# Arthur Hayes On The Early Days Of BitMEX, Where Bitcoin Is Going, And Why We Are Over-Indexing On AI 1000x · 2026-10-14 · 49 min · https://www.youtube.com/watch?v=clWKbuJG5-M ## Transcript Avi Felman Are you bullish on AI stocks? Arthur Hayes No. I think we're wasting trillions of dollars of capital globally. Avi Felman Seriously? Arthur Hayes Yes. Avi Felman What? Arthur Hayes Absolutely. I know I'm starting an AI company, and I have to have you explain to me why we're— Avi Felman Wait, wait, wait. You think that we're wasting— in data center build-out, or in what way? Arthur Hayes Yeah. Avi Felman What are we wasting money on? Arthur Hayes At a very high level— Avi Felman We're coming here from Token 2049. I have a very special guest today, Arthur Hayes. Thank you for sitting down with me. Arthur Hayes Thanks for having me. Avi Felman I think it's fun. You live in Singapore. Obviously, I don't think a lot of introduction is needed here, but I do want to start by asking you a little bit about your background. Arthur Hayes Sure. Avi Felman I know that you were born in Detroit, Michigan, right? Arthur Hayes Buffalo, New York. Avi Felman Buffalo, New York. But your parents— Arthur Hayes I lived in Detroit for a while. Avi Felman You lived in Detroit for a while. Arthur Hayes Yeah. Avi Felman Your parents worked for General Motors. Arthur Hayes Yes. Avi Felman That's right. How did you end up coming from a world where your parents worked for General Motors to founding the largest crypto perpetuals exchange in the world? You invented the perpetual itself, if I'm right. Arthur Hayes Yes. Avi Felman Take me through that journey. ### From Detroit To Crypto Arthur Hayes If you've ever— I love Detroit and Buffalo, but they're kind of boring, at least for me, in my own sense of being. As a kid, I always wanted to go somewhere else and travel. Obviously, traveling in the early '90s was a lot more expensive than it is today, so you were limited to what you saw on television in terms of your hopes and dreams. What I remember distinctly is that when I was in high school, I read a book called Ugly Americans. It was about this guy who graduated from, I think, Dartmouth or some Ivy League school. He was a second-rate football player, and he didn't know what he was going to do with his life. Somehow, he made his way to Tokyo and started doing index rebills on the Nikkei and different Japanese indices. The book talks about his time in Tokyo. I think it was maybe the late '90s, somewhere around there. I thought, "Wow, this sounds like a lot of fun. It's much more interesting than being like everybody else and going to New York or Los Angeles, or whatever most Americans think they're going to do after they graduate from university." I thought, "Okay, I want to go to China," because I thought that was the new up-and-coming economic power. This was around 2000, when I read the book. I made that decision in my living room right before I went to university: I was going to study Chinese and move to China. When I got to the University of Pennsylvania, where I went to undergrad, I basically took a minor in Chinese and studied the language. I did my first study abroad in— Avi Felman Can you still speak it? Arthur Hayes Very poorly. Avi Felman Yeah? Do you have a sentence to say for all the Chinese listeners out there? Arthur Hayes No, I'm not going to embarrass myself. Avi Felman Oh, man. Arthur Hayes I moved to Hong Kong for a study-abroad program and spent 6 months there at the university. I loved it. Somehow, I found a job at Deutsche Bank in Hong Kong, doing sales and trading during my junior-year summer. I got an offer to come back and work there, and I have never left. I've lived between Hong Kong, Singapore, and a little bit in Japan for the last 20 years, or half my life. My entire professional career is in Greater China, essentially. I've never had a job in America. Avi Felman Really? You've never— Arthur Hayes I've only worked out here. Avi Felman Wow. So you've only ever come back to America to visit, to see your family or hang out? Arthur Hayes Exactly. Avi Felman That's kind of nuts. Do you still feel American? Arthur Hayes I don't like passports as a general concept, so I don't think that you have to be American, Chinese, or Japanese just because you were randomly born in some country and they have a piece of paper that says this made-up name on it. I completely reject nationalism as a concept. I am Arthur Hayes. Whether that's American, Chinese, or