Wednesday. We're in this beautiful venue. I have to say I didn't know as a train person. I have become a train person after riding the little train here. Excited to be here. This is going to be a fun talk. We're not going to really talk about crypto that much. Maybe towards the end. So, excited to be here with all of you. Let's make this a little interactive because it's just a small group of us. So, feel free to shout stuff out. Not financial advice. Do your own research. This is us. All of this. And here we are. I don't know if you ever get philosophical, but I've been in crypto. This is year 11 for me. And sometimes I just take a step back and I say, 'Wow, we just live in the weirdest timeline.' And after spending a long time building two investment firms in the crypto space, I took a step back and I said, 'Well, what do I want to do with all of this?' Like, are we really just going to make better casinos for more degenerate gambling? Maybe there's more to this. I started my career in the oil and gas industry on the commodities side. And I started to get really excited about the first half of my career which was in energy and the second half of my career which has been in crypto. And then now I guess I'm starting the third half because this is a long career. But I started thinking about this energy to compute value chain that's emerging. So, here's a fun little thought exercise for you. On this chart, there are three assets that I charted. One is Bitcoin, one is Nvidia, which is the largest chip maker, and one is Vistra, which is one of the largest energy companies. Which line do you think is the green line? Is it Bitcoin? Is it chips? Or is it power? What do you think? Just shout it out. No wrong answers.
Yeah, power. Absolutely right. All right. What do you think is that blue line? The blue line at the bottom. Is that Bitcoin or chips? It's Bitcoin. Out of these three, Bitcoin performed the worst. This is a narrative violation. We love to talk about how crypto is the best performing asset class. Last year, Vistra, an energy company, was the best performing asset. Second best performing asset, Nvidia. Bitcoin did all right. The S&P did great. The Q's added 39%. So, we are living in an environment, we're in this little crypto bubble, but there's a huge bubble happening outside of crypto in energy and compute. And we need to go figure out instead of evangelizing to the same hundred people that you see at every conference around the world, instead of evangelizing to people who already have wallets, who already are invested in crypto, we need to go get people who are investing trillions of dollars in energy and computation and get them to come build with us and invest onchain. That's what we're going to talk about. I hope that excites you. If it doesn't, I will not be offended if you tune out. That's fine. Okay. So, the energy to compute value chain is a very simple one. You develop a site, you build a bunch of infrastructure, you deploy a bunch of hardware, you get GPUs, A6, TPUs, right? There's a bunch of different chip types. You get a bunch of hardware. It's very expensive. You have a bunch of software, you manage all the data, and then you have to manage all the energy and infrastructure around that data center. Okay, there are two types of companies that exist in this value chain. There are these hyperscalers that integrate the whole value chain. And the hyperscalers used to just sell managed services. Now they're developing their own chips. Apple has M1, M2, Google has TPUs. Now they're also becoming power operators, right? All of them have nuclear bets. They're building SMRs. They're buying IPs. They're bringing nuclear plants online. So the hyperscalers are going to vertically integrate. And then there's specialist firms who just work at one layer, but they work across these different compute verticals. So a great example is a data center infrastructure management software company. They build software that's compatible whether you're building a GPU AI data center, a classic CPU, cloud data center. You can use this across all these different verticals of compute. So that's the value chain. Value chains are really important because in every value chain, there's only two or three points of leverage. And our job as people building companies is to find points of leverage and to extract margin from those points of leverage. In crypto, the only place where there's leverage today is an interchange. If you own an exchange, you have distribution, you have durable margin that you can extract. Nowhere else in crypto have we found a durable point of leverage. So I need you to think about value chain. Energy demand growing materially after a decade of stagnation. Data center power demand accelerating double-digit compound annual growth rate. Insane. And it's not just data centers, right? It's not just AI cloud data centers growing double-digit compound annual growth rate. So there's a huge story here. EVs, double-digit compound annual growth rate. What do EVs use? They have to tap into the grid. Huge opportunity. But here's the thing. All of this capital expenditure that's getting deployed, right? And we'll talk about just the magnitude of money going into compute and power today, not just by PE firms, not just by the hyperscalers, sovereigns. Every single oil-rich sovereign is going to stop exporting barrels of oil and they're going to start building data centers and slinging compute. Why? Because like Bitcoin, compute is fungible and portable. A barrel of oil is not. Right? So if you're a sanctioned oil-rich nation, what's the best thing for you to do with your oil? Turn it into compute and sell flops. So new world order, we need to go back to the basics, right? I think in crypto, we've sort of gotten carried away with like the pump fun casino on chain. There are only three things that matter on planet Earth. And frankly in our known universe in our version of reality matter, energy and information and the global economy is basically just comprised of these three things right the largest industry in the world is the oil industry. That's energy. It's matter. It's atoms. We have lived in this world of bits and bytes of software for the last 15 years. And everyone thinks bits and bytes are what the world is built on. You don't have bits and bytes without atoms, without energy, right? They don't exist. So, we're now I think at this new point where we're going to see this