Jake Brukhman – Inside CoinFund’s $1B Strategy & The Future of Web3 – ep 27

The thing that people are probably most sleeping on in Web 3 is the fact that this is going to be the best Web 3 go to market of all time ever. This podcast is for informational and entertainment purposes only and is not investment, legal or financial advice. Opinions are those of the host and guests and do not reflect any affiliated entities. Investing involves risk and past performance is not indicative of future results. The host and guests may hold positions in discussed securities. Now please enjoy the show. Jake, thank you so much for joining me today. Super excited to chat with you. All right, Jake, you are the founder and CEO of Coin Fund to one of the world's first Dizlasa funds. Coin Fund invests in both liquid and venture strategies. You guys manage roughly a billion in AUM and you previously worked in quantitative development, Amazon and an alternative data company. Before diving into nitty-gritty, you tweeted recently, 2025 is the first year of blockchain adoption. That's a spicy tweet. You need to tell me how is that possible at 2025 is actually the first year of blockchain adoption. Amazing. Hi, Andrew. Thanks so much for having me. It's a pleasure to be here. It's also not the first time I'm on your podcast, so I'm excited to be back. I think 2025 is the first year of blockchain adoption because I think historically blockchain adoption was said to be a grassroots bottom-up kind of process. It said that if we had enough normal people, enough Ethereum wallets and Bitcoin based and buy enough Ether and load up our wallets and use enough DeFi applications, then we would achieve broad mainstream adoption of crypto. Now I want to say that we have not achieved zero. We have achieved a three plus trillion dollar asset class. We have achieved about 8% penetration of the world population. I have held crypto in some capacity, mostly, you know, that is probably Bitcoin. And we have achieved, you know, some technological progress and some developer progress. But Andrew, this is not mainstream adoption. It's not like your mom and your neighbor and, you know, your kids at this point of school are using DeFi. We have not actually converted a lot of infrastructure to blockchains out there in the world. Like most payments, they'll take place in banks, on Swift and so on. We have a developer base, which is about 0.07% of the world's developers. When you think about like the iOS app store and the Google Play App Store together, they put about 100,000 applications into the world a month. And a few of those applications turn into something profitable and good and growing and things that customers like. And we are nowhere near doing that in with decentralized applications. We're nowhere near putting 100,000 decentralized apps into the market, into the market of money. And so after 10 years of this, I would say that like the piece that's missing from this equation is the broader institutional and enterprise adoption of blockchains. And I also am keenly aware that for some people in our industry, you know, the idea of like giving blockchains to banks is a little bit countercultural. But I actually see it as a huge positive thing. I think that we can achieve a lot more saturation of blockchains in the world if banks and enterprises and Amazon and IBM, etc. So start to use blockchains technology. I think that will increase the developer base. I think that, you know, they will have an impact on regulation, which we're also seeing this year 2025 is the first year of blockchain adoption primarily because one of the biggest bottlenecks that the blockchain industry has had for the last decade, which is kind of a regulatory bottleneck has been lifted. And so we're in a world where, you know, we're possibly weeks or months away from legislation in the United States. We have had a regulatory complete 180 degree turnaround in terms of the attitude toward this industry. There's broader acceptance. There's broader technological acceptance. There's broader regulatory acceptance. And the 2025 is going to be the first year when the enterprise and institutional side will also start to be able to achieve adoption. And I think that we will achieve mainstream adoption when both of those sides can come together and interplay when, you know, when retail people will be using more of these products and when institutions will be integrating a lot more of this infrastructure because they now can. And to me, that is positive because the more that happens, the more we can actually try the centralization experiments that are more decentralized rather than less. So that's sort of my, the reasoning behind my tweet. And so you can see the adoption of blockchain and crypto broadly. It's going to happen through businesses that are utilizing technology to become more efficient and just become, you know, improved operations and whatnot. And through that process, it'll expose regular folks to products that use blockchain. And that's how it's kind of, you know, that's kind of, they can be the main pathway of adoption. Absolutely. There was a time in our industry when, you know, we were kind of looking at the market, like think like back to 2018 or so, you were around back then, Andrew. And people were saying, yeah, of course, banks will use blockchains, but they're going to be private blockchains. So they're going to be permission blockchains. And they will defeat the purpose of a blockchain. And like, why are we doing it? It's stupid to add a new kind of database to bank. I think the reality in 2025 is that a lot of banks think they're going to be in public blockchains. Right. And, you know, maybe they'll have permission blockchains to and their toolkit for doing certain things. But I think a lot of banks will be able to go on public blockchains and track with Bitcoin, interact with the Ethereum network, interact with an EVM based L2. And that is going to be insanely positive for, you know, the mainstream adoption of those types of networks. But don't you think building up that question? Don't you think that there has to be some sort of, I don't know, sub chain L2 something that where if these banks, let's say North Korea hacks, you know, a big wallet that JP Morgan owns, there has to be some sort of rollback mechanism for that because clients of JP JP Morgan don't want. It can't even fathom the idea of, oh, guys, sorry, we don't have your funds. Lazarus group just like hacks are wall with a billion and sorry. You know, so, so I agree with you. You know, I do think public blockchains are going to be the primary settlement layer. But when there have to be some sort of guard rows in there for like these in social players. I absolutely think you're right on that. I don't necessarily think those guard rails have to be built into the blockchain. Like, for example, you know, if you're dealing with a large corporation that is holding your private data and that private data becomes hack, they're exposed as, you know, in our day and age happens all the time. Those companies are liable to you for protecting your data. So I think the same thing would apply if JP Morgan or whoever offered you a product on the blockchain and then like something went wrong. It all kind of depends on, you know, what terms of service you're accepting, like with this, with this provider. And the other thing to say is I think there's also been a number of proposals and technologies and technological progress that I think makes those situations over time less and less likely. And one of those things is like just making blockchains more secure and smart contracts more secure through formal verification through better security practices through better coding tools through. I don't know AI generation and evaluation of code. There's been blockchains proposed that actually can do rollbacks or have safety mechanisms where a kind of recourse is possible smart contract schemes and do that right. So it all sort of, you know, all these