CoinFund is one of the earliest crypto-funds to have taken shape, forming in early 2016. The fund is well known for running an active Community Slack, conducting great podcast interviews with cryptocurrency projects and an emphasis on building network nodes / services.
We are joined by Jake Brukhman, CEO, and Alexander Bulkin, Chief Alchemist, to discuss their latest thoughts on investing in the cryptocurrency space. We cover a wide variety of themes such as their opinions on the “fat protocol hypothesis,” thesis on value capture in the cryptocurrency space, their efforts to build network nodes; and their effort to build an open source token-less technology that allows entrepreneurs to launch their blockchains.
Topics we discussed in this episode
- Coinfund’s history
- What is a cryptofund?
- Coinfund approach to investing in the cryptocurrency space
- Generalised mining – what it is, and how it offers cryptofunds a competitive advantage
- The ADAPT toolkit – a tokenless toolkit for rapid blockchain innovation
Meher Roy: Hi, welcome to Epicenter. This is a Cryptocurrency Podcast that interviews academics entrepreneurs and thought leaders in the cryptocurrency and blockchain technology spaces. I am Meher Roy and today I am really pleased to have a new host join me in this interview. That host is Friederike Ernst. Friederike, welcome to the Epicenter team.
Friederike Ernst: Hi, good to be here.
Meher: So, tell us a bit about your background Friederike and how you started in the blockchain space.
Friederike: Oh, yeah sure absolutely. So I am actually a physicist by training, so I studied physics and then did a PhD on Low Dimensional Complex Quantum Systems, and I quite enjoyed science, so I stayed in science for a little bit longer, after that, I did a post grad at Columbia University and then a second one at Stanford in SLAC, but all the while I had actually become pretty interested in the blockchain space, so I heard about bitcoin for the first time over 5 years ago in 2013, and at first I thought this is a load of, whatever, but it was enough to actually get me interested, and I started thinking about what money is and what actually makes money valuable and the social contracts that actually underlie all of that and at a similar time, two friends of mine started Gnosis, so I was actually close to Gnosis from the start and after my time at Stanford decided to join them fulltime, so I became the COO. This was before the token sale, so early last year, and I have been in the blockchain space fulltime ever since and have never looked back.
Meher: Wow that must be quite a switch from low dimensional quantum physics to cryptocurrency.
Friederike: Yes and no, so I was an experimental physicist, I always really enjoyed threading things and blockchain, I mean it’s a completely different subject matter, but a lot of the things that you do on a daily basis you question hypothesis, you build test systems, you think about what could possibly go wrong and that is actually very similar as in physics.
Meher: So, Friederike tell us about some of your interests in the cryptocurrency space and what kind of episodes could the listeners look forward to from you.
Friederike: I am super interested in business models that were impossible before hand, so basically capturing unused potentials in a way by actually making things peer-to-peer, that’s one of the things. I also love governance, I mean governance mechanisms and I think governance, I noticed that it often goes wrong in politics for instance, but then when you actually sit down and think about what could be done better, it’s actually pretty hard to come up with a decent system, so in hindsight I actually have to apologize in my head to a lot of people that I didn’t really say any of those things, but I thought this is a system that was devised in a weird way and actually thinking about these things I find really pleasurable and asking myself how would I design this if I had to design this and kind of blockchain gives you a platform for doing this. I also really like deep tech so things like scalability and the sort of load of other language that you use for uploading stuff, so I am interested in just the actual bidding blocks and how they all fit together as well.
Meher: Cool, so I mean I am really looking forward to recording quite a few episodes with you Friederike. For our listeners, please do drop Friederike a note as an iTunes review or to one of our email addresses, so feedback would be really good to get her started on her Epicenter journey, thank you.
So today we are talking to Jake Brukhman and Aleksandr Bulkin from CoinFund. CoinFund is one of the earliest crypto funds to have begun business and therefore it’s one of the oldest crypto funds, so we are pleased to have Jake and Aleks on this show. Jake, welcome.
Jake Brukhman and Aleksandr Bulkin: Thank you Meher, pleasure to be here. Thank you Meher.
Meher: So Jake, of course we’ve known each other for a while now, we met at last Devcon and I am familiar with your story, it’s a very interesting story, so for our listeners, tell us about how you came into the blockchain ecosystem and why did you start this fund?
Jake: Awesome, thank you. You know, I am a technologist, I worked on Wallstreet in Hedgefund world for a long time. I’m the kind of a person that is an early adaptor of technology, I got my first bitcoin sent to me by a friend of mine in 2011 and at that point I don’t think I fully understood what it was, but I definitely started paying attention and then around mid 2013, I bought some bitcoin on Coinbase in order to just make a random investment and hold that and see what happened and that’s what really got me engaged in this space and then you know, it wasn’t really until the Ethereum white paper came out and I read that and I started to think about what does it mean to have a world of many different kinds of digital assets and along with Aleks kind of said hey, maybe we should create a portfolio that diversifies the risk across all of these different kinds of assets that are likely to appear given the advent of a platform like Ethereum and so that became basically the basis for CoinFund and we launched this kind of a proof of concept fund in 2015.
Friederike: Fantastic. So you worked at Triton Research prior to that right?
Jake: Yeah, so my career is basically you know kind of in Hedgefund world, for about five years and there was a technical product manager and engineer at Amazon.com working on ad tech products for about two years and then as a CTO of TRIAD Research which is a company that did a lot of interesting financial modeling in private technology companies.
Friederike: Good. Fantastic. Aleks, so how did you get into the blockchain ecosystem and what made you want to start a fund?
Aleksandr: Yes, so I believe in Jake who you know started telling me about crypto in 2015 and then eventually showed me a couple of, at that time they were extremely new and advanced ideas such as Ethereum and Augur, and when I combined together Ethereum and Augur in my mind, I kind of saw just how powerful this technology was to be able to enhance real life and their options between people and provide alternative finance technology and so I got completely bought in. That came on the heels of me also studying some amount of organization and social psychology, which made it extremely interesting for me to think about consensus in both technology space and the human space and how they kind of flow into each other, so I joined Jake in early 2016 and eventually we have got into CoinFund full time.