Japanese, who cares? It doesn't matter. Avi Felman You hate the idea of a passport. I don't think I've ever heard anyone say that in my entire life. Arthur Hayes I think it's offensive that I have to present this piece of paper to get in and out of places. Avi Felman If it were up to you, there would be no borders—full globalism. Arthur Hayes Exactly. Yeah. Avi Felman How has it been? You were at Deutsche Bank, trading there. At some point, I assume you found Bitcoin and crypto. Tell me about how that happened. Was it like everyone else? Were you on Silk Road, or— ### Bitcoin Leads To BitMEX Arthur Hayes I was fired from my job at Citibank. I went from Deutsche Bank to Citi, trading ETFs in Hong Kong and Singapore. This was after the crisis. My first day on the trading desk was the day Lehman failed. Avi Felman Your first day on the trading desk was when Lehman failed? Arthur Hayes Yeah, when they filed for bankruptcy that Monday morning. Asia had opened right before New York, obviously. On Sunday night, they had declared Lehman bankrupt, and there was mayhem. I was just a graduate sitting on the desk with no idea what was going on. The head of trading was saying, "Don't trade with this bank or that bank." Everything was blowing up. Your concept of this amazing future you thought you had in finance—making hundreds of millions of dollars in a few years with your list of rock-star traders—was completely not happening. Avi Felman Wow. Arthur Hayes The riches that we all thought we would make in banking evaporated with the change in the trajectory of financial services after the crisis. I thought, "Okay, I need to do my own thing at some point." I got the opportunity when I was let go from my job at Citibank in 2013. Then I happened to find Bitcoin, got engrossed in the philosophy, read the white paper, and started figuring out how to trade this thing by doing arbitrage. That's what led me to the idea of creating BitMEX. Avi Felman I do want to ask you what you were feeling at the time, when you were on the desk and everything was falling apart. What was going through your head? Were you thinking, "That's over. We're done"? Why did you stick it out in finance if— Arthur Hayes It was a well-paid job. It wasn't like I did anything special. I was lucky. Some of my friends who had jobs at other banks didn't have a job after a few months. I had friends who had Bear Stearns offers, and then they were wondering, "Is JPMorgan going to accept my offer?" Previously, you would do your summer training program before you started work. You would just get drunk and party; it didn't really matter. But in 2008, they started saying, "Okay, we don't have so many jobs because of the crisis. If you don't score a certain score on your aptitude test during the summer training, now you either don't have a job or you have to go to commercial banking." All of a sudden, either half your pay gets cut because a job in commercial banking pays way less than investment banking, or you don't have a job at all. That was the environment we were all in. We were all thinking, "Okay, maybe we'll have a job. Maybe we won't have a job." I remember my managing director pulled me over to his desk one day. He said, "Do you think you're going to have a job in 3 months?" I said, "I don't know." Avi Felman He just straight-up asked you that? Arthur Hayes Yeah. Avi Felman What a psycho. Was he asking genuinely, or was it— Arthur Hayes He was genuinely asking. He didn't know. Avi Felman I thought maybe he was messing with you or something like that. I guess he probably didn't know if he was going to have a job either. Arthur Hayes Yeah, exactly. Avi Felman Do you think that ability to thrive on uncertainty has helped you in a way, or do you just not really think about it? Do you just power through? Arthur Hayes I don't really think about it. Obviously, if you come into crypto, especially when you start early and there's a lot less structure around it— You only survived if you were very cognizant that everyone was trying to pull one over on you. There were all these scammers, and you got very good at noticing the ways in which people were trying to pull one over on you. This was early crypto. Not that there's that much protection today, but there were a lot more things that could have gone wrong back in 2013, when I started investing and trading in this space. Avi Felman Before we get into that, you got into crypto. What were some of the craziest things that you saw in