"I'm Not Super Bullish on DeFi. We're Using This Tech to Enrich a Small Group of People:" James Prestwich

The founder of Summa speaks to The Defiant about cross-chain interoperability, building on Bitcoin, ETH2, and about why he has become disillusioned with DeFi.

Hello Defiers! In this week’s interview, I speak with James Prestwich, who has focused on cross-chain interoperability as founder of Summa, and now with proof of stake blockchain Celo. We talked about how against his expectations, users have opted for wrapped tokens, instead of cross-chain swaps; it all comes down to ease of use. He believes in the future, all dapps will have to be built with cross-chain, cross-shard capabilities as a default.

We got into DeFi on Bitcoin, which James says he’s not very optimistic about, as he’s found it’s extremely difficult for developers to build on, and for users to access.

About the common criticism that DeFi on Ethereum is centralized, he believes decentralization is just a marketing term and not a very interesting debate. What people should really be asking is: who has control over the funds and what are their limitations?

In general, he has become disillusioned with the way DeFi works in practice. Stories of financial inclusion are much rarer than stories of somebody raising hundreds of millions in a token sale or people throwing money at the latest yield farming protocol.

Still, he’s encouraged by the fact that it is much easier to launch a financial service than it ever has been before.

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


You’re a paid subscriber, which means you get the full transcript below. Subscribers also get exclusive access to The Defiant’s Discord chat for the community, here’s a new link to join.

🙌 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

Solana Wormhole Hackathon - Up to $200K in prizes

Registration is open until October 27th.

Solana is thrilled to introduce Wormhole, a decentralized bi-directional bridge connecting Ethereum and Solana.

With the introduction of Wormhole and more exciting partnerships on the horizon, the Solana Foundation is hosting an online hackathon bringing ideas and crypto enthusiasts together to create highly-scalable and cost-efficient crypto applications. Solana is written in Rust/C/C++, so familiarity with these languages is a plus.

Join Solana along with leaders from Compound, Gauntlet, Curve, FTX, Balancer, Aave, Audius, Circle, and more to explore and showcase how Solana can help projects propel into the mainstream.

The Wormhole Hackathon will run from October 28th - November 14th, with up to $200K in prizes. Applications are open to everyone worldwide. Don’t have a team to hack with? Join our Discord and team up with the hundreds of solo hackers who have already signed up!

Register Now


James Prestwich: A lot of people have these like crazy stories about getting into crypto and feeling like it's their life calling or something. But for me, it kind of happened because I was bored. At the time, this is late 2013-2014ish, I had just finished college, I was moving to rural Japan. I was mostly doing freelance translation and consulting work for small businesses.

I was living out in rural Japan with gigabit internet like a very small social group. So I started spending more time, like becoming part of these online communities. That's how I got involved in crypto. Was I had an idle interest in this and decided to pursue it more and more over time. So it's kind of like this gradual slide down the slippery slope more than the fall down the rabbit hole that other people seem to have.

Camila Russo: Do you mind telling me what you were doing in Japan in the first place?

JP: Actually, before I got into software, my bachelor's degree is in Japanese literature.

CR: Oh, interesting. That's kind of very two different subjects to specialize in.

JP: There's not a whole lot of overlap there.

CR: So you were living in rural Japan, I guess, driven by your interest in Japanese culture and literature, doing consulting work and so you got bored, and just were online all the time and got slowly slid into this world of crypto?

JP: Yes.I mean, it was a very exciting time for crypto. After the kind of 2013 bubble pop, we went into this long, slow phase where everybody was building things and that gradually led back up to the year 2016-2017 hype.

CR: In what ways were you involved?

JP: While living in Japan, I co-founded a company called Storj, which is based out of Atlanta. I think they're probably 60 or 70 people now. They work on distributed cloud storage, and I worked there into 2017 when I left and founded Summa pretty shortly thereafter.

CR: What led you to found Summa?


Founding Summa

JP: One of the things that I was working on at Storj was kind of this treasury management with crypto is like, you have this company you need to plan to make payroll in dollars, you have assets in five different cryptocurrencies in the bank account. I started thinking more and more about how you can hedge your risk and how you can make smart contracts to do this capital management planning for you.

We founded Summa to work on a lot of those things. We started off building derivatives, options and other kinds of complex financial instruments. We wanted from the start to work with Bitcoin and Ethereum, you know, the two most adopted liquid cryptocurrencies out there, and thought that being able to work cross-chain would be a really interesting and compelling story for the technology.

What we realized really early on in this process is that the cross-chain ecosystem has a bunch of unique problems and user experience issues and that solving those isn't something we can do just for one product. We could build out this whole bespoke solution for one product and it would take years, and we would probably end up with something we were only a little happy with. Instead, we started pivoting more and more into just building out the base interoperability technology.

