can start updating based on that. I knew that's like another variable there and that also matters, and what scales it matter at it all depends. It depends for different people, but, in general, humans' reaction time, 100 milliseconds.

That's like one guidepost. If something that's the time scale in which humans are really going to be able to tell how much latency something has. There's a real difference between 50 milliseconds and 500 milliseconds and five seconds. More and more of that, you're just cutting out more and more. You say you have Twitter or instant messaging on-chain? Well, it depends at 100 milliseconds latency, yet 10 seconds of latency. I don't really know, maybe. You start say, "Well, how'd you need more, do you need like one millisecond? Do you need a microsecond of latency? The answer is for some things, there are things that are extremely latency-sensitive and some HFT versus HFT trades qualify as that.

There a lot of scientific instruments qualify as that. There's some things that's just going to be very hard to have on any decentralized system. It's just the communication time between different nodes and speed of light is too much. You can chuck that out, but you're still left with probably two-thirds of what the world does. What I'm thinking about, is how can we start to approach that remaining two-thirds of the activity of the world getting on-chain? When we are looking at blockchains, the thing that really stood out with Solana was not just that they quoted a higher TPS number, and that they could actually, in fact, hold that higher TPS, but that it mattered to them.

They're trying to get faster. They had a road map for scaling over time, it wasn't like, "Y'all, congrats. We got 2000 TPS checked that box." That [unintelligible 00:16:37] is not going to be enough forever, even if it's enough right now. Thinking about scaling more, so thinking about squeezing out another few orders of magnitude those all really matters.

Jeremy: The architecture of Solana, obviously, just being able to like leverage Moore's law and HPC.

Sam: Exactly.

Jeremy: Commoditization of HPC and just that constant cycle. You can scale with that growth, which is so compelling.

Sam: Exactly. The whole thing was designed from first principles with actually really caring about this.

Jeremy: People don't, people don't realize how fast that stuff changes. 20 years ago I would be as crazy sounding to say that we're going to have one gigabit at home. I am today saying we're going to have one Terabit at home. literally, there's going to be one Terabit connectivity 20 years from now, which is insane. I can't even imagine what the hell we're going to do with it.

Anatoly: That whole evolution of like the bandwidth on the internet and what the internet made possible, it's a really good analogy here. ISDN was like, we have copper wire and we had dial-up and it was really terrible. Then we're going to use that copper wire and we're going to just like roll-ups on copper wire. It was like, okay, but you couldn't livestream HT video on that. You've gotten these orders of magnitude jumps that are completely different architectures. It just radically changes what society and the economy and humans can do. I think here in this trustworthy compute and transactions arena. That's what we're doing.

Jeremy: Cool. I think the layer two efforts that are out there it ties into what's happening with DeFi now. I think probably all pretty interested in Vitalic original white paper, and a lot of the ideas around DeFi were there back then. What galvanized so many people behind Ethereum in 2016 and '17, and then this real buildout that happened around these ideas in 2017 and '18. Really getting off the ground in the last year were those ideas. I think the ideas are really profound. I look at where we are now. We introduced the HPC because something like Ethereum made that possible. Now here we are in 2020, and there's all these protocols, and there's composability.

We're seeing the "harvesting of that" quite literally, and all this business model innovation around it, which is amazing. It feels like we're, potentially, hitting a wall. I think just at a time when the potential for humanity, and whether it be how people move value around or how markets can function, potentially, hitting a wall. That's why these choices around are we building on a new infrastructure or not become so profound?

Sam: I totally agree. I think something you've said, just really want to echo is it's so to underestimate the importance of future improvements in capacity and speed. You often think about them in terms of what you're doing right now and think that a factor of two is what you need, because you're not thinking about really unbelievable things you could accomplish if you had a factor of 10,000. The internet is not something that people were envisioning as it is mostly 50 years ago. A lot of that, it's not even the structure or anything just they didn't realize really, "Wow, we're going to get so much throughput. They would just fucking put everything there."