things have been kind of around. But what hasn't been done is that they haven't been integrated into well understood products yet. And the reason for that is we haven't really had that much kind of customers for these things. As soon as the customers will flow, all these problems will become very prominent and the solutions will, I think, follow in quick succession to us understanding those problems. All right. I want to move the conversation to coin fund and coin fund, coin fund business lines. You guys have venture fund and within that you guys do seed, you guys do A and B rounds, kind of that, you know, many growth will call it. And then you guys do liquid opportunities. Can you describe kind of these different business lines and where you focus your time. Yeah. So we actually have a seed fund. That's where I spend most of my time and do most of my investing. And, you know, it's been publicly announced our last seed fund was $158 million. And then we also have a series A and B fund, which is about a $323 million fund or so. And that's led by David on our team, which focuses on slightly later stage, more traction, more product market fit, more revenue, et cetera. And then we have Seth Gensfield leads our, he's our head of liquid investments and leads our liquid strategy as well. I do a couple of investments in the, in the venture fund, a sponsor there, but most of my work I do in the seed stage. So are you guys doing anything, you know, across like yield or I don't know what we'll call like GP stakes and funds or secondary or any of these kind of more opportunistic. activities or kind of more vanilla? - I would say we are very kind of purest in terms of taking a venture approach and bringing it to crypto. So we don't have like a fund of funds allocation. We are open to yield. In fact, there's certain like when you're making investments in highly decentralized open networks, there are sometimes ways that we can benefit our portfolio companies by being active network participants. So like we're open to that, but we, you know, like generally like the vast majority of our returns are created by being early contrarian and right in the best companies with the best founders in this industry and keeping it sort of there. - So when you're looking at different opportunities in the seed landscape, how are you slicing dicing it? Are you looking at, okay, I'm really excited about, you know, decentralized AI, therefore I'm now going, so I have this thesis on decentralized AI. I'm not going to talk to all the founders there or some more so, hey, you know, we're sourcing all these great founders from our network and I'm kind of reacting to what I'm seeing. - Yeah. Well, we say that we're thesis and trend driven and what that generally means on the thesis side is that we try to be, again, early contrarian and right to areas. So the centralized AI is like a great example of that. I think the first investment we made in that category would be world at the end of 2020. And it wasn't like fully obvious yet that there was going to be this huge intersection of, you know, AI and Web 3 as it is today. But what, you know, the second data point in that, you know, kind of two points make a line curve was our investment in Jensen in March of '22. And here we met a company with folks who were like from the AI world who were seeing that using Web 3 primitives was going to create some like really, really interesting opportunities and products. And then we published our thesis on decentralized AI in September of '22. Now, I think at this point, there must have been like six companies in the world who were like looking at this intersection. We called three of the founders and the other three called us and they were like, oh, I can't believe like you're thinking about this. But, you know, the space started to develop from there. And I would say it became consensus in like the summer of 2024. I remember went to Super AI and Singapore. And that was the first conference where like, crypto AI people, and traditionally AI people were, you know, mingling together. And if you look at Cato statistics, I think the centralized AI as a topic represented 30 to 60% of crypto and the mind share for like most of 2024 and into 2025. So it was quite a confirmation that we were like pretty early to that space. And that's an example of us kind of hopping early on to a thesis. And then that thesis like turning into a trend. So when I think about investing, I'm like, well, what are the, what are the two or three like major areas right now? Well, I'm still doing a lot in decentralized AI. I think there's a lot to do in terms of data, you know, in terms of decentralized AI training, which has been a big pet thesis of mine. I'm also very, very excited about the intersection of DeFi and traditional finances. I was just saying, I think we're coming into this convergence time when, you know, the financial world, for example, will be starting to integrate these different like Web 3 primitives. I think stable coins are part of that. I think what Rob is doing at super state tokenizing equities on chain as part of that. I think new banking, we made an investment Dakota.com recently is part of that. And then, and then the other big area that I'm super excited about is like finally in 2025, people, projects like World are showing that we can actually do large scale mainstream consumer adoption of crypto. So World is one in our portfolio, 26 million registered users growing. Gunzilla is another example, 14 million registered users playing games on traditional platforms, but also using NFTs to tokenize in game assets. So those are the three areas, you know, the centralized AI, you know, tradify convergence with DeFi and consumer adoption that I'm broadly like very, very excited about. How are you formulating these these these? Are you talking to founders and, you know, you're just like blown away by the drive and intelligence of these founders and that's giving you the indication that, OK, you know, DeFi is tradify these founders in the sector are so amazing. There must be something big here. Or is it more so you guys are putting your head together as a coin fund team and being like, all right, what is, you know, what are we seeing? Have adoption. You know, what's going to, what are things going to move the needle? You know, like, how is that? How are you guys developing these these? I mean, I think coin fund folks do it in different ways. We tend to be like encouraging our team to be generalist, but have one or two areas of specialization. So for example, you know, Georgia and our team will look at a lot of like stablecoin payments projects and he's developed quite a, you know, thesis there and knowledge grace. Excuse me. I would say I'm like the go to person on a team for decentralized AI with a couple of my colleagues that help me do those deals. And you know, I think it's a little bit about projecting into the future. Like I like to think that, you know, coin fund and like what I bring to the table of coin fund is a technical understanding of a lot of these and structural and deep tech projects, which helps me kind of theorize about, you know, what might win or be more pressing in the future. And then it's also a combination of, you know, operationally being good. Like one of the goals in coin funds investment team is to cover the vast majority of the market. And I think we do, you know, we very deliberately have a process where we actually compute like our coverage rate, we see deals that we missed and like try to understand like why we missed them. And you're never going to get 100%. Obviously, because not all kind of opportunities come to market, especially in seed, but it's a discipline. It's a discipline of like very consciously wanting to see all the best stuff in crypto, all the most relevant stuff for us because we want to be in like the best companies. So having that like thesis building, you know, discipline and kind of this operational discipline of like seeing the market, I