“What we realized really early on in this process is that the cross-chain ecosystem has a bunch of unique problems and user experience issues and that solving those isn't something we can do just for one product.”

Along the way, we got to work with Ethereum, the Interchain Foundation, Cosmos ecosystem, with Nervos, and Celo and tBTC, and a bunch of other different interesting platforms in the space. That's what we did, basically for the entire lifespan of Summa was go out and try to work with everybody possible to just push forward tooling around cross-chain interoperability.

CR: You obviously have a really good grasp on what that ecosystem is working with, all the different players, so would love to get a state of interoperability overview from you. What are the different approaches that these teams are taking to achieve this goal of having interoperable blockchains?

The State of Interoperability

JP: I gave a talk at the MIT Bitcoin conference earlier this year about the state of interoperability. It's about six months out of date right now, but it's still pretty accurate and descriptive. So for anyone looking for more context after the podcast, I would definitely recommend going to look that up.

In terms of the state of interoperability right now, the really condensed version of the talk, is that a lot of the work that we did over the last two to three years is finally panning out. We're starting to see a bunch of things like tBTC, and the Near Rainbow Bridge go live. We're starting to like see more of these systems in the wild. What we're seeing though, is, everybody is using the light client-based interoperability that we pioneered with Summa and the Bitcoin relay we built. Basically, nobody is doing just simple cross-chain transactions, which is what we all expected back in 2014-2015. So over the last three, four or five years, we've learned really a lot about how people use this stuff and how teams approach it. It's not at all what we expected.

CR: I guess, just straight across blockchain transactions would be, you said, atomic swaps but right now, what people are doing are these wrapped tokens. So you can take tokens from one blockchain, put it in a smart contract and issue some sort of derivative of that coin to interact with another blockchain, which I guess has its own set of tradeoffs; you need to trust whoever is controlling that smart contract. So what's your opinion on how the space has evolved in this way? I mean, do you think it is a good solution? Would you see it go in another direction?

Pushing Complexity Away

JP: The reason it's evolved in this way is it pushes the complexity away from the user. When you go and do an atomic swap the user needs to have a wallet and this wallet needs to have Bitcoin and Ethereum. So already, we've got two sets of keys and two different wallets built out next to each other. They have to fund those wallets with Bitcoin and Ethereum, so they have to go out and buy two different assets, and then they have to send them to this wallet.

There are six or seven major steps to onboard our user, and nobody goes through those. We had something like 100 downloads of our original wallet. Because you need Bitcoin, you can't use MetaMask and because you need Ethereum, you can't use any of the Bitcoin wallets. With atomic swaps and cross-chain transactions in general, you're trapped into using these special-purpose wallets, and it's impossible to convince anyone to use them.

“With atomic swaps and cross-chain transactions in general, you're trapped into using these special-purpose wallets, and it's impossible to convince anyone to use them.”

So this kind of motivates the move toward all of the wrapped tokens and bridging assets, things like tBTC. We're trying to move all of that complexity to a small group of users, people managing these specific nodes that handle wrapping the asset and doing so safely and taking care of failures so that the end-user who only has an Ethereum wallet can just hold tBTC.

CR: Of the current solutions, it was surprising to me how quickly WBTC grew beyond all the other more decentralized solutions. Is this again, because it's easier to use? Why do you think those more centralized solutions have seen greater adoption?

JP: Well, WBTC came to market a lot sooner than almost anything else and that's a huge point in its favor. tBTC has only been out for, I want to say a random month now, we launched in September, so it's still very early days. renBTC has a significant supply as well. So we're seeing building momentum from these raft Bitcoin projects and we're seeing things like Nervos, and Celo launching Ethereum bridges to try to wrap ERC 20 tokens on their chain. At Celo, we're kind of working in that direction slowly. But we're still in very early days of launch and user adoption for these things. I definitely wouldn't say there's a clear winner yet.

CR: That's a great point. You mentioned Celo, can you get more into what Celo does and how that transition was from Summa to Cielo?

Celo Blockchain

JP: Celo is a higher throughput proof of stake chain that's compatible with the EVM. So the goal here is that you can take whatever Solidity contract you've already written and front end for it that you've already deployed on Ethereum and move it directly over to Celo. By doing that, you get access to a native stablecoin, cUSD as well as a higher throughput, lower gas fees, etc. Celo is targeting a kind of mobile-first customer base, so a lot of the work we do goes into building out the Android and iOS wallet called Valora, and so it's not trying to compete with Ethereum directly, it's trying to address a different set of users.

CR: What users is it targeting?