Jeremy: Yes. All the world's knowledge, every human communicating instantly, [inaudible 00:20:39] every business connect, all these ways. It's phenomenal. Obviously, you're now, I want to turn actually a little bit in terms of DeFi. You've looked at a lot of innovation that's happening in DeFi, both of you. I think a big project here is saying, "Hey, with this greater capacity what can be built that's different.

Project Serum is, obviously, a lot of people go like, "Oh, it's a DEX or whatever." Obviously, it's way, way, way bigger than that, in the concept. I think that's what makes it so exciting, but it feels like you're building a composable, comprehensive financial market infrastructure. Talk just for a moment about the scope of that vision, and what that is. Then we'll get concrete because you like launching stuff in velocity as well. That vision, that composable financial market infrastructure, talk about it.

Sam: Yes, I totally agree. I think the first thing that came out really, the first substantial thing that came out was an order book. I think the reason for that is that first of all, it is a fundamental building block of finance. It's just all over the place in finance. Most volume happens on an order book one way or another. I think it's a super important primitive thing allows each to just express arbitrary pricing opinions. The other reason is that it's a proof of concept. It's a like, this was the hard thing.

This is the thing that you couldn't really get to work on other chains and here you can just do it. It was finicky, to be clear, it's a real pain to launch because, first of all, just building a pain and exchanges the pain. It takes a lot of work to build one, and it's actually quite a bit harder if you're building a decentralized one because theoretically, it's the same. In fact, it feels a lot more theoretically the same, in some ways on Solana, because it's built to mimic what normal computer programming feels like, to some extent. It's way less forgiving building a DeFi application.

It's not just, you can't fix bugs later, but you should take that to the extreme, and usually when you're launching exchange, you just don't bother thinking, "What if someone tries to submit a bid for $13 quadrillion? Will that crash the exchange? You're just like "No, one's going to do that." If someone does do it, and it happens to bring down the front end, then whatever will put some limit on its like, yes, okay no more than a billion dollars per bid, no one will care. Unlike if that turns out to hit some edge case with [unintelligible 00:23:20], because its tick size is too small, then whatever. Once that'll happen, they're like, "Okay, fine. Yes, actually its number of tokens on," whatever.

There are a lot of cases where you just do something that you're sure will not be catastrophic and will probably work. If something moderately bad happens from it, whatever, you'll put some juristic in later to patch it up. It's ugly and unseemly, and it's also just the truth of how the world works. As you dig into these platforms, you just see they're fucking everywhere. As you're building one of these, obviously they're there. It would take 20 times as long to build a platform. If you had to protect against literally, any possible input sequence, from the outset.

Someone asks for a new API endpoint, and you're like, "Okay, we can launch this in two years. Once we've done fuzzing on this with all possible input sequences and stress tested every single DDoS protection. Okay, let's do something reasonable." If someone brings down this endpoint at one on point, then we have to put in some reasonable protection. The problem with building a DEX, an on-chain DEX is just lose a lot of that. You're like, "Oh no, actually if someone, if it turns out that that's submitting a trillion-dollar offer because the price is insane, crashes the DEXs, because it just has some overflow error. Does that break everyone's money in it in an unrecoverable way. That's the kind of thing.

Jeremy: [crosstalk] little bit higher.

Sam: Yes, that's what you have to worry about. A lot of the trickiness of building something DEX, a lot of the time with just the intricacies of each of those and thinking out, this is the number of bits in this number. We have to make sure that no matter what, as long as it's no bigger than this, it won't overflow any of these buffers. Theoretically, technologically, it's easy. Nothing is impossible there. It just is finicky unforgiving, but you can do it. After some amount of time, we did, in fact, do it and now it's live, there's an order book on Serum. It's basically the only performance, easy-to-use, fully on-chain order book and matching engine in the world.

You can do the math and if your tours magnitudes is lower than Solana's, you can support one of them to get your order book. You use it and if you're a human, it just feels like a centralized exchange. If you're a bot, you notice it it's slower, but as a human, you click a button, a second letter, everything's settled. Yes, it's about response time anyway. I think it's super powerful and fundamental and angry proof of concept. As you said, it's just a start. What really is the end vision? The end vision is what if you got 30% of the world's infrastructure on-chain.