think is what helps to do that. And when you find a group of or a founder or founders building a company within an area that you're really excited about, it's called, you know, say decentralized AI, how are you waiting the different aspects of that startup? You're looking at, you know, is it just like, okay, they're on on thesis. So now it's all about the founders or no, it's like, you know, founders are we wait that 50% and then the product is another 50% and they're already in the market that we like or, I guess, where do you wait these different factors? What when you're looking to make a seed investment? - Well, I mean, the general wisdom in early stages, like team product market, like when I think about market, I think a lot of people who are in Web3AI already sort of think it's big, right, for various reasons. So we, you know, we're already kind of there. When I think about team, you know, the profile of what we've done in decentralized AI looks like traditional AI people who have a lot of credibility in the eye. This is because we typically invest in teams that are like doing like solving hard problems. Like pluralist is, you know, it's extending the state of the art of AI by doing model parallel training as Alex has talked about in various podcasts. And this is not something that, you know, the average person is particularly qualified to do, right? You kind of need an AI PhD. You know, this is not something that a normal person is typically like very qualified to do. It's something that you need an AI PhD to do. And we want those AI articulate founders to like look at Web3 and say, wow, that Web3 primitive, if we use that in our project, it's gonna make us like incredibly competitive, you know, in the broader world. So like one hypothesis we kind of have is, for example, that hey, if you do, you know, AI model inference on decentralized networks, hey, maybe that will be like significantly cheaper than people who do that over the data center. Now it turns out, Andrew, that thesis is like totally, totally wrong. And that's actually like not the case. Yeah, so I was, you know, so on the team, again, we're looking for AI experienced founders that are coming into Web3 and they're using Web3 primitives to be competitive. And what I was saying is, you know, we might make a hypothesis that, you know, if you're doing decentralized AI model inference on decentralized networks, hey, maybe that's cheaper than doing it kind of the way that, you know, big tech companies do it. And it turns out that that's totally not the case. that thesis would be invalidated, right? 'Cause it actually-- is more costly to maintain a decentralized network that can coherently and safely and correctly deliver those same level of results at the pay more when a decentralized network to do that. So that's kind of the profile we look for. And then in terms of products, the way I think about it is, there's a lot of noise and a lot of different ways that people look at AI. The things that I think that are gonna be the most impactful are, can I use Web3 primitives again to like actually advance the state of AI? It's not that I wanna give like, like there's nothing wrong with this. I, you know, if you wanna give USDC payments to like agents or something, I just don't think Web3 kind of quote unquote "anching agents" today are like very competitive with stuff that opening AI is putting out or Google is putting out these big companies who already have so much access to your proprietary like zero-party data. But what those companies are not thinking about is like how do I train a model into centralized network? That's a hard problem. That's a problem that requires, you know, like in pluralist six AI page days from Amazon to sit down and write a bunch of papers of how to do that or like prime intellect has shown that you could train a 32 billion parameter model, you know, on a decentralized network given their scheme or Jensen is showing that there's a swarm reinforcement learning that you can do in a decentralized fashion. So what we've done is we've gone from 18 months ago where people are saying this stuff is never gonna work. It's impossible, it's too slow. Why are you even like thinking about this to a world where like multiple companies have this in market and are making fundamental advancements in AI? That's the kind of investment that I think long term is gonna be really, really impactful. - When you're doing diligence on, actually let's kind of build your ideal start-up diligence process where you're like, okay, wow, this is 10 out of 10, these founders have, you know, it's called relevant experience. They're from open AI and they're building a decentralized AI thing. They have some early traction that shows their growth as week over week, absolutely stellar. And you know, their market is huge. Like, can you give me the green flags of these factors that you're looking at when doing DD? - Well, it's kind of interesting, right? 'Cause if you pull someone out of open AI who might be great at AI, like chances are they've spent a lot less time in Web 3, right? So the challenge I think with our space and making like those really great investments is catching the people who are bridging that gap. Like by default, open AI people are not gonna know about token economics. They're not gonna know about TGEs, they're not gonna know about going to market, they're not gonna know what it's like to build a community on Twitter for like the Gens, right? All these things that are like very cultural to the crypto space. So it's always a delicate balance. It's like, yes, you want people who can, I don't know, advance the state of the art of AI or show traction or have connectivity in the real world where they can credibly get a customer funnel into their product. But also if they're like doing Web 3, they need to know all these other skills too. So how do you find that balance? It's hard. - What about red flags when doing diligence? Obviously there's, you know, this versus no relevant experience, there's like the basic things. But are there any unusual or kind of like not often spoken about red flags when you're doing diligence that you're like, oh my gosh guys, hey, this person, you know, based on their history or XYZ clear red flag, we have to skip this deal. - Are you asking an AI or like in general? - In general, just when doing diligence on any sort of startup that you're looking at. - I mean, look, it's hard to be like super, you know, set in stone on this and super prescriptive. - It's hard. Yeah, like I would say, you know, at the end of the day, there's a lot of very different approaches and, you know, some founders get away with like raising money from, I'm not saying from us, but they get away with raising money without a deck. Just on the trust that they build with their, you know, investors and others do not get away with that. So like what's, you know, at the end of the day when the return is created, like everything is great, right? But what are the red flags? Well, I would say, you know, just very, very generally, I think that there's a sweet spot around a number of founders. I think like two or three founders are great, but like one and four plus tends to like perform less well for various reasons. I think you definitely want to try to support like early entrepreneurs sometimes because they turn into the best projects than the greatest founders, but there's also something to be said for kind of coming at a problem with experience in that domain and that being a really, really important aspect of success. So that's why, you know, that's why I think we see the like people with kind of a proven track record of, you know, whatever it is, exist or at least raising money. I know a coin fund, especially in this last seed fund that we've done, we've had it like a really phenomenal graduation rate by which I mean that, you know, the precedes or seeds that we've done in that fund, you know, overwhelmingly find the next round and find a lead investor for the