JP: Celo is a very mission-driven organization and so we're primarily targeting users worldwide, rather than in the US. We're doing pilot programs in the Philippines and in Africa and other places around the world, trying to reach users who don't traditionally have access to financial services.

CR: I think those are the hardest users to get and it is kind of the holy grail of or like the main goal of DeFi to provide access to a broader range of people. So how do you think Celo can differentiate itself and actually reach this group?

JP: One of the things I've really been interested in during the brief time I've been at Celo is how many just people we have on the ground around the world. Celo doesn't just talk about reaching these people. We try to send people out there to do things and I've been personally really impressed by it while I've been here.

CR: It’s like onboarding people on the ground, is that the idea?

JP: This isn't my like area of expertise. I'm not working on user acquisition or onboarding. I could go and talk to someone and have a bunch of answers for these questions, but I'm not going to be the best person to answer them right now.

CR: You're still working on getting Celo to become interoperable with other chains as you were doing at Summa, right?

JP: For us, this move to join Celo was a very vision-driven thing. We got into this originally because we wanted to give more people access to financial instruments. So we've known the Celo team for a while. As we were building out all of this abstract technology, we wanted to get back to working on a product that someone would use. We built out a lot of technical showpieces, and some very impressive bits of code, but they're all very far away from end-users, which is where we want to be building.

What we're working on at Celo right now is contributing to the core tech and wallet that is out there helping people today, and working on how we can give Cielo more access to the DeFi ecosystem on Ethereum and how we can give people using the Cielo chain access to Bitcoin. So that's why Celo foundation is partnering with tBTC to bring tBTC on to the Cielo platform.

We're hoping to take these great things that have been built on Bitcoin and Ethereum and help provide access to them to the billions of people out there with smartphones and no crypto experience.

CR: Well, so about Celo, I don't know much about it. How long has it been live and how many users does it have?

JP: I don't have those off the top of my head. It has been live since earlier this year, I don't know the exact date. Fortunately, for me, the chain launched before I joined and I did not have to deal with that particular headache. I have a lot of respect for anyone who can ship software of that size and weight.

DeFi-Curious Bitcoiners

CR: So talking about tBTC and just Bitcoin on Ethereum in general, it's interesting to me to talk about where these users are actually coming from. Is it in Bitcoin holders who are curious about DeFi and Ethereum? Or is it more Ethereum holders who just want exposure to Bitcoin? In your experience, where's demand coming from?

JP: I think we've seen a lot of this from both sides. There are a lot of Ethereum users who would like access to Bitcoin’s liquid markets and the brand recognition of Bitcoin as well. But we've also run into a bunch of Bitcoin holders who are looking over at the DeFi ecosystem and would like to, in a safe Bitcoin-friendly way, check it out. So obviously, there's a lot of interesting things being built and launched on Ethereum and for people who have Bitcoin holdings, this gives them a way to keep their Bitcoin holdings and still participate in this new ecosystem.

CR: I'm interested in your view, you have deep knowledge on both Bitcoin and Ethereum so what are your thoughts on DeFi on Ethereum? There's like this kind of argument that always brings up with bitcoiners saying the real DeFi is on Bitcoin, everything on Ethereum doesn't work, it's centralized. What's your view on that? Can DeFi on Bitcoin happen? What are your thoughts on DeFi on Ethereum?

DeFi on Bitcoin

JP: Building out DeFi on Bitcoin is extremely difficult, and there are a bunch of development and user experience challenges to doing it. I'm not saying it's impossible, but typically, the lead time to launch a Bitcoin product is two to three times longer than the lead time to launch an Ethereum smart contract. I've spent a significant amount of time building out Ethereum and Bitcoin wallets and integrating them and it's much, much quicker to bring an Ethereum product to market. That's why I am not immediately optimistic about DeFi on Bitcoin.

That said, we're seeing a lot of interesting things going on in Layer 2 on Bitcoin and I think we're going to be exploring a lot more of the design space there over the next few years. But it's still difficult to see how you could build a Bitcoin product with the same user experience that we expect out of these DeFi apps.

“Building out DeFi on Bitcoin is extremely difficult, and there are a bunch of development and user experience challenges to doing it.”

CR: Some of these Layer 2s, are these Rootstock or Liquid, what do they look like right now? What’s the ecosystem of Bitcoin on DeFi?

JP: Rootstock is really interesting. It's a whole complex system that I don't want to get in-depth on, because we'll spend the rest of the podcast on it. But I think it's a really interesting attempt to bring Bitcoin to an EVM. But the goal here isn't to bring Bitcoin to an EVM, it's to bring Bitcoin to the rest of the users, to the rest of the applications out there.