Meher: Cool, so I understand it like Jake and Aleksandr, you founded CoinFund with another co-founder and then other people joined on board from the community CoinFund created, right?
Aleksandr: Well, we started there was us and a couple of more people who were and still are slightly passive, they were just curious, but then Jake was the first to start and launch the CoinFund Slack, which I think by the way you know Jake is the best community manager I’ve ever known and is extremely insightful in creating sort of community spaces, which I think is still our strongest point of coin funds, but CoinFund Slack very quickly became a source of information, ideas, people, you know allies, arguments, and so on and so forth, so our current partner Aleks Felix came as a Slack member and many other people we worked with down the road.
Meher: I mean, I personally appreciate Jake’s community management, in fact I have to admit that, while preparing for many of the episodes at Epicenter I have listened to Jake’s Podcast because he had the guest first. So I listened to that Podcast and then my Epicenter questions were like, you know variations on Jake’s questions or questions that I think Jake did not cover, so you know I actually thank you Jake for your almost invisible contributions to Epicenter.
Jake: Thank you very much, it’s very nice of you Meher, but you know I think overall, if you kind of go on the CoinFund twitter account it says you know the best way to invest in a community is to be a part of it and that was a sort of a little tag line that we had from very early in the game where we really figured that in this very early space where it was really quite small still in 2016, especially in a space where you’re making investments and teams, one of the most important things is to engage the communities and from the decentralization standpoint, a lot of these projects are creating open decentralized networks and those networks depend on their constituents, their participants, their governance of those networks, I mean, so as an organization we’ve always like been very community oriented, we for example throw regular quarterly happy hours here in New York, kind of get some face time with our community, we try to participate in our community events and also as investors we try to participate in the networks of the projects that we hold in our portfolio.
Aleksandr: And that was by the way one of the things that drew me to CoinFund in the beginning is how easy it was to reach to all these teams doing these projects and ask them questions and pick their brains, which is something we still do and with not a small amount of pleasure.
Friederike: Good, I think we’ve segued into the CoinFund discussion, so what would you say makes a crypto fund a crypto fund and would you mind disclosing what assets and what amount you have under management currently?
Jake: We would mind disclosing that, but I am happy to talk about crypto funds, and you know ultimately from the very beginning I thought that a crypto fund is kind of a slightly different structure, so I think what we have seen happened over the last couple of years in watching and I think most people in Wall Street would agree, is that, there is a new asset class here. This is the digital asset class and whatever you have in a new asset class presumably has different properties and when it has different properties presumably the structure of an investor or of a fund that is participating in that asset class should be kind of commensurate with the asset class. So there’s a few things that set it apart right, for example the liquidity, like you could invest in early stage projects as we’ve seen many people do in 2017 and you get liquidity much faster using blockchain technology than you would in private equity, you know what other forms of investments and you know a lot of people when they kind of come into the crypto funds space, they try to put a box around crypto. They say we got to do a VC fund in crypto or we are going to do a hedge fund in crypto, or we are going to do a fund-to-funds in crypto, but ultimately we have always thought that you have to create a structure that conforms to the asset class and takes advantage of the features of the asset class that you are investing in, so we’ve spent a lot of time thinking about what that structure is and in short a crypto fund I think is, it’s a little bit of a highbred right between a VC fund, in a sense that you are always at this stage of the game, you are working with early stage teams, you are really trying to push those projects forward, get them to production, and it is very much a venture capital location, but at the same time you are then dealing with assets that are liquid to trade on 24-hour markets and that’s very much a hedge fund location, so at a high level I think a crypto fund is sort of a hybrid structure between those two and it keeps developing, so now as we talk about kind of staking invalidation, I think the role of crypto funds begins to change even more.
Aleksandr: And just to add to that I think if you want to do a good job in crypto investing, it requires you to be a professional synthesizer in a space where past experience doesn’t necessarily inform future experience, because this space gets re-invented every six months or so and so you know you asked me initially why do I call myself Chief Alchemist, that’s part of it, it’s like how do you look back and then project a picture into the future, which is a good picture, which is not going to be outdated by the new ideas, which does not prevent you from taking new directions. As a crypto fund you always have to be ready to go into a new direction such as for example, we are going to talk about generalized mining that was a direction that became obvious to us early, but not last year, in this year, and so you have to adjust and you have to be structurally fluid and you have to be fluid in your thinking, you can’t just decide oh, it’s all protocols, because that actually changes.
Friederike: You’re very flexible as to what to invest into, so how do you pick projects to invest into?
Aleksandr: Yes, and so on that point, just like the core principle of CoinFund is that we are looking at a technology space that’s first of all very nascent and constantly changing right, so you know to paint a quick story, in 2015, you know we see-saw teams building the centralized networks and they said we want to invest in that, so they invested in the equity of the teams, but then the value accrues the tokens and that’s kind of the high level like the fat protocol thesis, but then the technology shifted again and now we have all these layer two projects. The layer two projects only a tiny minority of them actually have token models today and very few of them have equity that gives you exposure directly to those projects, in fact about half of them are open source projects in the way that founders tend to monetize that is by building apps on top, so if you’re VC, suddenly you’re in a position where like oh, I thought I was investing in tokens, but now the landscape has changed again and how do I find exposure to layer two.