next round. And those investors have been really, really great. Like we co-led or else with USB in the seed round, right? Or founders fund followed on to Primint Elects with seed in the series A, with seed we led, right? And one of the key sort of criteria that might not be like, as obvious is like, can this founder raise the next round? (laughs) Like sometimes, you know, there are ideas that are worth, you know, trying till the very, very end and that could take five years, but what's gonna be the difference between the success and failure of that project is whether the founder can raise money, you know, in that interstitial period. So that's another thing that might not be as obvious. - Yeah, and building off that, you know, like founder personality traits that you guys look for or that are just notable, it's also difficult because, you know, if a founder is, you know, hard charging and kind of airing it in some ways, that, you know, in some sense can be positive, in the sense of like, they will stop at nothing, nothing, they will keep going. Alternatively, you want the founder to, you know, certain founders, you want them to be less, arrogance and more, you know, charismatic in terms of their, or more like, you know, relaxed in their communication with people to be like, hey, you know what, you're right, here's why I believe this XYZ. And so it's like, there is no, there is no one rules all personality trait, and there's pros and cons to each, but in your opinion, what are the personality traits that you like to work with? - Well, I mean, I think it's, I think like, as an investor, in many of my portfolio companies, I get involved, or at least that's my goal is to get involved, especially the deals that I lead, you know, as a, you know, for example, with, with World, which is a, which is a phenomenal deal that we were very early in, I was just over in Toronto and like, went on CoinDesk TV and was talking about World helping them out with their PR and just making sure that everybody in the world was aware of the product and effectively launched in the US and the possibilities there. So I, you know, I tend to work with teams that are great, teams whose products are working in terms of personality traits in varies, but it's like, you know, if you're working with someone in a close capacity, are you guys getting along? One measure of that is like, how much is the founders susceptible to feedback? Like, there's founders who will listen and integrate every piece of feedback and there's founders who will integrate none and I think there's a sweet spot. I think like, sometimes they will and other times, they have more information than the investor and they're better placed to make that decision. So I respect that kind of founder. You know, it's also true that you like, don't necessarily have to personally like everyone, you know, on a very deep level. I don't see why that investor has to be true to have a productive working relationship. I mean, I used to work at Amazon and an organization of hundreds of thousands of people and I'm sure like we got along fine at work, but maybe we don't have to kind of hang out outside of work. So like, you know, but I think at the end of the day, you know, the really successful founders will be the ones who whether or not they're nice people will act in a professional manner and in a fair manner. And that's a big part of like building trust with your investors like, how much your transparency are you giving down? How much trust are you building? Are you saying what you're gonna do and then doing it? - When you're looking at different investments at Ed Seed, are you leaning more equity, more token or 50/50? Like what is your ideal structure? - I mean, I think if we're integrating Web 3, then you know, ideally that means there's a, token that is adding something very competitive, right? So like I would say in SuperState's case, the fact that you're tokenizing equities and and treasuries and different strategies is a really important aspect of that strategy like the token sort of is the thing. Now will they have a SuperState dedicated token? We'll see or we're also like investors in blockade and blockade is a very like traditional infrastructure as a service company that's really you know their main focus is like providing kind of tooling and nodes and infrastructure is their room for a token there? Who knows? But I think in in the middle like 90% of deals that we do, I would say it's typically structured as something like an equity investment potentially with board seats depending on the stage and so on and some kind of token instrument like whether that's an optionality I'm token if the token is created in the future or whether that token is like a core primitive in that system that will actually capture most of the value long term. I would say that's a big deal and I'll say one more thing you know why do we like crypto funds over like traditional VC funds? Well I think the the new asset class that we are operating in. First of all is creating all these like new business models and interesting networks and like ability like openness and ability for like people to participate in them and that is all done through tokens. But the other implication of having this liquid asset class is that I think crypto funds can achieve larger returns and I think they can achieve larger DPI you know cash and cash returns to their investors and I think that's a really important aspect for the LPs of these funds because you know DPI has been I think harder in the traditional world. Okay so on tokens I feel like in the past I don't know we'll call it six months two people been talking about token fundamentals again it's you know it goes in and out of style as you know very well. But recently it's been all about okay tokens what you know ideally they should be used to the company you should generate cash and then the company protocol should generate some sort of revenue and the revenue should be used to buy back the token and burn it or some sort of mechanism like that. But tokens have basically unlimited design space you can do so many things with them in your opinion and this is a very broad question because it's like what what it's a product doing is kind of that Bob questionnaire but in your opinion what is the ideal use case of tokens should they be used as these kind of quasi equity instruments to you know I use revenue to buy back the token and burn and drive value or is it more of like kind of the Bitcoin style where I know this is an incentive mechanism for a certain behavior we want to drive on this network. I mean that is that is an excellent question Andrew in a very very very difficult question answer I think look I think we've seen some incredibly innovative like I would say Bitcoin is a like coordination mechanism instrumented by a token or you know a large part of why people by ether is explained by that same mechanism or like the fact that we used helium tokens through bootstrap a global telecom network that otherwise would take like tens of billions of dollars and as next to impossible right like these are all like very very innovative use cases that I think showcased some of the power of tokens and by the way you know and on some measures like Bitcoin is the largest computational network in the world right like it's it's the largest super computer by hash power right and you know that's an impressive demonstration of tokens I think that there's other experiments that we've run that have been like less successful I think you know one experiment is in the 2017 as you know we went through this like exuberance with ICOs is a new form of capital formation and I think there's like nothing particularly wrong with like forming capital by getting it from community members in fact people today do it all the time in these like echo sales and beach and sales and there's definitely like a value to that what sometimes people say hey you know like this capital formation method will just be better than every