You also mentioned Liquid specifically. If you've ever used POA, Liquid is basically analogous to POA; it's a small federation of trusted custodians holding on to the coins.

CR: It doesn't sound like they're really very good replacements.

JP: Yeah. Right now, none of the Bitcoin Layer 2 stuff is going to compete with Ethereum. The work that's being done on Lightning right now is really interesting, but we haven't seen as much Lightning adoption as we've seen of Ethereum dapps.

“Right now, none of the Bitcoin Layer 2 stuff is going to compete with Ethereum.”

CR: Why do you think that is?

JP: Again, like, it's much more complicated to launch because of the technical concerns involved. Because of the way Bitcoin works, it's much more complicated to launch these systems than it is on Ethereum. Lightning has to manage all of these UTXOs, they have to manage the keys and the revocation secrets and all of this complex state off-chain. Where in Ethereum, you just kind of shove that on-chain and let everyone else take care of it.

So the tradeoff here is, in Ethereum, all developers are used to externalizing the costs of their application development to the network. So every time we launch one of these nice big smart contracts, what we're doing is we're making it more expensive to run an Ethereum node. We're having that cost paid by all the full nodes on the network. Bitcoin doesn't let you do that. What that means is essentially, that the developer needs to pay those costs. They need to spend extra development time ensuring that the state gets maintained somewhere off-chain.

CR:That's a really interesting way of looking at it. So would you say that on the other hand, the downside of building or using Ethereum, is that because all of this work is being done by Ethereum nodes, using Ethereum becomes more expensive?

JP: I think there's some really interesting effects here, and this may seem very abstract, but it comes up almost every year. Some Opcode gets repriced in terms of gas as storage gets more expensive. The cost of calling other contracts is going to go up. It's going up, because there are a lot of contracts and a lot of state on the chain and all of the full nodes need to replicate all of those. What we want to do is make it more expensive to add stuff to that state and more expensive to read out the state that's already there.

Sharding Solution

So yeah, the cost of using Ethereum is going up over time and it doesn't seem to be stopping or slowing down. The whole project of sharding and ETH 2.0 is trying to make it so that you only need a small portion of the state. So maybe someone out there needs, for any given part of the state, you want to be sure that somebody has it, and you can get it when you want it. But unlike in ETH 1.0, you will not have to store the entire state yourself. You will rely on other shards to store parts of it for you.

CR: What's your view on this solution of sharding?

JP: I mean it seems fine. I'm sure it'll work in practice. When I started developing on Ethereum, everybody was sure that serenity would be out in 6-12 months and this was 5-ish years ago. So I'm sure that sharding will work when we eventually get it. But I have no idea when that will be, and I'm not making any plans that rely on it.

CR: The timeline says it should be out next year, do you think this time that's more likely to happen or I guess?

JP: I think that if Phase 0 launches on time, then it is much more likely that we'll get sharding next year. If Phase 0 gets delayed again, I'll have to revise my estimates. But it's worth noting at this point that even when we get sharding, the shards won't be able to do anything yet. It's not really expected until Phase 2 or phase 1.5 that we'll be able to use the shards.

CR: Talking about interoperability and sharding, these shards, I guess, are like mini chains, right? The big question there is what happens to DeFi that relies on being interconnected and composable and money legos and all that stuff? How does that work with sharding?

Cross-Shard Communication

JP: You bring up a really cool point, which is, shards are just chains themselves. When we talk about cross-shard communication, we're just talking about cross-chain communication. We call them shards because they were designed to make this easier. The people who are specking out ETH 2.0 are designing the shards with cross-shard communication in mind.

Most of what we worked on at Summa was kind of hacking in communication after the fact for chains that weren't designed for it. So I'm really excited to see a real sharded system launch. We’ve seen so much innovation in this model from NEO and ETH 2 and Polkadot and others. But in terms of the developer experience, I gave a short talk on this at Devcon last year, but I don't think they actually recorded it.

I think that the changes from ETH 1 are going to require us to rebuild essentially the entire stack from the ground up. ETH 2 is not going to use the EVM, it's not going to use the same signatures, it's not going to use the same key custody arrangements or call semantics or encoding formats. So I think it's going to be a year or two after the launch of sharding before we even have a good idea what developing forward is going to look like in the future. I'm optimistic about the ETH 1.0 shard kind of a special case shard that runs that EVM. But if you use it, you have to give up most of the benefits of sharding.

CR: Because it'll be just as slow as ETH 1?

JP: It'll probably be slightly higher throughput, ETH 2 blocks are a little faster and you can probably stick a little more in each one. But yeah, you'll end up hitting the same throughput caps that we’re hitting sometimes today.