So what are the foundational principles of how we think about structuring exposure to projects is we say we want to be maximally flexible in how we can structure exposure, so that means if the appropriate mechanism is very traditional private equity, you know we’ll do that and an example of investment that we’ve made out of CoinFund is for example Coin List, a very traditional equity investment, but all the way on the other end of the spectrum you know we might be investing in a decentralized network directly and that might mean buying tokens on liquid markets, and then somewhere in between it might mean buying a SAFT or buying a convertible note, or buying equity that may convert to tokens in the future. So that’s the first thing to say about how we structure investments. What we invest into generally falls into two categories, we kind of consider ourselves first of all full stack investors, means we don’t just invest in protocol, we don’t just invest in middle wear, we don’t just invest in apps. We try to find the best opportunities in the iterating cycles of how these technologies develop, and we think that, you know, for example applications that have hundreds of millions of users that can put blockchain in front of their user bases, they’re still very, very compelling in the same way that protocols are compelling in their kind of adoption case, but we are also kind of practical, right. We say like in order for the Web3 or the decentralized world vision to take place, money has to move from traditional markets, founders have to have financial services, there needs to be compliance clarity. Everyone needs to be able to operate and experiment with these different projects and for that reason we will invest in what we call key financial infrastructures. These are things like issuance platforms, these might be things like secondary trading exchanges, and might be banking and financial services for startups, and in that way we are a very general investor. Our team is general and multidisciplinary, and we can apply kind of our knowledge and expertise in blockchain to select what we think are a compelling projects in both of those areas.
Friederike: Good, that is super interesting. So once you have selected a project, what kind of support do you offer them and how actively are you actually involved in governance and staking, for instance?
Jake: So, Aleks, do you want to take this, because Aleks has worked very closely with many different projects. Also, he is for example the author of the stablecoin model at Sweetbridge. I am the author of credit risk model for Etherisc, and so these are examples of how we deeply work with projects.
Aleksandr: Yeah so, we really try to be as useful as we can, and I think over time we’ve helped projects on many levels, you know, understanding the economics and kind of product market fit in crypto was one of the big value add we provided to non- cryptonator companies that were coming into crypto and trying to make use of these innovative models to operate. It starts from the very first conversation we have with the team even before we invest, you know, all the questions we ask the team, are the questions that everybody should be asking their team. So, one of the biggest question we always ask in projects is, why blockchain? Is that just the publicity stunt or does it actually make sense? And on that trajectory, we always inevitably come to a place where we look for the right integration and balance between centralization-decentralization, user experience, and crypto economics and so on. You know, we’ve helped projects build out their kind of investment structures, first and foremost for our own benefit, but then of course, that then later helps them go to other investors. We have advised projects what sort of help they may need, who they should be working with in this space, and who they should be hiring with interviewed people on their behalf. So it’s a very broad space of value add that we provide.
Jake: And I think now that’s shifting and continuing to progress, and this kind of ties into Friederike your comment about staking. So as the market of decentralized networks develops what we are starting to see is more and more diverse and domain specific network. So in the beginning, you had the cryptocurrencies, you may be had some asset issuance platforms like Counterparty or BitShares, then Ethereum comes along and they are like, no this is, here you go, this is a general smart contract platform and those very low level technologies that’s what we kind of thought of as decentralized networks, but we live a little bit more and the technology continues to develop, and now we see resource networks. These are networks like FileCoin which give you decentralized storage, these are computational networks. Like Go Loan for example where you can rent your video. These are networks like Livecare where you can transcode and stream live video. You see social media kind of content networks and decentralized Twitters and decentralized Wikipedias, and what you start to anticipate is a world where these networks are ubiquitous, they are globally accessible and in some cases they might even grow, potentially might grow larger than their kind of centralized counterparts, especially if you think about like decentralized storage space right. Like ultimately there is more storage in the world across people’s devices than probably Google and Amazon put together and so in that world of decentralized networks, every different kind of network has to engage third parties to perform services for that network. In the beginning it was minors who were processing transactions on Bitcoin, then we went kind of the proof of stake systems, and then we had validators and delegators, and now if we go kind of case-by-case, what we realize is the number of participants in these networks is growing. I was just at the Polkadot meetup last night here in New York, and in the Polkadot system they have a participant type of network called Fishermen, and what Fishermen do is they basically submit proofs of wrongdoing about what people are doing in the network to ensure kind of the security of that network. And so in a world where all these different kinds of participants are providing services, it’s sort of natural for a fund to say, “Hey, I might have access to this network because I am an early investor here” or “This is a project that is in my portfolio and by engaging their network directly I can add very measurable, very obvious benefits in network,” like think about if you are in layer two, a lot of success of kind of state channel payment networks depends on well how much liquidity can I have in the hubs of the payment network. And if you are a fund and you can kind of make an investment where you are providing liquidity, then that’s great for your network, right?, because then you increase the network’s input. There are some examples of how funds can participate.
Meher: This almost has to touch into this topic that you call generalized mining, Jake, I think we will get to that theme later on, and like your thoughts on generalized mining. It’s super interesting that with the times like CoinFund as a fund has kept changing, right? Like it started off doing liquid investments, then it went into stuff like illiquid things, and now it’s also like building validators and some of the service providers for these networks. Another key idea behind CoinFund was expressed in one of your blogs, which was titled “Fat Protocols Are Not An Investment Thesis”, in which you talk about CoinFund’s approach to investing, but you also like sort of not admonish, but talk about some of the downsides of adopting this fat protocol investing approach in the blockchain space. So, I am actually curious about your thoughts there, so could you tell us what this fat protocol investing approach is first?
Jake: Absolutely. So fat protocol is a famous post created by Joel Monegro who is now partner at Placeholder VC, and it’s sort of describing the view cultivated at Union Square Ventures, and it’s actually a really important piece in the sense that it tells the reality of decentralized networks. That reality is overall at a very high level, very generally, value of decentralized networks will tend to pool in their digital assets rather than in the equity of the team that created that. And this is a key early insight into how an investor should be structuring exposure to blockchain projects, and so in that sense it is very important, but the other thing about the fat protocols post is that it’s one of the earliest attempts to kind of make sense of these things and it’s also not very precise, and I think a lot of people who were coming into the blockchain space as investors, they needed to anchor themselves in the space somehow and fat protocol seemed like a really kind of very compelling narrative just to say, “Hey look, just like another technology investing in a new space, I am going to go and basically like invest in the infrastructure of that space rather than the apps” and in the case of blockchain, those are called protocols. And here’s why. It’s because the value will accrue there.