other capital formation method or you know obsolete like venture capital and I now having worked in venture capital I I appreciate why that might not be exactly right and one of the reasons is like when you're building a blockchain you're not really building a blockchain you're building an organization that is building a blockchain and when you raise money from like very very unsophisticated market participants or just sort of there to like hold your token they're not helping you with this really really difficult problem you're running a startup you're trying to hire the best people you're building like the best organization you can build you have to constantly you know solve problems right retail participants don't tend to be cost company builders but there are professional investors that are and also professional investors like they kind of get a bad rap for like you know asking for a discount or something like that but one of the reasons that this kind of is there is because their long term locked into the stock and supply so you know like everybody gets their tokens the traders exit two years in but the investor is investing with the team for four years and by that time the four years comes around maybe like the prices and is good because of a volatile market right so to control for that risk there there's those mechanisms but the point I'm trying to make is that like actually like professional investors are much more long term aligns with what founders are doing than you know the average like retail network participant is aligned with them who can just sell their token at an opportune time for them so yeah I don't know thoughts on that yeah okay so I guess that building off that tokenomics because I feel like I feel like these go kind of hand-to-hand but for is there a red flag that you identify that you've identified or your team has identified with tokenomics where you guys see something I don't know you know just to make something crazy but like 70% of token goes to team something outrageous like that we're like okay red flag like you know this deal was amazing and now we can't do it because the tokenomics are so messed up or so it's like is there some major red flags that you you've identified within tokenomics and then on top that if you're if you had to go in and fix that red flag like all right guys can't you 70% of the team we're gonna do the normal 20% of the team or whatever it is what is like the ideal uh tokenomics kind of look and feel like I know there's not one size fits all but what's like the red flag and what's like the the in a perfect world I mean look tokenomics is another one of these areas that you couldn't design a token supply any which way that you want I think that you got to do a principle base so principle is the first law like as an investor you want to make sure that your team has enough incentive to be like really into their project and really like keep in going and getting motivated so you don't want to like have them to lip you know own too little of the token supply also when the team owns too little the token supply and they're being too idealistic that like no no no wait like the vast majority has to go to the public and I've been in projects like that which um you know which there are some like sometimes down the road they run into a problem where like they don't have enough token to sell to continue to the next round and that actually in danger is the project so so like one principle is like you've got to like give the team enough token material to work with to be incentivized and to have kind of longevity and the other principle is like you don't want short-term behavior like we've you've had plenty of experiments on this and the ICO boom of 2017 and different like supply economics but we always sort of converge back to the same thing the best teams who are going to be long-term successful are the ones who are long-term aligned and that's why by the way coin funds been here for 10 years supporting founders and hopefully we'll be here for another 10 years um we love working with founders who are long-term oriented so how do you achieve that in token supply you achieve that through investing you know very clear transparent terms of how everyone is dealing with their tokens um and finally I'll say you know well what is that line of like how much should go to the community how much should go to quote the insiders I know there's been this like kind of obvious sounding heuristic that like well at least 50% of the supply should go you know to the public and I think one of the motivations behind that kind of rubric has actually been like the lack of regulatory clarity it's it has been like the desire of founders and investors structure the token that supply in such a way that they would later be able to argue to Gary Gensler at the SEC that they're doing stuff that is decentralized, quote unquote, and fair and it doesn't fall under a securities regime. But is it really the most effective way of splitting up the supply? Again, I would argue in some cases, no. Because as I said just now, 50% of those retail participants are not going to help you build this company. So if there's a reason why your token should be distributed to them and they're very well-could-be. For example, in worlds, worlds ideal is to get out to as many people out there in the world as you can and distribute this token to have as wide distribution as possible. They think that creates the best customer acquisition on the network of any project. But in other cases, we've seen these tokens like be dropped into atresses that make the tokens effectively burned, they're being wasted. Or in other cases, rewards allocations are being pumped into teams that are not really doing very productive things to improve the standing of the network. And so it's really a case-by-case analysis of like, in my network, is giving away 50% of the supply, 75% of the supply, 80% of the supply, is that the right thing for this network to be doing? Well, it depends. And then, you know, I think it also goes the other way. Like, if you're giving out very little and most of it is owned by insiders and the intention of those insiders is to like flip this token supply and dump on retail as is the parlance of our times, then I don't think that's productive either, right? But everyone's goal should be building the best product, the best network, the best team, the best company, the best organization we can. And for every situation, there's sort of like a right way to do that. And you should just do that right way. I want to talk about a few themes that you are excited about. You keep mentioning world. So did they change the world coin, change your name just a world? World coin rebranded recently into world and world coin still exists as the cryptocurrency, but the overall project is called world. You have the world app on your phone. You have world ID, which is their biometric proof of personhood. And you have the world chain, which is an optimism based EVM blockchain that they get there running. Okay. So I am a admitted world coin hater only because of the reason that the orb is spooky. I want to give my eyeball to some orb, you know, like what is that orb for? Yes. I'm glad you asked Andrew. I have, by the way, an orb, I don't know if there is have an orb right here. This is the V2 orb. Is it actually the world orb? This is actually the world orb. I'm actually setting it up. So I'll be able to scan people who who visited me. A lot of the pushback around the biometrics of the orb, like the biometrics functionality the orb, in my opinion, Andrew has been like around people kind of misunderstanding or just not being clear about like how it works. So maybe I'll try to do it. You don't actually give your biometrics to the company or the orb. You simply use the orb to take the biometrics, turn it into what's called an iris code, which is a hash of your iris. In this hash, it's one way. It's like once you have the hash, you can't really recover the biometrics. The