The takeaway here is that the difference between working on just one chain and being cross chain is expected to go away. When ETH 2 launches, cross-chain communication needs to be just a normal part of the application. You can't get away with being just one chain anymore.

“The takeaway here is that the difference between working on just one chain and being cross chain is expected to go away.”

CR: Do you mean that will happen just within the Ethereum ecosystem, like cross-shards, or also between Ethereum and other blockchains as well?

JP: I'm optimistic that this is just going to become the default. If I am thinking about cross-shard communication, it is very little additional work to make it work cross-chain as well. The formats can all be the same and the communication models can all be the same and there can be a little bit of different under the hood. But for the average developer, you can make the developer experience for cross-chain and cross-shard communications extremely similar.

I'm optimistic that this is just going to become the default. If I am thinking about cross-shard communication, it is very little additional work to make it work cross-chain as well.”

CR: But I mean, first we would need to have to bring the entire dApp ecosystem to ETH 2 in the first place.

JP: That's going to take a little while.

CR: Especially I mean, since you said we're it's a completely new stack that's on ETH 2, how do you see that playing out?

ETH 2 Uncertainty

JP: It's still in flux. This changes very, very often. Two years ago, it was Ewasm, 18 months ago, it was execution environments with custom languages written in Ewasm. I think about six months ago, execution environments got phased out because of performance impacts. So, even the latest is this switch to a rollup centric roadmap. But you can't implement a rollup without some sort of underlying logic, some VM.

So it's still extremely unclear what this ecosystem is going to look like, what language we'll be writing in, how it's going to plug into the shards. I don't have a firm view on this because the party line changes every few months.

CR: So it's a little bit scary, to be honest. I mean, are you still optimistic on DeFi on Ethereum? Or do you think all these complexities will just force it to go elsewhere?

JP: I think that anywhere it goes is going to face the same problems. There's no magical scaling solution that's going to make a happy home for DeFi forever. So while I get called a critic or a Bitcoin maximalist a lot, I'm still pretty optimistic about the future of Ethereum. I think that we have committed to spending a huge amount of time and resources on this ETH 2 roadmap, and I think that there are some very interesting and difficult challenges along the way. But I'm still pretty optimistic about this.

“There's no magical scaling solution that's going to make a happy home for DeFi forever.”

CR: You said there's no magic scaling solution that will be the right home for DeFi in the long term. But everyone's really excited about rollups right now. Synthetix and other DeFi teams are planning on moving there to the “suburbs of Ethereum.” Why do you think this isn't a good long term solution?

Roll-Ups Surface Their Tradeoffs

JP: Let's back up a step. It's not that I think it isn't a good long term solution. It's that all of these solutions, sharding, roll-ups, channels, plasma, come with tradeoffs.

There's a lot of this cheerleading that happens, where people don't acknowledge the tradeoffs, and they get convinced very early in the life-cycle that this will be the solution and that's kind of the thing that I take issue with, is the uncritical acceptance of whatever the latest cool solution is.

So with my issues out of the way there, roll-ups are really cool and exciting. The thing that's great about them is that they surface the tradeoffs. As opposed to side chains or plasma or stuff like that, roll-ups rely on the base chain for data availability, which means that anybody can go back and look at just the Layer 1 chain and sync a roll-up node from scratch. The kind of tradeoff there is that in order to participate in the roll-up, you have to be running a Layer 1 full node, and if you're not doing that, you're relying on someone else to do it for you.

It's really like the strength here is that we're setting up these partitions, where I can choose to be in the roll-up or not. If I'm in it, I can validate the whole roll-up, and I have a guarantee of that. If I'm not, I don't have to worry about it at all, other people are working on that part. What this partition does is, it allows us to experiment with other state and validation models.

So a lot of people are talking about optimistic roll-ups, witch use fraud proofs instead of direct verification. So in the happy case, that will be cheaper than running on Layer 1. There are zero-knowledge roll-ups which have validity proofs for verification instead of running everything yourself like Layer 1 does. What people aren't talking about much is the UTXO-based roll-ups like Fuel. UTXOs have this cool advantage where they can be parallelized, and you can run all of the block at the same time rather than doing it in order. It means that you can process a lot more UTXOs then account state updates in the same amount of time on normal hardware.

CR: But is that for UTXO blockchains or would they work on Ethereum 2?

JP: Well, that's the great thing about rollups is Ethereum uses accounts, but Ethereum can be used to run a rollup that uses UTXOs.