Now what happens when you start actually spending time with protocols and applications, and basically the decentralization stack is that you realize like there is no really clear distinction between like, what is a protocol and what is an app. It’s sort of like, is Twitter an app or is it a protocol? Well in fact, kind of the Twitter front end is an app on an API that exists, and that API is kind of like the Twitter protocol, right? And then people build other apps other than Twitter on top of it. Or if you look at Augur; well Augur is an app, right? It’s the prediction market platform, but what it really is, is a protocol because someone could come in and they can build another front end on Augur. They can build some kind of other protocol or application on top and so you begin to like realize that the distinction between app and protocol is not quite clear-cut, I mean, there are certainly things that are like lower level in the stack and things that are higher level in the stack, and things that are more user facing and things that are more developer facing, but ultimately in order to understand where value accrues, you kind of have to understand how that individual protocol works and then my post, “Fat Protocols Are Not An Investment Thesis” is just trying to point out very simply that, for example, if you are investing in Ether because you think some applications are going to take off on Ethereum and thus appreciate Ether Price, you might not actually be making the correct investment, because in order to determine whether the value of application usage flows to Ether Price, you really just have to examine that particular application or that particular protocol like in its own right. And all that post is saying is that you need facts and circumstances and sometimes it works and sometimes it doesn’t, but if you don’t understand the cryptoeconomics, you can’t really like make the decision of whether this is the appropriate value accrual that you are investing in.
Meher: Super interesting. So the way I have kind of thought about, like when I read your blog the imagination I got was, if you look at like the internet, you have all of these protocols – TCP/IP, TCP, HTTP, and you have like TLS, and let’s say like you were a VC, like let’s say you, I don’t know, you are Adam Draper’s grandfather or somebody like that, and you had a chance you were a VC through the 80s and the 90s, maybe you are like Tim Draper or Tim Draper’s father, and if you think of a career like that, that career is pretty interesting, because I am sure they have had the opportunity to invest in some team that was, I don’t know, maybe Tim Berners-Lee’s team building HTTP and they did have the chance to also invest in a team like Google that was using HTTP to build a search application. Difference is Tim Berners-Lee probably made like 20 or 50 million whereas Larry Page made 50 Billion, right? So that’s almost probably a thousand fold difference in the outcomes. Whereas, it would be hard to argue that Tim Berners-Lees’ contribution was smaller than Larry Page’s, I would say technologically both are equally valuable, but the financial outcomes for both of these things are so radically different, right? So when you see that kind of VC journey, you start to empathize with the VC class in always thinking about where value accrues on this stack, right? So I think, if somebody is a VC, their fundamental question is – “Okay, cryptocurrency is coming, blockchains are coming, and they’ll power all of these applications, and there will be many good technologies invented, but it’s not necessary that every good technology will make people rich. So they want to invest specifically in those things that are going to make people rich because investing in those things is what is going to make them rich, and so this fat protocol thesis argues that it’s the base layer protocols that are like Bitcoin and Ethereum that are going to make you rich, invest there, whereas like your approach is, don’t be so hasty about that judgement, it sort of depends on the particular application we are talking about. Would it be correct to say that, in your idea, there might be cases in which a wallet makes most of the money or like some UI facing element makes all of the money, and the protocol layer and even like the application logic doesn’t make any. Do you think a scenario like that could happen?
Jake: Absolutely, I do think. Let me give you two quick examples to maybe illustrate right. So let’s say you want to capture the value of an app on Ethereum like Augur, you have two choices, you can go into Augur REP tokens, if you feel that REP will reflect the value, you can also try to go into Ether because you are saying Ether as a base layer technology will accrue the value of Augur. What is the value of Augur? Well a lot of traders, they make trades on the platform on all these prediction markets and then the platform and the market basis kind of takes fees and then they pay some of those fees to the resolvers of markets, which are REP holders, and so you have this concept, like turnover, so now it’s fairly easy to see that, you know, if you are an Ether then what you are doing is, you are capturing kind of the utilization of the Augur protocol and the Augur app, you are not so much capturing the turnover. In order to capture the turnover, you actually have to hold the REP. And this is because, when Augur transactions go on chain, there is sort of the theory would be, “Okay, a lot of people want to use Augur”, people are then buying Ether in order to utilize the Augur protocol. But then, what you are capturing is the value of utilization of Augur, but not the transaction of Augur. For that you probably want to hold REP in that case. I will give you an example from the totally opposite end of the spectrum. Let’s say that you have a centralized company and you have a user base of a 100 million users, what you are taking on is a blockchain strategy where you can actually put some kind of crypto or blockchain features in front of this existing audience. Well, who is more likely to convert tens or hundreds of thousands of users into crypto? A protocol that has been built with no obvious utilized applications or an existing company with a track record of converting users and that already has a measure of user trust and market fit. Why I’ll argue? Maybe they both have a shot, but the application phase is still pretty compelling. And so that’s why I think, as an investor, and especially now that we are couple of years into smart contract platforms and dapps, they’re definitely compelling opportunities where some companies will be able to put applications in front of mainstream users and capture value that way and there are compelling opportunities where someone might create a protocol that everybody adopts and captures about.
Meher: I also personally feel the question of value capture can be very complex, right, like say you mentioned like the example of Augur, now if I look at something like decentralized exchanges, that’s an even more complicated question, because it’s like when you see like a protocol like 0x. So you have Ethereum, which is the base protocol, then you have 0x, which is like decentralized exchange protocol, then you have these order books, so centralized companies can build order books that run on 0x and then you have this UI or wallet layer where ultimately the user, the one they want to trade Ethereum from maker their inner wallet and they click a button, and then that wallet is going to decide whether to use which order book, which protocol 0x, or a competitor, on which platform, so you have these 4 different parts of the stack, and it’s entirely unclear which one is the one that’s going to make the most money, right? So of course, all transactions will go through Ethereum, but Ethereum just gets some transaction fee per order, is that going to be the biggest amount or 0x which is like a governance token? Is governance going to get the most of the value capture? Or does the order book get a lot of the money, because it traps liquidity? Or whether in the end it’s the wallet that makes the money because if you see the best hardware wallet is Ledger, right? So if there is lots of decentralized exchanges happening, the way Ledger the hardware wallet maker structures their integrations with decentralized exchanges are going to determine what players, when at the order book and exchange level, and therefore they will extract their own commission to root users to these order books. So a Ledger can also make money on the stack. And so it’s a highly complex question on like which part accrues value, isn’t it?