device deletes the biometrics and sends them to the user's phone. So you have an encrypted version locally, just like you have private keys in your, you know, crypto wallet on your phone. And that's self-savarin data. That's a very web-3 compliant way of posting this data. And then once the iris code is kind of in the database and you want to prove your humanities someone, they don't even send the iris code. They send the zero-knowledge proof that you own some iris code there, but not which one it is. And so it preserves even the privacy of that piece of information. And by the way, World Coin today is by far, it's not even close. The vast majority of zero-knowledge proofs that are verified on chain today is from World Coin. It's something like over 90% today. Which, you know, all is to say that this is the most privacy preserving biometric proof of personhood that you can buy in the market. You can compare it with some others. There's clear at the airport, you know, clear will actually store your iris biometrics in an actual cloud database that they own. They have the onus of protecting it from hackers. They also have a terms of service to say that they might be able to share your biometrics with third parties. Just like, you know, any startup. And I've heard people tell me like, "Oh, you know, I was at the airport and the line was very long and then clear was like, well, if you want to make your plane on time, you know, sign up." And then I did and then I gave my biometrics away and that's that, right? So this is actually the solution that prevents that scenario, right? That's what most people don't understand about World Point. And then, you know, once you have this identity, I would say it is the most credible, you know, decentralized identity play on the market. The app has 26 million registered users. 12 and a half million have been verified with this, with the orb. I think we just opened up in the United States. I was at the launch with Sam Altman and Alex Blania just a couple of weeks ago in San Francisco. We're now open in the US. There's six flagship stores where you, in the US where you can go and get verified, including Miami where I live. And also one in Austin, LA, San Francisco, Nashville, and Atlanta. And that's kind of the identity piece. And the reason is very simple, you know, like we've had app stores forever. I think in 2015, when Ethereum launched in the summer, I think like we had an app store right away. We've had decentralized app stores for 10 years, but they've never worked for one very simple reason. You need tens to hundreds of millions of thousands of developers on the supply side and you need tens to hundreds of millions of users on the demand side. And no Web 3 app store has ever had that. My proposition is that we will have that in World Coin soon. I mean, we already have 26 million. If we're not at 100 million by the end of the year, we're not doing our job. Like we will have just opened up in the US. The number of devices will proliferate from about 1000 right now to 8000 in the US. Orb mini will ship in 2026. Orb mini is a iPhone form factor device, which you'll be able to do everything you do on an orb. You know, there'll be like 20,000 orbs in the world. And it works. It's like the user experience is phenomenal. I want a single click log in into like a bar or lend protocol like Morphos currently on the world app store. And I already have some world tokens that have been promotionally like vend it to me. And I can try a DeFi protocol, which I couldn't really do before, right? Not without a lot of friction. So I think what will become very, very obvious very soon is that the world app store is going to be a major channel for like Web3 consumers who want to try to centralize the applications. Wow. Okay. Well, I am no longer a world co-inhater, but I still think it's spooky, but I like it now. I like it. Thank you, Jack. I appreciate you doing that. So tell me about crypto and compute because you know, I've seen a lot of these decentralized compute networks. And that to me has made a ton of sense because I'm like, man, if you're trying to provide compute to this global AI usage, being paid in US dollars, maybe you're in the UK and then you got to do the transfer. It's just a nightmare on payments. So sending a token makes a lot of sense. But I guess my kind of pushback on that is wouldn't it make more sense if they all use I don't know USDC or like ETH or just like the coin like some, if they are getting paid in some currency that we're already using versus those decentralized compute networks creating their own token or does it make sense to have their own token because they need a boot shop to network, etc, etc. Are you referring to any project in particular because I feel like I've seen like a, a, a cauch and you know, it's like some of these more where it's like, you know, hard, it's a quote, quote, quote, deep in or you know, hardware, not, not pure software. Well, I think I've seen a bunch of models, right? Like if you look at live peer, which is also a compute network that uses GPUs, you know, they're talking is very, very much supply side. It's like you have to use the token to join the network and to secure kind of your position. It's like collateral to make sure that your node isn't misbehaving or doing something wrong. Whereas when you actually want to buy compute in live peer, you know, you very much use a different currency, right? I think we will see. Maybe like a resurgence of hey, can we buy like resources with the native token of the thing? You know, we buy we use the theorem to buy kind of a theorem block space and computational power in that block space You know, there's other projects like we're gonna be Investors in a while. I won't say the name yet. It's not been announced But we're gonna be an investors and a decentralized Storage protocol which does programmable data, right? And so people would pay You know again to execute those smart contracts with a native token So I think you see like a few models there. I think you know when people do GPU networks today They're very very AI focused and I would say it's really what you're gonna do with that compute versus the Compute itself, you know, I think the compute Aspect of it is very like competitive. I think prices are being driven down I think if there's a shortage of any kind of like high-end computer day That absolutely Kind of goes away over time as more compute supply hits the market also there's um You know compute is like high-end compute is about 2% of GPUs in the world, you know So yes big tech owns 90% of high-end compute but night, but but that's only 2% of the GPUs in the world so there's a huge What kind of computational bandwidth to be assessed by going to kind of mid-tier or commodity hardware or consumer devices, right that are out there in the world um so When I see just kind of these pure plays um, I think that tends to be like less of a defensible business Then Okay, like I have compute and now I'm gonna do something really interesting with it. I'm gonna train Models into the centralized way or I'm going to trans code video 10 times cheaper than Amazon or you know some some kind of use case like that So on the decentralized AI You know, I guess what I have heard from some folks and founders is that um kind of the the primary reason that decentralized AI is so exciting is Because you know, we don't want all the the AI power and compute and would not to be held by a few companies. It's like well, we we want you know, a decentralized network and access to models and whatnot and Compute because it provides almost like a a safety mechanism Uh, you know instead of having these three corporates these gigantic companies own all that Is that is that how you see it or is there some also like monetary financial reason like oh no decentralized AI makes a lot of money sense because of xyz I think I think on average I mean like it really depends which issue you're sort of Disgusting, but on average, I think