Inside the roll-up, you're using UTXOs, outside the roll-up, you're using accounts and you can send back and forth. So it lets you choose the state, UTXO account, and it lets you choose the validation model. You can do full verification. You can do validity proofs. You can do fraud proofs. Lets you kind of target the roll-up to the application and the user you're trying to serve. That's why you get such great throughput benefits from roll-ups is, it's not all of a sudden we're scaling, it's that we can target the structure of the rollup to specific use cases and to specific like tradeoffs.

CR: So you have for finance-specific applications, ones where it's important that the transactions are secure because you're dealing sometimes with high-value transactions, and you don't want people to lose their money, which of these rollups are better?

Zero-Knowledge vs Optimistic Rollups

JP: I think it might not just be about like how much value is being transacted, I think it's partially about how people are using it. So for example, zero-knowledge rollups are really great at preventing validation costs from being paid. Only one user has to pay those validation costs. But the tradeoff there is that you have to kind of set in stone the circuit of the zero-knowledge proof that you're using, and that limits your applications. So it's not as flexible.

Whereas an optimistic rollup can be much more flexible, but you're still going to have to do a lot more of that validation work on every roll-up full node. So there's a bunch of different interesting little nooks and crannies to get into here with how these things work and who exactly we're targeting, and what the average usage profile is going to be. Right now, I think we're just starting to launch these mainnet and we are learning an awful lot from things like ZK Sync and Fuel and OVM.

CR: It's good to hear that you think that they can still provide a relatively good solution for scaling considering the tradeoffs.

Througput, Not Scaling

JP: I wouldn't say scaling, I would say throughput. We have this kind of idea of scaling is just processing more transactions. In which case, the most scalable thing is just one server that's really big and beefy does all of this. That the EF runs one node that processes all the transactions really fast. That's the most scalable situation. But in reality, when we talk about scaling, we're talking about how many transactions we can process on specific hardware in this network that we have.

The reason that I wouldn't call rollups scaling is that to get the same security guarantees that you get from Layer 1, you have to run the Layer 1 node and do all of the Layer 1 work and you also have to run the Layer 2 node and do all of the Layer 2 work. So in order to like get the same security, you actually need more hardware.

You need to spend more money to get that same security. So I think these are going to be great throughput solutions. I generally don't call them scaling, because I think it gets a little too sensationalized.

CR: Scaling in a more specific definition would mean higher throughput, maintaining the same costs on the same hardware and in that sense, rollups aren't doing that?

JP: But again, they may not be scaling, they're not letting you do more transactions on the same node on the exact same hardware. But what they are doing is giving you this opportunity to explore different tradeoffs, to explore using the UTXO model, to explore using fraud proofs, and that's a really compelling advantage all by itself.

Decentralization is a Marketing Term

CR: I also wanted to ask you about another common criticism of DeFi by Bitcoin maximalist since you're called a Bitcoin maximalist. But I assume you're not.

JP: I've written a lot of Solidity for a Bitcoin maximalist.

CR: So the really common criticism of DeFi is that it's not really decentralized. That many of these applications are actually controlled by multisigs run by a handful of addresses, that we've seen so many cases where something malfunctions, or there's a hack or an attack, and the team is able to just come in and stop everything and get things in order, which doesn't seem that decentralized. So I wanted to get your thoughts on that. Does it even matter or is this actually just regular finance and not decentralized finance? What are your thoughts?

JP: I think decentralization has always been mostly a marketing term. There's not like this test you pass, and then you're decentralized officially and you get the stamp. People started off back in the day calling Bitcoin decentralized because it was something they wanted out of their financial system. They wanted to get away from this reliance on the government for money.

I saw this back in 2014 with the original round of token sales of people using it as a way to tap into that ethos, right? It has definitely grown that way over the years. I don't really pay attention to the word decentralized anymore, just because of how functionally meaningless It is these days. So if you're saying it's a criticism of DeFi that it's not decentralized, it doesn't really mean a whole lot to me, it doesn't matter.

I think what we should actually be talking about isn't this very binary, is it decentralized or not? It's what kinds of trust assumptions do you have to make in order to use the system? When I put money inside this thing, who has control over that money, and what are they saying about it, and what are the limitations on their abilities?

“What we should actually be talking about isn't this very binary, is it decentralized or not? It's what kinds of trust assumptions do you have to make in order to use the system?”

So, for a very concrete example, I found out earlier this year, that Liquid instead of being 11 of 15 keyholders needed to move Bitcoin around, was regularly falling back to a worse security model, and was to have three a significant portion of the time. So the way Liquid works for a few hours every few weeks, two out of these three keys could take a significant portion of the Bitcoin and just walk away with it. All of these three keys are held by one company. So we were in the situation where the stated security model was 11 to 15 with 15 different companies, but the actual situation that was going on was two out of three with one company and control of all of them.