With a billion people using it, what would that look like? What's between here and there? Many, many things, obviously, many layers, but what might the next you be? Well, first of all, just building out all the important primitives, starting with maybe financial primitives, but someone's building a messaging application to it. It works, there's, "Oh, you can go to, see some opensource code. It doesn't have a great gooey right now. It's a little funky, but a lot just focus on financial and infrastructure, but the building blocks of all of this to make it easy for people to then build their applications on top of it. As you said, one of the really huge things here is composability, and why would you bother with DeFi?

There are lots of reasons, I think the permissionless and the trustless is a big part of it. I think another huge part of is the composability is that and totally builds some application. You want to build something that interfaces with it, and you can just do it. You can just read all of its data on-chain. You can interact with, you can send messages to it, and you can compose with it almost as if it's just natively one application without having to consult with him, without having to get his permission. That's just so powerful. I think as all of us know there's this really tempting thing to do these joint ventures with people. Where you're going to build together and they can be really powerful.

They're almost always unbelievably harder than they seem because of the seams in the middle. You spend all your time on like, "Oh fuck, we don't have an API endpoint for KYC data submission and the other side doesn't want to submit PDFs because they're too big. We can make this a JPEG." That's what you spend all your time with, is these little interfacing blobs and like, "Oh, fuck, we're now our DDoS protection is shutting off our white labels and stuff." If you do things on-chain that all just drops away and one really cool, powerful example of this is if you guys haven't tested this, you should do it. If you happen to know how to spin up a surfer is, do you want a DEX?

Do you want a Serum DEX? If so, it can be yours in 10 minutes. All you do is there's open-source, gooey code. You can make whatever modifications you want, put it up on your surfer point to whatever markets you want. New ones are existing ones, point this gooey fever rebates to your address and you're done. You can't do that in CFI.

Sam: Yes, [crosstalk] go ahead, Anatoly.

Anatoly: I had this realization that early on that these networks are equivalent to when we went from analog communication, direct copper line between two people to packet switching and storm forward. We're literally doing storm forward packet switching, but for programs and money and data. That's mind-blowingly awesome.

Jeremy: Yes, it's really transformative. I actually want to come to the bigger picture, how does this [unintelligible 00:29:26] the world stuff in a minute, but just back on Serum, obviously you launch order books, and they themselves that it's an order book engine that is now composable and anyone can build on top of it. USDC is now, obviously, a digital dollar that now can work in this infrastructure, and people can compose with that and program that. Obviously, that becomes a really valuable in those order books. What are the other major new term building blocks?

There's [unintelligible 00:30:00] automated market makers, liquidity pools, you have the pool's concept, which is primitive that people can use in a lot of different ways and what happens with all that.

Sam: I think everyone will have serve their own pet projects. I don't know that my sense is right, but I think it's probably not crazy. What are really the core financial primitives, order book's clearly one of them, some digitized dollar, clearly, one of them. You can't just have everything priced against themselves. What are the other ones, some margin or borrow lending or something like that is super crucial. It just underpins all of finance and there are all these different systems for it. There's structured products derivatives, there's borrow lending, margin futures, repo. They are diverse and have their own twists but at their core, they're all the same thing. At their core, they're all away.

They can all just be represented by a borrower lending protocol. That's the core underlying technology behind them all, is a way to be able to use one piece of capital, take one piece of capital and use it as if it were another one. Lend it out to borrow another one. Whether these are synthetic future capital, whether it's deliverable capital, whether it's some wrapper however you want to put it, and so something like that is really crucial. I know some people are working on some pretty cool borrow lending protocols right now on which I can just power a tone of things. I think one of my favorite examples of the power of DeFi and composability is how'd you build a margin trading exchange.