the centralized AI tends to be more expensive like we kind of talked a little bit about like the inference side I mean now I think there's ways of like potentially attacking that problem and making it cheaper but the fact is like With decentralized inference you always need to add a little bit of verification. You always need to add um Maybe like some privacy primitives and stuff and you end up paying more for that What's really really exciting on a training side is that people have always said well, you know I think Well, most people said decentralized training is impossible Then after it started to be like a little bit possible people started to say well, it's possible But it always be a lot more expensive And I think what we're actually seeing in practice is that At least this is true You know a couple of our portfolio companies working on decentralized AI training is that they've both done these Uh experiments where they like Run a model training process the traditional way using pie torch and kind of a data center And then they run the same exact model process under the centralized network And they show that the loss functions sort of like track each other so like we know that there's at least kind of approximate Efficiency parity more or less right now and this is very early stages of like centralized versus decentralized training Which is a really really encouraging early result So it could even be that like Like the centralized AI training is not more expensive But I think you know to answer your question I think for me like the primary reason why I want to invest into the decentralized AI stack and solve those really hard problems is because I want to see AI be built You know as a web 3 public good. I think what are what are in general blockchains and the centralized networks good for You know, I think I've answered that question for myself over the last decade and I think it is building Public goods that are collectively sort of owned and and live outside of You know kind of government ownership and corporate ownership I think we could do the same thing with AI. I think that's really healthy in a number of ways. I think it keeps the Um kind of the science of AI more open and lowers the barrier to entry like if you are a scientist You don't have to go work at Microsoft or something to you know try and experiment now you can you can crowd funds You know some resources and you could potentially or you will be able to train in train your model in the centralized network I think that's extremely healthy the other thing is like well, what is open in models Well, you know, we say that llama Meta's model is open But then Mark Zuckerberg goes on all these podcasts and he says, you know as soon as it becomes expedient for meta to close them We'll close them and Even though you know the weights of these models you don't really don't know what data they have been trained on You don't know what setup they have been trained on you don't know what hardware they have been trained on And you know when you talk to Alex of pluralist He has this really really great idea of openness, which is to say Openness means openness of setup. It's the data. It's the setup of how you train the model But then the weights can be closed and that's actually a good thing from a network perspective because it creates a business model Where the network can actually make money like no one really wants to pay service providers to Infraints, you know stable diffusion for or some of these open models because most people can do that now on their MacBook but when you're training a competitive frontier Decentralized publicly-owned model It's actually in the interest of the network to have the weights closed But the setup open if someone wants to go and compete with you just like GitHub right? It's like I can one click Clone your repository doesn't mean I will have the same number of eyeballs on that thing You have to develop my own audience. I have to distribute it and so on Save thing here. You could clone the model setup But then you have to provide your own resources to actually like train it and get people behind it But that creates the optimal competition environment while also Having a monetization scheme and also keeping the models open And I'll say one last thing is some people are scared of AI But I actually think the more transparent and open it is the less there is to be scared of because they can be Better analyzed. We'll see problems coming sooner. We'll have more eyeballs on the problem You know, and we'll have like better checks and balances if you know they More more power balance around You know who owns what kind of model so I personally think that like openness is the key reason why I am in decentralized the eye and I think it is Critical that we Develop open models on a competitive level to like the big tech companies You know if we want to maintain sort of the checks and balances in the world Yeah, I mean, I have to completely agree with that. It's it's we're talking about You know potentially in the future hyper intelligence and if there is a couple corporates that control that that is a very You know, it could potentially be a very large issue versus what the future you're talking about which is more community Hobbies professional driven which I think you know obviously That that is a feature that you know out prefer versus three or four corporate zoning or in the full stack Totally and I'll just say like There has been multiple instances where like senior executives and CEOs of these big AI companies have come out and said you know what like Our proprietary models will always be one step ahead of everybody else and we'll always own the biggest data center and I actually something that's true at all I think deep seek has invalidated that already in practice to some extent But I also see and many other people around me do as well now the fact that we could build much larger computational networks with decentralized Networks and GPUs that are not necessarily like high-end GPUs We can linearly scale them and we can potentially train larger models You know just as effectively as this company. So you know, there's a lot of like Evidence starting to mount that open source can win Um, we just need to put a little bit of energy and research and Into it and put some products into production and I'm very excited about that future Amazing all right jake it is time for the closing questions all right already So you have to tell your portfolio companies to build on one smart contract platform. For next. three years. So, you know, like, ETH, Seoul, whatever. I don't know. Like, you had to tell everyone in your portfolio to build on this network, which would it be? Can I split myself in half? Yeah. Yeah. It's, that's a very hard question. I mean, look, the general answer to that question is, like, companies will see a proliferation of platforms, and they have to choose the platforms that are right for them. I mean, like, this has always been our thesis. When we look that, the blockchain space in 2015, as Ethereum was coming onto the market, like we literally said, our thesis is a proliferation of blockchains, a proliferation of architectures, different trade-offs between these things, and high degrees of interoperability between these things. I do not think that, you know, that thesis has been invalidated in any way from, from, from 2015 to now, I think that there are some applications that are suitable for Solana, there's other applications that are suitable, you know, for, for L1s, for L2s. It really depends on what you're doing. You know, if you're running like a sovereign global intercontinental, you know, cross continental world currency, I think you should be on a dedicated L1. You know, if you're running some kind of, you know, more niche business that requires high throughput, you probably should be on an, on an app chain, I don't know, maybe like a ZK sync elastic chain or something like, like that. I think the bottom line though is we