So it’s not about, is this centralized or decentralized. It's about who can take the money and what can they do with it? DeFi projects definitely fall into a very wide spectrum. So I don't think this is like a fair criticism of the ecosystem, but it might be a pretty good criticism of specific projects.

“So it's not about like, is this centralized or decentralized. It's about who can take the money and what can they do with it?”

CR: Going forward, are you bullish on DeFi with this wide spectrum of applications, some requiring less trust than others, some being more open to users than others, with what your mission or your vision with Celo to provide access to all these groups of people who generally don't have access to many financial services. Do you see DeFi on Ethereum or elsewhere achieving that?

Disilussioned With DeFi

JP: I think, since the birth of the DeFi narrative, I've become progressively more disillusioned with the way it works in practice, so I'm not like super bullish on DeFi as a concept. There are some really great success stories of getting financial access and financial services and financial inclusion to people outside of the US, outside of Europe. But they are much rarer than the stories of somebody raising hundreds of millions in a token sale or people just throwing money at the latest yield farming protocol. Overall, what we seem to be using this tech for is like enriching a very small group of people. So I am not very bullish on the state of things or where things are going.

Since the birth of the DeFi narrative, I've become progressively more disillusioned with the way it works in practice, so I'm not like super bullish on DeFi as a concept (…) Overall, what we seem to be using this tech for is enriching a very small group of people.”

If you look at some of the recent yield farming launches, you have projects like Based where the average holder in the Uniswap pool had $70,000 and you have all of these things which are driven by six or seven-figure deposits, not by average retail users. I don't think we're succeeding at our goals of fostering financial inclusion. I get a little bit upset about that sometimes.

CR: I think it's a really important and fair point. But there is an argument I've heard, which is, this is the early days, early adopters are going to be “whales”, and in the long term, they're building out the foundations for a system that will benefit the individual, retail user if you want to call it that, in the long term. Do you buy into that, do you think it's a fair argument or still no?

Sitting Next to the Money Printers

JP: I think it is really, really easy to say that we will start off centralized and then we will decentralize over time and it is really exceptionally difficult to give up money and power once you have it. So I don't buy the argument that the existing whales will diffuse their holdings over time.

Bitcoin people love to talk about the “Cantillon Effect”, which is that the people closest to the source of money printing tend to accumulate wealth and power, right? What we've done as DeFi, as the Ethereum community, is just normalize putting ourselves at the money printer. When somebody launches a token, and the team holds, you know, 20%, 30%, 40% of the total supply, they get congratulated. So I think it is more likely that we keep the in-group sitting next to the money printers, than we spread out this wealth we're creating.

What we've done as DeFi, as the Ethereum community, is just normalize putting ourselves at the money printer.”

CR: That's such a pessimistic view. Is there anything to you that can save the ecosystem? I mean, what's I'm missing, what can be done?

Things that matter

JP: It sounds like a really pessimistic view, but I'm still here, I'm still doing this work, I still think that this matters. I don't know how to change the ecosystem or people's views on these things and I don't know how to fix things, but I'm just going to keep doing the work and try to improve things where I can.

CR: So what's driving you?

JP: That's a very good question. I care about this work a lot, I care about financial inclusion, and I care about, you know, trying to reach people and improve their lives. So, I would be very bored if I weren't doing this.

CR: Do you believe this is heading, if flawed and everything, to a potentially better place than what the current financial system is at?

JP: Yes, and no. I talk a lot about tradeoffs. We're driving towards something where it is much easier to launch a financial service than it ever has been before where anyone can go out and deploy an exchange or an automated market maker or a derivatives platform. They can do it very quickly and without asking anyone's permission. So that I think is a really interesting and positive change from the existing system.

The kind of flip side of that, the tradeoff is that there are very few protections for buyers. There are very few protections for people who are not in the inner circle, or who don't know how all this works. That's why you get a bunch of these exploited yield farming setups is because there is no one there protecting the retail buyers from throwing money at whatever the latest fad coin is. So I really, really appreciate how much easier it is to work on money than it was even five years ago, and I am wary of the harm that this is going to do to individuals.

CR: That’s a great point.

JP: I don't know how to really address that, though, is the problem. It seems very difficult to educate people about these things.

CR:. Because the system is open, it just lends itself to people just throwing their money at these things. In the end, there is this thrill of gambling and just trying to catch the next big thing, which has been so powerful in crypto and just in finance in general, and it's just hard to control human nature.

Selling to Retail

JP: It's interesting having been around the space for six-plus years to see how things have changed over time. For example, a lot of the DeFi narrative is pushed by a small group of people who got extremely wealthy off of Ethereum. Because they were successful, they believe in this much more strongly than the average person would, right? They are pushing these DeFi narratives that made them successful on to other people even though circumstances are very different and other people are very unlikely, even though the risks that they are advocating may not be worth taking.