You have a spot deck, you have a borrow lending protocol, and you're done. You just tie them together. It's not like now you go build your margin trading deck. It's just like a wrapper on top that routes orders back and forth through them, and in one click can send a margin trading order through those. Theoretically, you can do this right now in crypto. You can go to Coin base buy a Bitcoin, send it to WoBe, lend it out, but it doesn't- you got to be a quantitative trading firm if you want to make that thing work. It's a huge production, whereas in DeFi, it's like some teams build a website on top and that just wraps the commands together. That's one of the really core primitives.

What else is there? Some pool object and a pool can mean a lot of different things but at its heart it's a place that you can put stuff and then [unintelligible 00:32:47] token that represents ownership of that stuff. What does that allow you to do? It allows you to create structured products, obviously. There's in some sense a core underlying component of borrow lending of AMMs of a lot of other things. It allows you to do things like yield, you can drop yield into pools or onto pool tokens, and it allows you to have multiple people joining together in the same trade or the same process as they join the same pool.

I think that's a really powerful building block for a lot of these things. Simultaneously, a lot of people are close to releasing instances of pools. A lot of AMMs and things like that. Also, is a generalized pool code so you can customize each little endpoints sticking out of it. I think that's another thing. Then the stuff around the edge is plug into the rest of the world. Whether you're talking about cross-chain stuff, whether you're talking about [unintelligible 00:33:43] on-ramps, off-ramps, somehow, it has to touch the rest of the world. I think stable coins are one example of that.

One way to do that is you can send dollars and [unintelligible 00:33:54] and send it to Serum. Another version is you can interface with [unintelligible 00:34:01] or Bitcoin blockchains or centralized exchanges that are supporting SPL tokens, but things that plug into it to let you access the system. That's another really crucial thing, and once you have those, almost everything that people are looking to build you can go ahead and build. You can put those together, put tweaks on them and just start releasing. I think as those core pieces start coming into place what you're going to see is that just composing those together allows a lot. It allows a lot of what we people want to do.

Jeremy: Even where you are today, it's just mind-blowing. Where anyone can create a market and, theoretically, create a market in anything. It can be a dollar price market and that's like in the [unintelligible 00:34:53]. When we talk about information markets and other things that go beyond people who are building financial primitives, but actually creating markets much more broadly is super exciting. I guess like that's like a bigger picture kind of question which is like we're talking about market infrastructure, we're talking about capital markets more broadly.

As you're saying, Sam, how do we move two-thirds of the world or how do we move a third of what is the financial system or some huge amount onto this infrastructure. The big question is why bother? How is this going to improve the world? What are the fundamentals, I have an intuition about that. Obviously, I have a lot of thoughts about that but ultimately, these market infrastructure innovations should create fundamental improvements in microeconomic market structures in macroeconomic market structures in the nature of what even a corporation or a collective vehicle of corporate form looks like. What's the why bother in both your minds about this?

Anatoly: I can go first. I guess I have a very- this is probably two reductionists of a view, kind of doesn't capture a lot of nuance, but you look at [unintelligible 00:36:27] reports or McKinsey reports, finances like $20 trillion, some 20% of the world's GDP just moving numbers around. If these systems that provide sufficient enough trustless censorship-resistant computing can take a chunk out of it. I think the cost will reduce dramatically, not by 90%. That's, literally, giving everybody in the world a day off without hurting their standard of living. I'm down for that. I want to work four days a week.

Jeremy: It's like, fundamentally, returning economic value to the economy or whatever you want to describe that as.

Anatoly: There's no reason why it should cost more than one like a few electrons to update numbers in the computer. No matter what those numbers are like how many billions of dollars to represent.

Jeremy: Sam, I'd love to hear you tell talk about this. Just the substance of what changes in the economy when we have a world like this.

Sam: Maybe one analog for this is something that's happened in the fund industry over the last 30 years. ETFs and mutual funds and things like that. I came of age as we were already transitioning to the new standard. 20, 30 years ago, when you had some money and you're going to invest it in the market just, generally speaking, who knows what. Maybe some broad market fund, what fund would you put it in? What fees would you pay? How much middleman taking away would there be? Who knows it could be anything? What would you [unintelligible 00:38:15] go to your financial advisor, or click on the one button that your portal has and have some random fee. Say a bit the year at 5% a year.

