"When Incentives Are Gone, What's Left? DeFi Gets Mixed Marks:" Sam Bankman-Fried
FTX co-founder Bankman-Fried and Solana co-founder Anatoly Yakovenko talk about the future of decentralized finance on The Defiant podcast.
In this week’s interview, I speak with Sam Bankman-Fried, co-founder and CEO of crypto liquidity provider Alameda Research and derivatives exchange FTX, and with Anatoly Yakovenko, the creator of the Solana blockchain.
Bankman-Fried, also known as SBF, recently announced the launch of Serum, a DEX built on the Solana blockchain. In this episode, we talk about why he chose Solana and Anatoly explains how Solana works and how it’s able to process tens of thousands of transactions per second. Anatoly hopes that a higher throughput blockchain like Solana means that people can start dreaming big about what they can build.
SBF has become known to be a ruthless trader, but he argues he’s simply using DeFi protocols for what they were made. He says his firms’ involvement in DeFi will be motivated by short-term profits and is not seeking to have a long-term impact in protocols via governance.
For Anatoly, the end goal of Solana is to find the fastest and cheapest way to scale cryptography so that people are able to have ownership and participation in the applications they use, and so that the financial system can become more open efficient.
🎙Listen to the interview in this week’s podcast episode here:
You’re a free signup, which means you get only part of the transcript below. Subscribe to never miss out on any content ($100/yr, $10/mo) by clicking on the button below. Click here to pay with DAI (100/yr).
🙌 Together with Zerion, a simple interface to access and use decentralized finance, Sorare, a fantasy football game with officially licensed cards on Ethereum, and Near, a high-performance proof-of-stake blockchain that interoperates with Ethereum.
Sponsored Post
3, 2, 1...Liftoff: One week to go till applications close for Filecoin Frontier Accelerator!
We enter the countdown for Filecoin Frontier Accelerator, with only one week away till applications close on 15 November! If you’re new to us, our 12 week-accelerator program offers a $20K grant, follow-on investment of up to $1 million, and a global network of mentors from blockchain-native and institutional companies including UNICEF Ventures, Winklevoss Capital and Fenbushi Capital.
More questions? We’ve rounded up an expert panel from Filecoin and LongHash Ventures in an upcoming Ask Me Anything happening on 11 November 7 PM PST/12 November 11 AM SGT. The session will be conducted in both English and Mandarin - come join us to make the most out of it.
If you’re part of the vanguard bringing Web 3.0 to reality, join us:
1) Core infrastructure driving the transition towards a more decentralized, efficient and resilient Web, supporting essential functions such as networking, computation and storage.
2) Middleware solutions that extend the utility of Filecoin beyond the blockchain protocol. Some examples include data oracles, plug-and-play blockchain toolkits, developer APIs, block explorers or search engines.
3) Applications that provide and enable end user interaction, running the gamut from gaming to social media to content marketplaces.
Apply by 15 November.
Sam Bankman-Fried: I was an ETF trader at Jane Street Capital before I got into crypto, on Wall Street and left in 2017, and sort of checked out a number of things, including briefly looking at CoinMarketCap and seeing some pretty big arbitrages and ended up basically diving into crypto trading.
Camila Russo: Driven by arbitrage opportunities?
SB: Yeah.
CR: So you left Jane Street to start a crypto derivatives trading desk. How was that transition?
SB: I didn’t know exactly what I was going to do when I left, but I checked out a number of things. When I checked out crypto, it was sort of with an eye towards, like are there good trades to do here? Are there good arbitrages, quantitative modeling type things that would be worth doing? Basically seemed after looking into it a bit, like there are some pretty gigantic opportunities there, if any of this sort of data was real, and so you end up basically diving in.
Really quickly realized there were, in fact, gigantic opportunities here. But it was not easy to do them. There's a lot of operational work that went into getting set up to do one of these trades, because you had to get together your bank accounts with exchange accounts and withdrawal limits and blockchain infrastructure, and you had to understand the trading and the flow, and sort of a bunch of these things together in order to actually complete even just a single simple arbitrage. So I spent the next month building out a team and building up the infrastructure to do that.
“Really quickly realized there were, in fact, gigantic opportunities here [in crypto]. But it was not easy to do them. (…) So I spent the next month building out a team and building up the infrastructure to do that.”
CR: That was kind of how Alameda came about?
SB: Yeah, that was Alameda starting up late 2017.
CR: Anatoly, do you want to give us your background and what got you into crypto?
Eureka Moment
AY: So I spend most of my career at Qualcomm, study computer science. When I ended up at Qualcomm, I kind of jumped into optimizations, really working on protocol stuff, embedded systems, operating systems. If you ever remember those old school flip phones, I was on the team that built the first developer platform for mobile. They were like half a billion of these devices out there, about 2 billion app downloads, stuff like that.
2017, I was at Dropbox working on compression. With Stephen, who is a cofounder at Solana, we had this silly startup to use crypto mining to offset the cost of deep learning hardware. We're more interested in deep learning, but we wanted to buy a bunch of hardware to build something interesting. The way to offset the capital cost is you might add in the background.
CR: Did it work out?
AY: No. Before we put in enough interesting amount of capital there, I had the idea to start Solana, which was we were getting into the weeds of proof of work, why it's slow, why is it necessary for consensus and I had too much coffee and beer, and I was like up till four in the morning and this kind of like eureka moment for using the same technique that you use in proof of work, but to track the passage of time.
Because I spend most of my career at Qualcomm, time is a fundamental optimization technique for wireless protocols, like time division, multiple access. The most boring radio protocol ever, 2G foundation for cellular networks, that can handle thousands of block producers. So that was it, right? That's what started it for me.