Jake: Absolutely, it is again in that example, right, let’s say if you have transacted at a 100,000 orders on 0x protocol right, you could have transacted a million dollars or you could have transacted a 100 million dollars. And it’s sort of like the actual turnover or the magnitude of the throughput of that system is not necessarily reflected on chain, it’s more like a projection of the utilization of that protocol on the chain and so in that case, should you be holding Ether, should you be holding the 0x token. Well, 0x token is interesting in its own right, right because you mentioned that it’s a governance token. I think it’s still, we’re still learning about governance tokens and exactly how do they value the ability to upgrade protocols and things like that but it’s certainly kind of a compelling case to say you know for a lot of people that protocol and they kind of rely on that protocol for their business or whatever activities they are engaged in, then they would want the stake in the governance. They wouldn’t want that protocol to get away from them in some way right, and present risk to their businesses. So, maybe that is a compelling way to value tokens that way.
Friederike: All right, so I think we are already moving in that direction. In the past, recently you two Aleks and Jake have started speaking about generalized mining. Would you be able to explain to us what that is?
Jake: Yes, so as we sort of mentioned before, you know I regard the generalized mining space as kind of the space of opportunities to provide what Chris Bernenski calls “self supply side services” to the centralized networks. So, these networks, they are we you kind of fast-forward into the future. If we accept the idea that there’s going to be a lot of these networks. They are going to be very different, they are going to be serving different domains. Some networks are going to be about social media. Other networks are going to be about the centralized storage, other networks are going to be about registering DNS, domain names and so on and so forth. If we live in that world, then all of those things – all those networks, they are not companies. They can’t hire you know hardware, IT departments. They can’t hire a compliance department. They have to rely on all these third parties to provide the services that make them function and many times those services are highly specific to exactly what those networks need to do.
If you are at the centralized storage network then you’re looking for storage. If you are LivePeer then you’re looking for people to build GPU center – you know GPU and data centers to transfer video. If you’re a social network, you are looking for a highly competent curators of content and maybe those are humans, maybe those are bots. But the point is that there’s all these third parties that have been cloud that can provide these kinds of services to the centralized networks. The question for us is how does a crypto fund fit into that scheme and the way that we kind of think about is, there’s a spectrum of opportunities and some of those opportunities are very hardware intensive and require out of the goal proprietary software? These are things like proof of work mining, right?
If you are building $100 billion data center and you’re competing on your ability to maximize your hash rate. On the other side of the spectrum you have opportunities that are very software intensive but these software, proprietary strategies you can kind of run on you know very basic and inexpensive hardware, maybe on AWS and so an example of something like that might be you know market making. You have a proprietary algorithm that market makes a certain market in blockchain. It might be what just happened at Livepeer network, which is the MerkleMine. The MerkleMine saw a bunch of third party miners. This basically distributes 63% of the Livepeer token supply to eligible, Ethereum wallets and those miners are doing that because they’re getting a little bit of a reward in the tokens – it’s interesting, crypto account system.
And if you go into the middle of the opportunities space and Meher you will know a lot about this, this might be something like Cosmos. Cosmos is not something that you necessarily have to write a lot of proprietary software for because the project is kind of like giving you the node software but as you well know you might want to write some. You might want to have kind of increased security in your data center. You might want to have redundancy. You have to worry about maintaining sentry nodes. You have to worry about physical hardware as well in some sense because you want to increase the physical security. And so this represents kind of a you know a staking opportunity but its somewhere like in the middle of the spectrum where it’s a little bit of proprietary software, little bit of hardware. They are certainly more expensive than what you get like a totally modified hardware side.
And where I think crypto funds fit in, they have a natural space in this spectrum right. If you are looking at the hardware end of the spectrum, well this is the natural place where proof of work miners are coming in and maybe they are repurposing some of their hardware for these different kinds of networks that are coming in. If you are on the software side of the spectrum, this is the realm of quants, this is the realm of you know smart people who are kind of building models and like trying to you know find the best algorithm that you know gets the best content and steem it up for the top. So, I think that’s a side of crypto funds and quants. And then kind of like in the middle, you have this staking sort of area of the spectrum where here you are competing on how efficient am I. How secure am I.
We’ve talked to a number of projects that have quantified at this point whose product is you know one click deployment of staking nodes on various networks. So, this is all about like how fast can you get into this network. How fast can you build stake. How secure are you and maybe even how many networks can you be on at the same time. So I think crypto funds are going to go to the more proprietary software side. It’s sort of a natural, funds don’t want to run hardware. Data centers, if they don’t have to, they’re putting their – probably rather invest in teams that are doing that and this becomes an incredible value add for fund I think and I think we’ll see this in the future. One reason is that you can add tangible, measurable value to portfolio companies, as we sort of touched upon before right. Another reason is that you might be – this might be a competitive advantage. If you’re doing this and other funds are not doing it, then maybe you’re generating returns that are not correlated to what most people will do which is long-term, long only investments in these like Saff nodes.
And in its flatter down market that might be an interesting source of returns for funds. And finally I think you have access, you have better access as an investor and so, you know for us Lifepeer is a great case study for that, in the sense that we were never early investors in the node. However, when the network came up about a year later, we were able to use their mechanisms of transcoding and mining to build up a stake in that network, kind of in a single digit percentage size wise. And it’s comparable to stakes of early investors and we think that’s really interesting because it democratizes the access to ownership of these networks. Now you just don’t have to be a VC to get the kind of ownership. You can be a technologically savvy participant and – well finally in summary what that says to me is that I think crypto funds over time are going to have to get a lot more technological and then VCs, who are sort of pure VCS and pure equity investors are going to get a lot more constraint because they can only kind of invest in this one stage of the blockchain company lifecycle which is the stage where you are funding the early equity but other folks can choose their stage at which they invest. They can invest in that equity but then it also buy the token on the market, and finally they can build stakes, like direct participation in markets. So, we think of this is like a set of features, like an amazing opportunity for funds.