have over the last 10 years, we've solved the three key kind of pillars that we're missing in these steps. One was scalability, how do we get a lot of throughput? One was interoperability, how do we get our assets to, you know, interoperate between the chains and three privacy, like blockchains can't just be like transparent ledgers, like there are, there is proprietary information that needs to be kept. And I think we've solved all those problems. And depending what your application needs, you will be choosing different solutions. But all that being said, you know, I don't think like Bitcoin is going anywhere. I don't think Ethereum L1 is going anywhere. I think we've solved linear and horizontal scalability very, very effectively with L2 right now. I think there is a case for more monolithic blockchains, like I know some founders who are going for like a million TPS and they're saying, look, if you want to build the next Facebook, the next Google, you know, you've, you've got to do it this way and it's going to have this high transaction throughput, you know, we'll see. There's reasons why I'm skeptical of that, but, but there could be, there could be use cases for that. So Andrew, I can't tell you which one to use. It has to be dependent on what the project is doing. Very diplomatic answer. I will accept it. Unfortunately, I will accept it. All right. What trend is totally overhyped? What trend is underhyped? I think, I think most crypto people really don't appreciate the institutional enterprise adoption angle. Sometimes for fun in the weekends, I'll call up banks and I'll say, you know, hey, like what are you guys doing in crypto? What are your, what are the products that you're building? What's your outlook on the future? And like almost every call that I've had has shown me that, you know, it's conservative as these places are. They all generally have a positive outlook. Many, many, many of them have house views that, you know, banks will be a public blockchains. I don't think it's super clear yet. Like what the biggest products will be, like whether it will be like customers, you know, holding digital assets in their counts or, you know, tokenization of stuff or payments with stablecoins, we'll see. But I think like the headline is like, they're all really bullish on crypto. They just need the right like regulatory environment. And when they enter, you know, we'll just have an influx of users. It will be really great for everyone in the industry. So I think that's an underappreciated view. And I think the overhyped view is like, you know, these like ginormous infrastructure rounds like guys, we've been building supply side for 10 years. There's some really interesting tech. There's great innovation. We've made a lot of progress. But I think, you know, building the next marginally better, you know, slightly faster blockchain or slightly difference, you know, interoperability protocol is starting to be a little bit less valuable at the margin because what we need now is customers. We need product market fit. We need to go to mainstream users and really like resonate with them, educate more people about how to use these things. And until that happens, I don't, you know, I think those things are kind of right. Love that. All right. How is being an entrepreneur slash fun manager change you as a person? Ooh, I mean, it's a certain lifestyle for sure. A lot of travel, you know, dedication to the team. I really like this industry because my brain, for example, loves technology, loves finance. This is the reason I'm here, but mentally is that this is the best intersection between technology and finance ever. But it's also like the kind of technology that gets applied in all these different areas, you know, sure, we've talked about finance and we've talked about, you know, like blockchain and currency and like defile these things, but we've also seen architecture. We've seen you and I Andrew have seen digital assets, people trading cats on the blockchain. Just all these like different areas and people from those areas congregate in the space. And that is what keeps me going like my brain needs constant stimulation constantly seeing new things. I don't think there's a more creative space than like crypto blockchain decentralization technology. And it will be sad when we specialize more and they'll be more boring actually, but you know, it'll be fine. And I guess like I've had a lot of like learnings too. And like at the end of the day, this is going to sound really right, but it's it's try because it's true, which is that like at the end of the day, you know, your your startup, you know, as good as like the people behind it. And the people are only as good as like, you know, you can bring them in and teach them and manage them and empower them and they can grow into like leadership role. So it's like especially as a manager who, you know, fund management is like very much of people business dealing with Alpiz also dealing with founders and portfolio companies. And so I think those are the dimensions upon which I have grown the most. Does it not? I love that. All right. What is one piece of advice that you would give yourself before launching your firm? So before launching, you could say, you know, this one tidbit of advice, what would it be? I mean, self critically, if I were to do it all again, I would probably hire marketing much sooner and I would hire people much sooner. Did you go a few years or, you know, did you go a long time thinking, okay, I can run every aspect of the business or what was that? Where did that thought arise from? Well, I think in our case, it was a combination of like us starting from like literally scratch, like, you know, I started to coin fund with 5400 bucks with a Bitcoin. So some of it was like not having like resources to, you know, to hire everyone. We need to hire. It was a journey. Another part is just like some things were in our sphere of expertise like deep tech and blockchain tech, but marketing was less so. And I say that also as like, you know, kind of the most followed member of coin fund on Twitter and so on. But like, you know, just getting those professional people wasn't on the menu in those early days. And if I were to do it all again with resources, I would certainly like push the gas pedal more. Awesome. All right. Last question. Is there a piece of content that had a major major impact on your life, whether it be like a book, a movie, a podcast or even like a some series of shows or, you know, podcasts or whatnot that had a big impact on you? You know, like in crypto, I always think back to this one talk by Andreas and to Nopolis. And I believe it is called the future of Bitcoin. And I believe it was either from like 2014 or 2015. But in that in that talk, he describes how we have this in that inevitable future of token proliferation, which I understood as blockchain proliferation. And as I said, like, that always like resonated with me and was always like the backbone of everything that we did and how I invested, you know, in the space, like you kind of have to have a model of like, what do you think blockchains will do in order to like build models of what will happen in the future. And so, you know, a few thought that like one model would take everything and it would everything, which is beyond Bitcoin, that would be like one future. And if you thought that like there be millions of blockchains, that would be a very different future. And so, you know, I think that talk really influenced me into like the blockchain proliferation and reinforced that feeling that I kind of already had and that has been an important You know way in which coin fund was right about the market. That's one of the things that enabled us to be here today Amazing. All right Jake. Where can people learn more about yourself and connect with coin fund? X comm JBR UK H follow me and also coin fund is COIN F you and D underscore IO amazing awesome. Thank you Jake. All right. Thanks Andrew. Appreciate it