CR: Using time as a core element inside a blockchain was kind of the Eureka moment?
AY: So if you look at like a bunch of consensus protocols, like a ton of academic work, almost in the first paragraph, all of them say, we make weak assumptions about time and this thing, this protocol is still valid, still has strong conditions, strong results for safety and availability. So those weak assumptions about time make it really, really hard to work from an engineering level, like it becomes a kind of an impossible thing.
CR: So the reason why you're both on this podcast is because FTX announced it would be building a decentralized exchange called Serum on top of Solana back in July. So, Sam, why did you choose Solana as your base layer?
A Billion Users
SB: We were sort of looking at what to do, and talking to a number of projects. Once we decided that we want to do something in DeFi, there are sort of two options. One was, do what it will work, and the other was, do what we're most excited about. And do what will work, what I mean is like, do the thing that definitely will at least be fine.
From that perspective, it's like, what's the thing that we're really confident about the path forward for? For a lending protocol on Ethereum or something like that, this sort of thing that people are doing, it's obvious how it plugs into the current infrastructure. It's obvious what role it is. It would just be a marginal improvement on what the current Ethereum DeFi ecosystem at the time was, but is definitely not going to be super ambitious.
The reason for that was, any time we tried to build anything we're excited about, we just immediately exceeded the throughput of Ethereum blockchain by orders of magnitude. At some point, it’s just clear that we weren’t getting around that. Either, we're going to build it on Ethereum, or we are going to build something that we thought was going to be really exciting, but not both.
Then when we sort of started thinking more about it, there's the sense of why half-ass it. If it's worth doing, it was going to be enough of a production no matter what. It wasn't going to be like a one-week side project thing. It's really sad to have a three-month project and release it, and you just know from the beginning, the project's not going to scale and it's not going to be that cool.
But we were excited, and the reason we were excited was that we're excited about DeFi, and the world was excited about DeFi and there's a lot of potential there. So if we really want to tap into that, the thing that really had upside was going to be the thing that had a chance to become huge. What does that mean? It means that you could imagine a world in which a billion people use it.
“…the thing that really had upside was going to be the thing that had a chance to become huge. What does that mean? It means that you could imagine a world in which a billion people use it.”
It's true of crypto and I think it's true of FTX. What's an enormous upside case for FTX, it's the world in which they are a billion traders with accounts on FTX. What's an upside for Serum was going to be a world where a billion people use it. As soon as we got there, we were like, alright, well, we have some instincts about what the fundamental building blocks of a decentralized financial ecosystem would be.
But if we're looking to scale to that number of people, it's not just Ethereum and it's also not just 100 times faster than Ethereum, we need like a million times faster than Ethereum. And that prices out basically every blockchain, and you need not just one of the fastest blockchains, but a blockchain that is built to scale, one where it's going to keep getting better over time.
“…it’s not just Ethereum and it's also not just 100 times faster than Ethereum, we need like a million times faster than Ethereum.”
When we talked to a bunch of different ones, it kind of quickly became clear, alright, one of the core principles for it was if you want to care about throughput and speed and efficiency when building a blockchain, how would you do it from first principles? How would you make it scale natively? For all the others, it was sort of not quite an afterthought but “and also we made it fast.” That's one of our 12 pillars and one of the 12 pillars is not enough to scale. That's enough to get 10 times the current DeFi ecosystem using it efficiently which is cool, but it's not huge. So that was the thing that was really exciting.
CR: Basically, you looked at this space and said, which is the blockchain that is the most scalable, that is capable of really supporting the throughput that we need for getting a billion users on this DEX?
[ … ]
Paid subscribers have access to the full transcript, including sections on:
Solana validator requirements
AY: “A PlayStation 5 already has faster specs than what we'd recommend for people to run a validator (…) The problem that we need to focus on is getting more awesome applications.”
AMMs Vs. order book DEXs
SBF: “AMMs have become popular because they’re a really simplified model of an exchange that doesn't require order matching, and doesn't require canceling orders, doesn't require a liquidity provider in the same way.”
Danger of paying for liquidity
SBF:“In the end, what matters is when incentives are gone, what's left? Did you successfully build a really great product and user base and ecosystem while they were there? I think it deserves mixed marks for a lot of DeFi right now.”
Revenue in DeFi
SBF: “Note that almost every single platform in DeFi right now is not collecting fees. There's this myth that DeFi has revenue, it doesn’t, it's just people lying to themselves maybe.”
Solana’s DeFi
AY: “The Serum order book is a new primitive that's not available in Ethereum, and that's what makes it more exciting to me than just thinking of how do we replicate what's in Ethereum and Solana.”
Beating Web2
AY: “No matter what you're doing, even if you're working on something not related to trading, all these tools can be leveraged to really exponentially outpace what you could do in Web 2.0.”
DeFi villain reputation
SBF: “… it's f*cking ridiculous the central claim, that it's bad or evil or manipulative to use a borrow/lending protocol to lend one out asset in order to borrow another.”
Subscribe now so you don’t miss any of The Defiant content. Subscribers reading this post: Head to posts marked with the little lock to see the full content.
The Defiant is a daily newsletter focusing on decentralized finance, a new financial system that’s being built on top of open blockchains. The space is evolving at breakneck speed and revolutionizing tech and money. Sign up to learn more and keep up on the latest, most interesting developments. Subscribers get full access at $10/month or $100/year, while free signups get only part of the content.
About the founder and editor: Camila Russo is the author of The Infinite Machine, the first book on the history of Ethereum, and was previously a Bloomberg News markets reporter based in New York, Madrid and Buenos Aires. She has extensively covered crypto and finance, and now is diving into DeFi, the intersection of the two.