huge step function shift in how we use matter, all sorts of new chip form factors, really cool advances happening in generative materials. We're going to see massive advances in energy. We need really dense base load power that is reliable. That is nuclear energy. I don't care how you slice or dice it. And we're going to need new ways of aggregating, analyzing, disseminating information. That's where AI comes in, right? It's this intelligence overlay. This isn't a physics lecture. It's very simple. The modern economy requires compute. Voracious amounts of compute and as a derivative of that, bandwidth and storage. Okay, compute consumes a fuckload of energy, an insane amount of energy. Therefore, the next frontier for our industry where we're going to get trillions of dollars of net new inflows, finding, funding, and building markets for energy and computation and also storage and bandwidth. That's it. That's my idea. It's so simple. If you listen to Santi talk yesterday about what he's doing with inversion, what he's doing, I think, is another sort of version of this. But I'm very much concerned with this energy compute. And then the third leg of the stool is crypto, right? Crypto is the only credibly neutral non-sovereign currency that settles with finality on digital rails. And then we have underlying it this new layer of capital markets and financial engineering. Right? And that's what we've done really well over the last decade in crypto is we for the first time in a very long time have built financial markets that operate entirely outside of the existing market structure. And we've created new financial contracts. The most novel one is the perpetual swap which was pioneered by BitMEX, right? But we've built new financial products that allow us to do financial engineering in these value chains in new and novel ways. This is very exciting to me. It's going to be very exciting to you hopefully in 10 minutes time. Okay, are you with me so far? Okay, we're all on board. Okay, so let's talk about Bitcoin. This isn't a Bitcoin conference. This is about Avalanche. But I just want to talk about we don't give ourselves enough credit. Bitcoin is the blueprint. Proof of work is what has made the data center economy possible. And I'm going to tell you why. Bitcoin has pioneered the energy to compute value chain and created for the first time thermoeconomic money. What does that mean? Bitcoin is secured by proof of work. You have to expend energy and do computation to make Bitcoin. Now Bitcoin is backwards looking, right? It's energy and compute that's already been expended. But it is very novel, right? If we look at some of the top companies in the AI space today, they started life as Bitcoin companies. They're really good at finding low-cost energy and mobilizing compute in new and novel ways. If we look at Bitcoin miners today, there are 24 publicly listed Bitcoin miners with 24 billion in market cap. That exceeds the market cap of pretty much every single other blockchain. Just for context, okay, CoreWeave alone, it went public at 36 billion. That's half a Solana. That's one-fifth of an Ethereum. The numbers here are staggering. So the bubble we are living in in crypto is really tiny. We need to think much bigger. And Crusoe, by the way, they won the first two Stargate construction contracts. Stargate is the half trillion dollar data center that's being built here in the United States. We're in the UK. So, let's be really controversial. The AI boom would never have reached this magnitude without Bitcoin capitalizing hundreds of billions of dollars of data center buildout over the last decade. Fueled by lowest cost electrification, Bitcoin is a marginal buyer of excess energy. Driving persistent demand for hardware, ASICs have been at the frontier of hardware innovation. We're now on three nanometer and two nanometer designs. We're pioneering optical computing, right, which is non-transistor based. So you can get smaller than 2 nanometers because you're computing with light as opposed to atoms of energy. And we're pioneering behind the meter collocation. So this idea that I can have a latent source of energy that's not connected to any grid and I just rock up with a containerized Bitcoin mine and all of a sudden I'm tapped into the Bitcoin network and I'm tapped into global capital Bitcoin markets because I can trade Bitcoin anywhere without needing permission from anyone. It is really beautiful. This is what it looks like. This is what I used to do. I used to work in the oil field in West Texas where I started my career. This is where I'm spending my time now. Okay. This is what crypto is about. This is the future of crypto. So, let's talk about it. Investment framework. So, human progress is defined by our ability to manipulate energy and atoms over spacetime. Never in the history of humanity has there been energy transition. We're not going to use less energy. We're going to use more and more and more energy. We just need to get better at extracting energy. We have a giant nuclear reactor in the skies. It's called the sun. Okay. PV is so photovoltaic, sorry, so abundantly cheap. Putting solar on the grid, very cheap. Battery costs coming down. So the most obvious use case for blockchains and tokens and what's happening in the Avalanche ecosystem in my view, find, fund, and facilitate the most efficient use of these thermodynamically constrained resources at civilization scale. Not just here on planet Earth, but beyond, right? China this week shipped a 2 gigawatt space data center into outer space. Like, what the fuck are we doing making casinos on chain? We have people colonizing space and building data centers. What are we doing, guys? We have a ton of money. Let's do some really cool, innovative shit that actually matters to humanity. I'm very excited about this. There's a huge opportunity here, right? Crypto is this little orange bubble. We're like a $3.5 trillion asset class today. Tiny. I want crypto to be a massive asset class. But where are those inflows going to come from? It's not just going to be from people buying coins. Okay, we need to actually build things for which there is persistent fee paying demand. Today there are only a number of blockchains that have persistent fee paying demand that are sustainable. Most blockchains today are fully dependent on inflation in the form of token inflation and foundations funding their development. That's not