Friedericke: Great! Thank you, that was super interesting. And, so basically when you speak about generalized mining, you mean that more people actually actively contribute to the upkeep of the network? So, how far do you see that going? Would you expect complete lay people to actually contribute or is a select circle of people contributing? How do you – do you have to have some sort of gate-keeping and how do you make sure the people who want to contribute are actually capable of contributing? How do you know their hardware, their software is good enough?
Jake: I think it will vary and it will – you know again it will run from highly experienced, very technological companies, like for example the company that Meher is working on, which is providing like very advanced technological services, you know to just sitting on Steemit as a human being and kind of like using your mouse to curate articles but the point is you know even if you’re human, like here is an opportunity for you to earn value from the network by contributing a valuable service and not all the services situation. So, I think there will be all kinds of players in that market and they’ll all be doing like pretty different things. Some people will be casual, some organizations will be very serious about a very – like a narrow area and then other people might be just concerned with kind of like scaling data centers. So, the centralized storage networks might flow and work.
Meher: This brings a different dimension to a crypto fund. So in the beginning it’s like crypto funds are like, “Oh I want to invest in the liquid assets that can become the currencies or stores of value in the future.” Then the second generation is, “Oh, there’s going to be these application stacks and I want to invest in that part of the application stack that’s going to capture value.” And now this almost feels like now CoinFund is going to put a new kind of hat in, which is, “Okay there’s this new protocol coming and it’s going to need all of these off-chain workers. Can CoinFund build one of those off-chain workers, and if CoinFund does then is that workers – is that a competitive advantage to building that worker that lasts over the long run?”
And if you do find that there is a competitive advantage to building a worker, then CoinFund would build a sort of a worker node or a generalized mining node.
Do you actually think there can be defensible competitive advantages to generalized mining nodes or worker nodes? Like our experience with protocols like Livepeer, Tezos and even like Bitcoin is that with all of these mining and off-chain services, the market ends up highly competitive because there is free entry, anyone can enter and build one of these workers. So, there is going to be – there is probably like thousands of miners in Bitcoin. There is already a hundred bakers in Tezos. There is probably going to be hundreds of transcoders on Livepeer. Do you think they will ever be a protocol in which building a worker makes sense for a crypto fund, from a competitiveness and long-term perspective?
Aleksandr: Yeah, actually – so I have a whole range of thoughts about that but most importantly is the fact that as the number of networks that need this type of service increases, the networks will always compete for attention from you know different staking companies. So, a staking company has a limited amount of resources you know both engineering and hardware and cognitive cycles to contribute to every specific project. So, all of the staking companies are going to always choose the best ones and so networks that need this type of service they will always try to create the best incentives. This is a little bit unlike the mining, the hardware mining stays where the costs – the profits, the miner profits will always gravitate towards zero because in the staking in generalized mining space there is a upward push on the mining rewards from the networks that compete with each other. So, you know you as a company can stake Tezos or Cosmos or Livepeer right, which one are you going to choose? And so there is the natural two-sided competition, one between the staking companies that want to provide the best service and one between the networks that want to get the best staking companies.
And so, if you are a company, as a crypto fund, you get a very natural early entry point to all of these opportunities and so the competitive advantage for a crypto fund doing this is that the crypto fund is usually aligned with the network from the very beginning and has these dialogs early, and so for us to set up a generalized mining operation for a specific network is very natural, and I think long-term defensibility is a great question but we don’t mind doing generalized mining in the short term. It’s always the same discussion as which network do we invest in? Which network do we spend time on and so on.
Meher: So the trade-off space for a fund like CoinFund almost appears to be – okay there is this – there are these new networks that are coming up. Either we can build these service nodes or generalized mining nodes ourselves or we could delegate to other people that are building those nodes. And so, imagine like there is two parallel universes. One in which CoinFund adopts the approach – hey we will build generalized mining nodes and protocols we love, like Livepeer and there is another parallel universe in which like CoinFund says – hey we’re not going to bother like building these generalized mining nodes. Let’s just delegate, right. Let’s just delegate to the people that are. So in these two universes, like CoinFund’s performance over like 5 or 10 years is going to be different and CoinFund is building generalized mining nodes so you have chosen this universe where you do build. What do you think is the long-term advantage to CoinFund, given that generalized mining is so competitive?
Jake: So, yeah, yeah, yeah. So, – again this goes back to kind of my assertion that the space of generalized mining opportunities is actually much more vast than the space of staking opportunities. So, like if you look at most companies in the staking space like Alex was kind of getting at, are not funds and so then their business model kind of depends on their ability to leverage third party delegation and kind of take a commission on that, on that delegation. But of course that’s a limiting factor because not all networks will have opportunities that are staking opportunities and not all staking opportunities will have delegation, so at least on protocol delegation. So if you take first example that comes to mind is new cipher network. New cipher doesn’t – you can actually delegate stake and in new cypher you have to run a node. And so the competitive advantage for a fund I think is not so much in competing with the highly commodatized market of staking companies and people who are competing on efficiency. It’s more like competing in the kind of hard to replicate market of proprietary software and here you are kind of competing on how smart you are and how good of an algorithm you might build.
So, let’s take Steemit as my – it’s one of my favorite examples here right. Today there are a bunch of bots on the Steemit network where users basically send them a little bit of steem and those bots up vote their article and their article goes up higher on the Steem website and so they earn an ROI in Steem because Steem articles are compensated in Steem by the system. And they – you know may raise exactly the advertiser models. Like I go out to Facebook I pay Facebook some money, Facebook distributes my ads to men over 35 who have recently been divorced etcetera and then they buy my product and I get a ROI. In the same way I compare Steem bot and get an ROI.