CR: Like survivor bias of some sort.

JP: For an example, look at yield farming and YAMs in particular, there was a coordinated marketing campaign from influencers who also put their own money into the yield farms in volume. They use their influential position in the community and their existing holdings from the previous run up to profit off of the credible retail user.

All of this is really difficult to talk about and point to concrete examples, because so much of the info about it is just under the surface, you know?

CR: What do you mean? Isn't it all out there?

JP: Well, the chain is public, but in order to know who addresses are, you need either specialized knowledge or to subscribe to services.

CR: It is true that the way these things work, and what we saw with these “million percent yields,” was because there was a small group pumping up these different meme coins to then sell them off to whoever was willing to buy them at the top.

JP: The really great thing about this is that the yield farming protocols generated their own buyers. When you're putting money into the YAM Uniswap pool, in order to farm YAMs with it, what you're doing is a member of that Uniswap pool is putting a standard offer to buy any amount of YAMs. So it's not just they're pumping this up to sell whoever's willing to buy it, they have structured this to create buyers.

CR: That's right, because you're required to participate in these liquidity pools and once you're in, you have to be a buyer.

JP: Yes.

CR: But I mean, do you think this was done on purpose? Was it orchestrated in this way, or it just like resulted in these kind of forced buyers, but the original purpose of these things was to create something of actual value, at least in some cases?

JP: I don't know, I wouldn’t say that this was intentionally done. A lot of these things seem to be like a progressive evolution, right? You can actually trace this in the smart contract code, seeing different versions of the same contract modified and deployed by different protocols over time. You can see how things kind of went from the Synthetix reward contract which just distributes tokens, and the YFI vault contracts which handle investment strategies. So, the combination of a contract specifically for distributing rewards and a contract specifically for taking funds to try to find the best return combined with these liquidity pools created the situation we're in.

The YFIs are going out trying to find the highest rewards, the end dumping those onto the market. So they're going to naturally go towards the ones with the biggest liquidity pools which leads to liquidity mining. You have these big vaults which are liquidity mining and immediately dumping into the liquidity pool that they're participating in. Anyway, I don't think there was intent to do this. I think that people were combining a few interesting things and found an explosive combination,

CR: Do you think there's anything positive to rescue off of this, any lessons that maybe can be used in a more sustainable way?

JP: Well, I'm actually really interested in some of like, YFI-style things, right? We have this standardized contract that everybody is using to pool funds and to lessen the cost of investment for the participants. Rather than going out and making these 20 transactions yourself, we're going to all pool our funds and make those 20 transactions once and we're going to save a lot on transaction fees. So that I think is like a pretty interesting use case, just this standardized way to pool funds and execute a strategy. I don't have problems with the rewards distribution stuff and I am really interested to see AMMs finally taking off after the last few years of slow growth.

I think none of the tech is particularly bad. I think that we have unfortunately stumbled into a technical situation that rewards people pumping and dumping tokens. But we do that every few years anyway, so probably fine.

Governance Tokens Dropping

CR: I think it already is and hopefully, we'll take the good things. I think, one thing that's been kind of apparent is how powerful these token rewards are and if they can be used to bootstrap liquidity and just like an early protocol but do it in a more sustainable way, like not doing 100x rewards for the first three days and then everyone leaves, maybe it can work in the long term?

JP: Yeah, I'm kind of interested in this. I don't know if they can work in the long term, but we're finding out, right?

If you look at CRV, the curve governance token, it's dropped, like 60 or 80% of its value since launch, simply because there aren't enough buyers to maintain the original price. The actual usage of this token, and what it will be worth in practice is a little unclear and the same with Uniswap. The only thing that this UNI token does is vote on who gets UNI tokens. They notionally have control over 0.5% protocol fee, but it hasn't been activated and 0.5% is not very many percents.

CR: Yes, and same with COMP, I guess.

JP: There's been this explosive irrational exuberance around governance tokens over the last year. So I think that we're going to see a lot of these valuations come back to a reasonable level that accurately prices the value of governing these protocols. For most of these, value is not very much right now.

CR: You need something substantial to govern in order for these governance tokens to be valued the way we are right now. Great. Okay, we need to start wrapping up. But it's been such an interesting conversation. I really enjoyed it. I love kind of your balanced take on things. I think they’re enough cheerleaders in the space, I think it's great to hear some of the criticisms from a very kind of knowledgeable obviously, point of view. So thank you, James, so much for joining me.

JP: I know this is really fun. Thank you.

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.

Subscribe now

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.