sustainable. Okay. So, we need to find net new use cases that bring in fee paying demand. Credit bubble, right? All of the biggest credit firms in the world have become infrastructure credit firms. Hundred billion dollars in credit and growing. Every sovereign investing in energy and compute. Every massive company investing in energy and compute. We as an industry could be massive. What are we doing? So, let's build the future. This is what I'm investing in. New protocols, orchestration, optimization layers, net new commodities, and markets where we can trade them and aggregators and marketplaces. I'm going to walk you quickly through examples of each one and how it relates to what's happening on chain. I'm going to make these slides available for download so you don't have to take photos or try to read too fast. I like to cram a lot of information in. This is my problem. But they'll be available on my Twitter feed. You can download them, use them, copy them, do whatever you want with them. So, really cool example of a protocol is Protocol Zero. They're building a new dedicated ISP. So they're taking dark fiber. It's not public internet fiber, but private fiber that's owned by high frequency trading firms. And they're building a new ISP, a new fiber network on which we can run distributed systems much more quickly. Super cool. If you can imagine, if we can speed up connectivity, right, if we have higher bandwidth, higher throughput, we also can do disaggregated storage. All of a sudden, edge computing, edge storage becomes much more viable. So in my view, this is the first of many, but we'll see decentralized bandwidth, right? In a way that's actually viable. We'll see decentralized distributed storage. You don't need a token for this. You don't need a token. You just need to build a network and you need capital partners who own physical infrastructure that they want to dedicate to these things. So I'm very excited about this idea that we can go lower into the OSI stack beyond just the networking layer and actually start to get into the hardware layer and building new routing topologies, right? That leverage some of these novel traits of how blockchains function. So we've done a lot of engineering at the networking layer. What are we doing at the hardware layer? Right? Bitcoin is the only ecosystem today that has its custom hardware. The ZK ecosystem is trying. Right? In Yinyama and a few others are making ZPUs or dedicated ZK processing chips. What are we doing here? Let's go into those lower layers. Let's build our own dedicated hardware. Let's stop paying AWS and Google Cloud to run our compute workloads. Let's run them on crypto-native platforms and crypto-native distributed systems. Very exciting commodification. Okay, I get really excited about commodification. If we think about compute right as commodity we started with on-prem, right? You had to run your own bare metal, very expensive, very high upfront costs. Then we had this whole era of cloud that made web two possible where all of a sudden you could one-click deploy a container right and run it anywhere. That made compute more of a commodity. And now DePIN and DeFi combined together are working on financing the buildout of compute capacity in new and novel ways. A bunch of projects working on that and constructing compute networks in new and novel ways where you can consume latent compute capacity whether that's on a computer in a data center potentially even on someone's phone, right? That's what EXO is trying to do with their approach. Really exciting. Optimization and orchestration. So you have all this hardware, you have all these energy systems, how do they talk to one another and how do you actually drive profitability there? I think there's some really exciting opportunities. Blockchains are exceptional at solving complex coordination problems, right? And so this is an area where we're spending a lot of time. Made a number of investments in how we actually go from single player mode where you optimize one data center to optimizing a whole grid or optimizing a whole set of data centers. So I think blockchains are really beautiful for aggregating, optimizing and orchestrating using tokens or even stablecoins as an incentive mechanism to drive that coordination. And then last but not least, talking really fast, you can download these, pick them apart, aggregation of marketplaces. Daylight's a really exciting one. How do we take physical infrastructure that's really difficult to aggregate? So like power plants, right? Really simple. All of these consumer solar installations, consumer batteries, home smart thermostats, really difficult to aggregate. Cost of customer acquisition is very high. Lifetime value of a customer is very low. Pretty much every residential solar company has gone bankrupt. Why? Because the economics don't scale well. So Daylight's an interesting company. They're starting by building a platform where they're incentivizing people with their token to provide their residential energy data to the protocol. Developers, utilities, and other companies can build on top of this data set. You're effectively building a data topology. Then you can apply an AI like an agentic overlay to intelligently optimize demand response. So you can actually pay people to stop using power, to turn their thermostat up a few degrees, to contribute power from their battery into a local grid. And then you can actually start to see this pipeline where you start to see decentralized deregulated energy markets emerging where people can actually trade real-time electrons, right place, right time, right amount on chain. So these are the things I'm excited about investing in. This is how we win. This is how we win: energy, compute, crypto. When I started my career, energy companies were the biggest in the world. I worked at ExxonMobil, was the biggest company in the world, $400 billion market cap. Today, that's laughable. 2025, the biggest companies in the world are semiconductor companies, energy companies, and some software companies, which are now becoming energy companies, and hardware companies. Where are we going? 10 years from now, the biggest companies in the world would be energy companies, compute, crypto companies. Today, crypto is a tiny sliver of this. If we want to make it a bigger material slice, we need to invest in energy, compute, crypto. That's it. Thank you for listening. I'll post these online.