Now which bot is going to give you the best ROI or which bot is going to earn sort of like the best reward? Well, it’s the bot that can most effectively identify in deep what is the correct content and say most bots and Steem are really, really stupid. They are like whoever pays them they will just get up voted but then eventually someone is going to come along and say I’m going to use kind of a database and the machine learning algorithm and some other technologies to actually identify and try to predict which Steem post are going to go viral in the future. And if you can do that on Steem, then you’re going to earn a much better return and because that’s a proprietary algorithm nobody knows how it works because you made it up.
And so you’re going to earn a better return than others and that’s I think really the competitive advantage of funds versus kind of going to these more commoditized networks or networks in the future might be more commoditized like Cosmos and Tezos and really like anything that has staking. For those kinds of networks I think in the long term funds might be happy to delegate to other parties who spend a lot more time on security and efficiency and things like that. But in the proprietary software space I think this is where the really interesting stuff, there is certainly going to get competitive but not commoditized is going to happen.
Friederike: Just a follow up on that, a lot of staking tokens today in effect when you stake them, it’s a lot like getting hidden dividend right. Do you think general mining is going to move into that space and take a big cut of that or do you think these staking models where in effect you get a hidden dividend a thing that will exist in the future?
Jake: Well, I think generalized mining opportunities like from staking to proprietary algorithms to kind of like hardware mining they’re all different forms of getting dividends. I think as Alex touched upon you know I think in today’s world whenever you choose to engage in this activity in a particular network that it still an investment decision, right. So like if you look at you know staking opportunities like, Tezos, Cosmos you know you can try to calculate and estimate what is the token denominated return that I might get if I stake a certain amount of tokens. And then you might get a number like I don’t know exactly but maybe like a number like 10% or like 20% but I think a lot of the people that are engaging in this activity is not because they want to make 20% return on tokens. It’s because they believe, they’re actually speculating on the success of the network or these particular networks in becoming like large multi-billion networks.
What they are really doing is that they are speculating on the fiat denominated return but then they are trying to maximize it by also getting that 10% of tokens on top. So, it’s kind of funny, like there is a number of for example lending platforms in crypto right now where you can put your Ether and then you can earn a little bit of Ether kind of like lending it out right. But the amount of money that you earn in Ether is like 70 cents but the volatility of Ether is like thousands of dollars, like a new position right. And so, a lot of times in today’s market, it makes sense to still sort of speculate on the future success of these platforms rather than so much worry about what is the dividend? How much percentage of the dividend am I getting? I think that’s going to get like to be a lot more important you know down the road when these markets become a little bit more efficient.
Do I think that who takes a cut of that – so I think like in the commoditized area like when you are talking about specifically staking networks then as Meher was saying like, yeah this gets like really, really commoditized and so the companies that are providing these services are going to be competing on fees and it’s always like kind of a race to the bottom. I don’t know it ever gets to zero because these companies have to operate but again whoever can operate the most efficiently will handle all this sort of fee and then aggregate delegations stay there. You have any thoughts on that Alex?
Aleksandr: Yeah, I think it’s hard to distinguish and classify this space today. I don’t think of it as necessarily easily separable into you know staking delegation, generalized mining, you know it’s – it’s very easy to fall into the trap of imagining that you know once you’ve seen one network you’ve seen them all. That’s completely untrue. The opportunity to make returns on contribution to various networks is very broad and so you know you can say, well the funds can just delegate to other stakers and very quickly find out that that’s not true. You know you can say that staking is a way to distribute rewards and then very quickly you find out that that actually requires you to pay a tremendous amount of attention and provide a good service. And so it’s not just rewards but its rewards for something else.
So, I would be very careful drawing kind of black and white lines between these areas. I think of that as providing services to networks according to the rules that they set up. Networks are very immature and the way that these rules get created is often you know very quickly becomes clear to be not ideal and mistakes are not very quickly become obvious and new networks try to fix them and so this is very fluid – and like I said as a fund we are trying to position ourselves where we’re at agile and able to start in this development very quickly.
Friederike: Thank you that was super interesting! So, you said that the networks that are kind of immature, the way that you two are trying to mature them is by building or by helping them by supplying them with a trinket that you have started building named ADAPT. Can you talk about that a little bit?
Aleksandr: Yeah, so I started in blockchain working with projects very early and I very quickly realized a number of limitations that projects face when using existing technology. So, let me give you an example. So, one project I worked with before was trying to create a very advanced financial system and they were trying to use Ethereum for that and they very quickly realized that it’s very hard for them to create the user experience that they want to create for their customers using Ethereum because Ethereum transactions require paying Ether and so all the users will you know who need to transact on their network are going to have to call their wallet, manage their keys and call Ether, and it became obvious to me that in blockchain, and this goes back to our discussion on fat protocols to which I have a slightly different answer right now. You know is fat protocols like Ethereum, EOS and such, they are in fact designed to capture value. Of course, that’s true you know the teams have to ensure that their currency has value and as that value increases and what that almost always means is that their networks will compromise on the possible user experience of people who are trying to build and use products on top of that. And so, I started to think of ADAPT as a kind of a generalization of blockchain on many levels. The main premise of that is that even the network on which application runs has to be customized to your application’s user base. Generalized networks will inevitably compromise user experience and the choices that these networks make will inevitably make things more complicated for at least some new spaces.
To give another example, you know CryptoKeys was built on Ethereum. Ethereum was a network that was built from — kind of came out of the bitcoin thinking of sovereignty of money and radical decentralization but CryptoKeys is again, they don’t need radical decentralization and the moment you realize that, you realize that Ethereum is the wrong platform. And so, what I’m trying to do, is I’m trying to enable people to experiment and create smaller, much more customized networks for their users. Does that make sense?
Friederike: Thank you that made perfect sense. So, what kind of tools do you actually provide the projects within this ADAPT framework?