the early days of crypto? What were some of the scams or pitfalls that— Arthur Hayes I mean, I know— Avi Felman —you had to get through that people today don't have to think about? Arthur Hayes I know today everyone probably knows not to do this, but back in the day, everyone was so hungry for deals. It was like, “Such-and-such famous person wants to buy 10,000 Bitcoin,” or something like that. Avi Felman Right. Arthur Hayes “Can you help this person buy all this Bitcoin?” And you're like, “Wow, if I can charge a few percent, I can make a few hundred grand on this trade.” Thankfully, I was never scammed in one of those trades, but that was rampant back in the day. Avi Felman I remember that. It was so funny. You'd always get these messages like, “Hey, I have a friend who wants to buy 10,000 Bitcoin or sell 10,000 Bitcoin at plus 4, minus 4 from the market.” Arthur Hayes Yeah. Avi Felman And you're just like, “Yeah, I can introduce them to a desk.” And they're like, “Oh, they don't want to work with a desk. They want—” This random guy all of a sudden can move all this size. I just know it was always your least successful friend who found this person, and you're like, “How did you find him? How did you find somebody with $5 billion in Bitcoin to buy and sell?” I don't know. It's totally nuts. Those early days were crazy. What year did you found BitMEX? Arthur Hayes 2014. Avi Felman 2014. And you founded it with 2 other guys. Arthur Hayes Yeah, so I— Avi Felman So— Arthur Hayes —I went out in Hong Kong and found 2 other technologists. One was to help build a trading system, and that was Ben Delo. The other was to build the web architecture, and that was Samuel Reed. They were both in Hong Kong doing other things, but they were interested in crypto. I found them through friends of friends, and then I pitched them on the idea of building a derivatives platform because that's where all the money was going to be made in the exchange space. There were obviously successful exchanges back in 2013. You had Mt. Gox just crushing it, up until it wasn't. You had Bitstamp, Coinbase, and all the big Chinese exchanges. There were definitely exchanges being successful, so trying to operate a spot exchange was not something I wanted to start, even in 2013, because it was a very crowded space and there were clear winners already back then. Avi Felman Mm-hmm. Arthur Hayes Derivatives was something that didn't really exist. There was this one little exchange called ICEY Bit, run by 2 guys in the Caribbean who were anonymous. It was an interesting thing. They almost went belly up because their margining system didn't work properly. But they put in place some of the bedrock concepts, like inverse futures, and some of the standards that we still use today were invented by ICEY Bit. I said, “I think I can do a better job than these guys, so let's build BitMEX.” Avi Felman What was the problem with their margining, do you remember? Arthur Hayes Everything didn't sum to zero. Avi Felman Okay. Arthur Hayes In 2013, Bitcoin went from about $250 to $1,300 in 4 months. They were offering, I think, 3× leverage at the time. I remember the December-March futures contract was trading 100% outright over spot. If you sold the December contract and bought spot, you doubled your money in 3 months. Avi Felman Wow. Arthur Hayes I wish I had put all my net worth into that trade, but I did put it on. A lot of other people put that trade on, and the problem was that when the market corrected and that basis collapsed, they didn't have the margin on the other side. Avi Felman Right. Arthur Hayes The exchange wasn't good at liquidating people, so it had a massive hole. One of the large traders at Cumberland actually bailed them out to keep the exchange alive in early 2014. Avi Felman Really? Arthur Hayes The only reason we all got our money back as traders on the exchange was because Cumberland bailed them out. Avi Felman So did Cumberland then take a stake in their exchange? Arthur Hayes I don't know what the arrangement was, or whether it was a private deal between one of the large traders and the founders. But the only reason we all got our money back as traders on the exchange was because Cumberland bailed them out. Avi Felman So you were trading on the exchange at the time. Arthur Hayes Yeah. Avi Felman Okay. That makes sense. So then you started this exchange, BitMEX. What was the first product that you launched? Arthur Hayes We launched a standard, I think, quarterly