I don't know. It is what it is your, unlike your financial advisor, "I recommend this fund for you. I really think I have confidence in it." You're like, "All right, fine. I don't understand any of these names anyway." I just have an extra 3% a year [unintelligible 00:38:37] on in fees and so be it. Is that the fair cost us through the [unintelligible 00:38:41] notes? There are all these industries that existed in finance that were just products of this. All these subdivisions of investment banks, that their only business, the only real business was how do we create some PowerPoint presentations around this product we're trying to launch. To convince people to buy it even though its fees are an extra 2% a year. How do we convince people that whatever, Morgan Stanley's special opportunities, blue-chip fund is what they should put their money in despite the high management fee?

The reason is they even talk about the high management fee. They talk about the special opportunities. No one even knows what the management fees are supposed to be. Then, what happened? What happened was the internet happened. I don't think is obvious stuff at the time, but it's really fucking brutal for this business model, like really bad for it. The problem is they're just websites. You can go to a website now.

Jeremy: [inaudible 00:39:43] zero and information asymmetry was blown out of the water.

Sam: Exactly. You just type in S&P 500 ETF. You get the list of them. You're like it seems there's 20 of them that charge 10 [unintelligible 00:39:53] a year. Mine searching 2% a year, and I see [unintelligible 00:39:56] at place number 30,000, maybe I should not buy this thing that my financial advisor said, I should buy, and maybe just buy like VOO, instead. Just over a decade, completely killed these funds that never served any purpose except for overcharging people because there's just this global trusted database of all the options, and just a number in the end management fee. Now, you just can't really get away with high management fees.

Yes, there are things you can do. People are still finding ways to upcharge people, but it actually did take 80% of the money out of that game for getting people to invest in funds. Each year you can look at open interest graphs for ETFs and each year, it takes a little bit more in of the low-cost funds. Vanguard's just slowly going up. That's great. You cut out a ton of blue. That's very similar to a lot of what- you get by or what you could get by putting things on-chain is get an open ecosystem. You get an ecosystem where everyone has access to everything and everyone can see everything and just look at it.

The equivalent management feature, is just a number on-chain, and you can look at it and be like, "That's too big. Why don't you use this one instead?" You can compare the system you're using to the system other or people are using. I think it has the potential to take a lot of the duplicative work that the world has right now. Where people rebuild the same thing over 3,000 times, not to refine it, but to be able to rent seek on it in their own way, and to cut through a lot of that.

Jeremy: What do you think it takes to, tokenization is this bigger theme, and if you have a market infrastructure like this, it becomes intimately more valuable if the tokens that are on it or are ownership and/or assets or other things. Future cash flow of a business or [crosstalk] project. How do you see that maturation process and, what are the markets that are realistic to be on this now versus in 12 months, 24 months? 36 months?

Sam: Yes. It's a good question. I'd say a lot of things could be on it now. Not all businesses, but many businesses could be, but realistically, most won't because it's just the tooling around it isn't built out enough that as someone who has no idea what a blockchain is, you can just go click to put your business on a blockchain button. No, you need to have blockchain developers and stuff. That's just a threshold that most won't cross. It's going to be random. Which businesses are on?

Now, if you're building on Solana, it has the capacity to support a lot of businesses. It's less going to be, could it be, and more just going to depend on, does your particular company happen to have a bunch of blockchain developers on it? If so, then yes, maybe it will be, and if not, not yet, check again next year. We'll see what happens over time, but the goal is that over time, this just starts growing in snowballing. Each year, another 1% of all companies build out the infrastructure and support. It's not that hard.

Especially for a giant company we're talking hire a team of five people and they can do it. You have to actively decide, "This is one of our corporate priorities." Each year gets easier and more popular, the goal. Then, the other thing that I'll say is one thing, which is, I don't know if it's a downside of on-chain stuff, but it certainly is a pitfall, is that it's so easy to mint a token.