Aleksandr: So, is platform technology but it’s not a platform network and that’s a distinction which would have been very easy to make in 2005, it’s very hard to make in 2018 because everybody thinks of networks as platforms. But if you go back to 2005 and look at weekly GCC limits and so on, you kind of quickly understand the platform is software and so abou this software, it’s a software to build and launch networks and its modularized software so it basically is structured as a programming language on top of a very general data model that allows you to develop nodes, build consensus algorithms and then also develop the SmartContract infrastructure on that network and the business logic.
And so as a toolkit it looks like a star where in the middle you have the compiler of the programming language, that above that it has modules in that programming language and below that it has the primitives and the data structures and the networking components that can be accessed from that language; and the language is basically the glue that makes the toolkit work. So, using the toolkit the idea is that basically you can first load and use you know all the necessary modules that somebody else has developed for things like wallets, tokens, governance, TCRs, you know whatever generest components that can be built in decentralized space can be developed and be used and then you would add some business logic for your specific project and then you would launch that as a decentralized network that is specific to your specific project.
Meher: Very interesting! So, to me, because I’m building a validator, a commodity validator for the Cosmos network, I tend to think of ADAT – so the vision behind Cosmos was like it allows entrepreneurs to build their own spoke chains, their own blockchains and have those blockchains handle their own application. And like that, blockchain is tailored for exactly that application. So, you know like this vision of application specific blockchains would have for example, Augur built its own blockchain network, Gnosis built their own blockchain network and so on. And so ADAPT to me is a very similar vision but with a different technology stack but the key difference is one of incentives right. So, in Cosmos they have this token, the Atom and they want to create a platform around the Atom.
So, when somebody else builds a blockchain application, some of the value of the blockchain application they want it to leak into the Atom token, right? And so like Cosmos, the project has like launched this Atom token and their incentives are geared towards leaking some of the value of these blockchain applications into the Atom token. What’s special about ADAPT from this perspective is you don’t want to create that central Atom token or the central Ethereum token to which the application value should leak into, right and you want to keep builds just like this open source stack that anyone can use to build their blockchain network.
Aleksandr: Yeah, so at Cosmos, to me is probably one of the best thinking in blockchain and it’s actually like you say is very similar. What Cosmos is doing will naturally happen in ADAPT because you do need a hub, you do need a trading center. You do need infrastructure services for the entire ecosystem of these small to medium size networks. The difference I would say is not in the value capture process but more in the amount of freedom and the amount of sort of high level tools that developers get when they want to create their own networks. So, for example, Cosmos is completely based upon the Tendermint consensus protocol which has you know fast finality and some really good properties. However, it’s not necessarily clear to me that that’s what you want for all loose cases. Cosmos uses fast finality because they want inter blockchain communication. They want to move tokens from one zone to another.
I don’t think that’s correct. I think that’s too limiting because in some zones you can implement tokens using different code bases that don’t even exist on that zone, they have another, right. So, it’s not clear to me that moving tokens from one network to another make sense. What’s clear to me nevertheless, is that you need for example a decentralized exchange to live kind of in the middle of the ecosystem and it’s likely that you know that’s where this will go but I want a lot more freedom in every zone. I want zones to be able to choose their consensus protocol and if they want probabilistic finality, they should have it, right. If they want their zones to be not interoperable with the rest of the ecosystem they should be able to do it.
ADAPT is much more high level than Golan, in terms of developing code because it hides away data modeling and it provides high level tools for organizing data and creating usable security primitives inside your database. So, I claim that it will be a lot quicker to develop an ADAPT network than a Cosmos zone if you need – if you want to be sufficiently different from the rest of it.
Friederike: Thank you. That was super interesting. So, does ADAPT have a business model?
Aleksandr: We are working on it, at the moment is fund raising on donations because I am really sick and tired of seeing projects raise billions of dollars on a white paper and I have a very principled objection to that because to me that creates really, really bad incentives that will basically destroy this space in the long term but we are in discussion internally about how we can structure this effort because to us it is painfully clear that the kind of subtle shift in perspective that I’m offering – that we are sort of trying to put together is really important for this space. We want to do this as a fund because we are also tired of projects coming to us and painting an unrealistic picture of how their adoption and how their baseline technology you know will work because we’ve seen so much in the space and we thought about it for so long, in terms of real risk cases and real world adoption that we can very quickly point out what will be a blocker for this project and that’s what we’re trying to solve.
We’re trying to create a mechanism for people to build the next generation of blockchain technology where each application can be customized and their network and their you can say protocol but it’s a – overuse the word “will be customized to the users” to an extent where it will stop being like toy and will become a real world tool. So, think about questions such as governance, recourse, validator incentives, transaction economics, can your social networking application built on Ethereum today have free transactions? Can your miners of your network also be your oracles? Also be your service providers on many different levels for your application. The answer is you know LivePeer is running on Ethereum. Ethereum has miners but LivePeer has to list a whole new group of people to help them with their MerkleMine Sharedrop. Ideally those would be the same people.
This feeds directly into our discussion on generalized mining because you know and this takes me to the topic of initial witness offerings where it’s very explicit that early supporters and investors in this project, in a project should also be service providers on the miners that are highly aligned with the mission of the project. We don’t have that today, we don’t have that level of alignment between early supporters and later service providers. And that’s what I’m trying to create because to me that’s more important for the long term of this ecosystem.
Meher: Cool. Alex, we really look forward to seeing the development of the project. Perhaps we should have an episode on the project when part of the protocol is released on the first application, blockchain networks start building on it.
So, that takes us to the end of this show. Aleksandr and Jake, thanks for joining us today.
Jake: Thank you Meher – so happy to be here.
Aleksandr: Thank you Meher. Thank you Friederike.
Meher: So, thank you for all that who listened to this episode. As you know we release episodes of Epicenter every Tuesday or Wednesday. So, subscribe to us if you like the content. You can also watch a video version of this show on YouTube.com/epicenterbitcoin and you can chat with all of the hosts and the community members at epicenter.tv/contact. If you like this recording or you didn’t like it, leave us an iTunes review as it helps people find this show and it make us happy and we look forward to being back next week. Thank you.