inverse-style futures contract, Bitcoin-margined— Avi Felman Mm-hmm. Arthur Hayes —Bitcoin–U.S. dollar. Our margining system was a lot more conservative than OKCoin at the time, BTC China, and Huobi, which were the main Chinese— Avi Felman Mm-hmm. Arthur Hayes —platforms and our competitors by the fall of 2014, when we launched. It wasn't very successful because the leverage wasn't high enough. Avi Felman Mm-hmm. Arthur Hayes All the traders complained that there wasn't enough liquidity. They asked, “Why would we trade on BitMEX? We'll just trade on the Chinese exchanges,” because they were offering 20× leverage at the time, and it was quite liquid to trade these products. We had to pick our way through that and find our way to socialized-loss margining. We took some of the concepts being innovated in China around socialized losses and made our own little tweaks to them. Because our technology was better than theirs at the time, we were able to go to 100× leverage, while they were only at 20×. The main thing that kickstarted getting clients to want to trade on BitMEX was that we offered the highest leverage of any platform globally at the time. Avi Felman I think that's still the name of your holding group, right? 100X? Arthur Hayes Yeah. Avi Felman This is actually the 1000x podcast. There's actually a history here that I don't know if you remember. When I was first starting the podcast, I reached out to you and asked, “Do you mind if I call it 100X?” I thought it was such a great name, and you were like, “Don't do that.” Arthur Hayes Yeah. Avi Felman So I just added a zero. Arthur Hayes Yeah. Avi Felman I went, “Ah, that's 1,000.” So you did actually inspire the name of this show. What was the point when you realized, “Wow, this is really working”? You were getting some real volume. Was there a moment? Was it slow growth? Was there an event? Arthur Hayes Once we went to 100× leverage and Bitcoin popped from about $130 to $600 in Q2, Q3, or Q4 of 2015, we were cash-flow positive and making money. Not a lot of money, but at least we could pay our bills and all that kind of stuff, which was great, because for the first year we literally had no volume on the platform. We became a player in the ecosystem. We were nowhere near as liquid as some of the Chinese players, but at least people were trading on our platform because we offered something different from everyone else. Avi Felman Did they ever try to kill you, those Chinese players? Was there any conflict? Not physically—metaphorically, in a business sense. Was there any conflict? Arthur Hayes No, we all knew each other. Avi Felman Yeah. Arthur Hayes There was no beef, if you will. Everyone had their own idea of how to build a futures market, and we all just did our own thing. Avi Felman Interesting. I have heard some stories, especially once the industry got bigger, of people fighting and jostling a bit. I was curious if that ever happened to you. Arthur Hayes No. Avi Felman It was all just kumbaya with everyone? Arthur Hayes Oh, it was kumbaya. I was down in Hong Kong, and they were up in Beijing and Shanghai, so we weren't even in the same city. It was a different sort of ecosystem. Avi Felman Did you ever worry about the legality of what was going on? I guess eventually you had to. Arthur Hayes There weren't really any rules at the time, so you just had to cross the river by feeling the stones. Avi Felman That's a very strong Asian metaphor. Arthur Hayes Yeah. I don't know the Chinese term for that, but yeah. Avi Felman That was actually going to be my follow-up question, so you read my mind. When did the volume really start coming? Was that in 2017 or 2018? When did you think BitMEX really became a household name in the world of crypto? ### The Perpetual Swap Takes Over Arthur Hayes We invented the perpetual swap in May 2016 to solve a problem: our clients didn't understand the products. Crypto people being crypto people, they assume that you’re scamming them. They didn’t understand the expiry date, and they didn’t understand why the price of a futures contract was different from spot. These were mostly margin traders and people who had never really interacted with the trading system. Even a buy/sell, two-sided order book was complicated for some traders. We were like, “Okay, how do we make this simpler?” We were literally getting emails every single day saying, “BitMEX is a scam exchange. Why is all this stuff happening?” We also wanted to concentrate liquidity. We had