At some point, I think a lot of people in crypto have had this light bulb go off. The light bulb is all this jealousy towards other people who have a token and you don't have a token. Then, you're like, "What if I want to make a token, how hard would that be? Wait, is it really five minutes?" I just Googled it, and here's a website. We can click and make a token button. Does that work? You're like, "Oh God, that works. That's what was stopping me all this time. I didn't Google mint token please," and then people just go wild.

Jeremy: Well, this is the composability thing of if you have stable coins, you can actually have the treasury of a business, and all of its flow. You can have a contract with it where it's got output to you, token holder. You can put voting and governance. What is a corporation, after all, but [crosstalk] treasury and it's a voting system, and it's a contract for token holders? That's the innovation curve. We need people who are thinking about building these "digitally native corporate firms" that exist on global internet. They exist on-chain. They exist in the enforcement, if you will, of that is just, is done through smart contracts, and we have the building blocks there. I'm not sure [crosstalk] the big companies first, it's probably going to be more, yes.

Sam: It's going to be small and medium-size companies that have someone high up, who's excited about this, at the beginning. One of the most amazing and powerful things is how relatively way easy to do this. One of the pitfalls is that it's easy to do this, even if you don't have anything. We've definitely seen a little bit of that this summer where people realize, "Oh, I could just mint token tomorrow," and people don't go through the work of maybe I could create a useful, valuable project and then put a token in it. They skip that first step. They're just like, "I'm going to have a token, no one has grabbed tacos yet. I don't know why, no one claimed the taco token. I'm launching the taco token. We're going to figure out what it does later."

It does mean that, it's really easy to mint vaporware. I think that's something that this space is just going to have to contend with. It's a really fundamental property of it, the openness and you can deal with it. It is not fatal or anything, but it does mean that people just have to be a little bit cognizant of that, and figure out how to have some trust systems in a world where the barriers for building something that looks like a product aren't so high, that only a legitimate company would be able to do it.

Jeremy: Go ahead, Anatoly.

Anatoly: I think even that is also something that's really remarkably innovative is the ability to organize people around a fungible thing is governance. It seems like very vaporware right now. I remember six degrees in those super early like internet social networks that just seemed like, "This is cool, but I don't know what to do with this. I don't know why I care if I'm separated by two degrees from somebody famous." I think these kinds of experiments are going to be doing something interesting. When we get to 300, 400 million people all organized by single token, I think that's going to be probably worth more than Facebook for that many users. There's something really cool.

Jeremy: Yes, I think that's what I'm excited about is, we're very likely to see, I use the word corporate form, but basically, new corporate forms that people participate in and that they have an economic relationship to, and that it organizes work. It organizes creative output. It's all run entirely in software, on-chain on this infrastructure. It seems like now, the building blocks for that are falling into place. All right, cool. We're running out of time. I have to ask a USDC question, which is, I think you guys have all seen USDC grow, and it's been a big year, and we've talking about all this infrastructure opening up. All these opportunities opening up, something USDC, it's 3 billion today. In 12 months, 24 months, where do you think USDC will be?

Sam: Yes, and it's a good question. I think a lot of people really radically misunderstood the stable coin business model. Do you remember, what was it? It was week 2018, but month of stable coins when everyone is doing nothing but talking about all the different stable coin companies? It made no sense because, first of all, everyone seemed to think that you get $150 million under management as a stable coin company. It's the best company in the world, but the profit, the margins are not big on these things. That's not worth that much. Also, everyone seemed like allied to the question of, "What are the advantages and disadvantages of various ones here?"

You made a stable coin and all of a sudden you were a hot company. I think that it missed' two big things. The first is that all the really huge outcomes here, aren't getting huge. How does a stable coin become extremely valuable? It's when you start measuring it in billions and tens of billions and hundreds of billions. It really isn't the scaling of it. Second of all, people try to be way too cute with the model it, and the structure of it. People are like, "Oh, my stable coin is algorithmic and you can only create it on Tuesdays. There's just lots of ways to make it a little bit worse.

It's like everyone had their own thesis on it, but if it's a tokenized dollar, there's just an obvious way to do it. A lot of slight improvements you can make that are really intentionally thought out, but random perturbations, just make it worse. What's the fundamental thing here? The fundamental thing is just building a good product. Making it work really well and going back to the scaling thing, making it, able to massively scale without friction. I think that that's always been USDC's calling card has been, "How can we make creations and redemptions of it as easy and cheap and fast as possible? How can we make sure that there's as little bullshit as possible, that slows it down so that it can just seamlessly scale up?"

Anyway, it was this really mysterious month of everyone building products that were obviously later to the game and 20% worse than USDC. Trying to become the hottest crypto company with no roadmap, Anyway, so where could it get? It's really hard to tell. I think that one thing Tether has shown is that most people haven't thought hard about that question. Every year, Tether breaks a new record for open interest on say, it's holy fuck, billions. 5 billion, 10 billion, 15 billion. What happens if Jeff Bezo wakes up one day and decides he wants to put himself on the blockchain, might go from 15 billion to 50 billion overnight.

Jeremy: The block size is no pun intended. They get they're really changing.

Sam: That's right.

Jeremy: The step function changes of the types of capital that get deployed into this form factor of money. I don't think people quite see whatever dollar money markets or how many trillion dollars?

Sam: Yes, exactly. It's the same thing at TVL&D five people are shocked as it kept getting bigger. I don't know, guys you're paying a million dollars a day. How much TVL are you getting for that? A lot? You need a billion dollars at TVL. It's not that hard. If that's what you want. Similarly, we hear this isn't the paying for it, but they're being demand for it. Yes, how much demand is there for liquid money things in crypto, of billions, hundreds of billions, maybe trillions at some point.

It really depends, it could be anything, and I wouldn't be shocked if it shrank a little bit. I wouldn't be shocked if it added two orders of magnitude and over the next year or two. What's really going to just determine that is just how much demand is there. How fast is the crypto ecosystem growing? How many people are trying to get into how much money is trying to get into it? It was a quarter of that can end up in stable points. I don't know something like that.

Jeremy: We're seeing tons of people just doing investments in USDC. We're just seeing so many people, the receiving firms and others. Now, obviously, with USDC on these new platforms and just, obviously, like Solana, a 300-millisecond digital dollars. Pretty awesome, so we're excited.

Sam: Yes. I'm super excited for that and I think, obviously, I'm a little biased here, but I do think it's great for USDC as well. From thinking about, from the perspective of how would you position yourself as the stable coin? I think he's done a great job of just making it really easy to use, but I think also it sure would be a shame if the answer is the scaling platform ended up centering on some other stable coin. I think making sure that you look at the self like plausible routes to where a lot of this could end up, and that it supports those is really key.

Jeremy: Yes. That's awesome. Any closing thoughts from you on this, Anatoly?

Anatoly: I think people often talk about these big companies coming into the space, like Visa, MasterCard, or even Facebook Libra. I often think that it's the innovators and the early adopters. You guys that are going to take these very obvious [unintelligible 00:53:59], a dollar that's cheap and fast to use and make it global. I am just super excited about the boring stuff. We can do point of sale now under two seconds, but that's trillions of dollars get moved every few months.

Jeremy: In this case, explosion is the awesome thing and that's just where everything's clicking into place. It's awesome to collaborate with people like you guys. This is a lot of fun, so it's going to be an interesting couple of years.

Anatoly: Awesome. Yes.

Jeremy: Listen guys, thank you so much for joining today, and looking forward to all the stuff we're going to keep working on. Awesome. Thank you.

Sam: Yes, of course.

Jeremy: Cool. Great show this week. I'm really grateful to have, Anatoly and Sam, and excited for next week. Until next time, stay well, stay safe, stay informed.


Sam Bankman-Fried
CEO, FTX and Alameda Research
Anatoly Yakovenko
Co-founder Solana

Get more insights from The Money Movement in your inbox. Sign up today.