all these different expiries, so how do we make just 1 product where everyone trades 1 product, and then compete effectively with the quarterlies offered by other Chinese players? We tinkered around with different designs for this perpetual swap. The first one we launched was a failure. The funding cost was basically referenced from Bitfinex—their dollar and Bitcoin rates—which weren’t updating fast enough to reflect the change in the market. The derivative was trading at a massive premium to spot because crypto was ripping in 2016. Overnight, we had to make some changes, and we came up with this self-referencing funding mechanism. That’s essentially the key innovation that allowed the perpetual swap to have staying power. We had this massive premium, so how do we charge people for that and create an interest rate high enough to punish people who are going with the trend, and then attract market makers to go against the trend? That was the funding rate, the premium index, the 8-hour lookback, and all these things that everyone else has copied since then. It was meant to solve the problem of the derivative trading at a massive premium to spot: How do we make it collapse? Avi Felman What was the process of building this? I mean, there was some complicated math involved with this, or…? Arthur Hayes No, it’s not really complicated. It’s more about whether your trading system can computationally do these things fast enough to process the transactions on a tick-by-tick basis, which was a nontrivial problem. Hats off to Ben and Sam, the engineers who actually implemented this stuff. This is why perpetuals launched in 2016. It wasn’t really until 2020, when Binance and FTX came out with their own products, because it was technologically very difficult to do this. You really had to rebuild your trading system from scratch to handle this type of product and margin everyone asynchronously, tick by tick. We were able to do it because we used KDB, which is a very esoteric language used in trading applications at banks. As you know, Ben Delo is a KDB engineer at an investment bank. Over time, everyone figured out how to do it, and that’s why, call it early 2020, all these other people could technologically offer the perp. It took them a while to figure out how to build the trading engine to support the product. Avi Felman Did you know at the time that you were building something that is now eating the world? Did you have this vision, or was it…? Arthur Hayes No. We just wanted to make our lives easier and have fewer customer support requests. Avi Felman Yeah. That’s pretty crazy. How does that make you feel? You look around now, and you’ve got the Hyperliquid of the world, you’ve got the Lighter of the world, and you’ve got the CME looking at it. Everyone is now thinking about perps. Arthur Hayes I think it’s great. We’re very proud of what we were able to accomplish. Avi Felman Yeah. Arthur Hayes I know we shut down BitMEX a few weeks ago, and that was a sad and happy day for us. At the end of the day, we created something special, everyone is still using it, and we’re accomplishing our goal. Crypto’s taking over TradFi. The CME—I think I saw a Bloomberg snapshot where the CME CEO, or whoever, said, “Oh, we can’t go to 24/7 because of some staffing issue or something.” Avi Felman Yeah, that was totally— Arthur Hayes It’s hilarious. Avi Felman What are you talking about, a staffing issue? Arthur Hayes Yeah. It’s a database, man. You guys trade quadrillions in notional contracts, and you can’t do 24/7 trading? Your computers can’t handle 24/7 trading? It’s just a joke. Avi Felman How many people were working at BitMEX at the peak? Arthur Hayes I think 250. Avi Felman 250. Yeah, I mean, the CME has far more than that. Arthur Hayes Yeah. Avi Felman You’d think that they’d be able to handle it, but that’s totally nuts. I do want to ask: Why was it a happy day for you to shut down BitMEX? I would assume that would be quite… ### The Cost Of Running An Exchange Arthur Hayes Because running an exchange means you are the bearer of billions of dollars of assets. Obviously, our assets under custody had declined a lot since the peak in 2019 or 2020. But think about CZ, for example. I don’t know what the signing architecture is at Binance, but I assure you the buck stops with him, and you’re talking about hundreds of billions of dollars, probably, of assets under custody. Avi Felman Mm-hmm. Arthur Hayes The same with Brian at Coinbase, and the same with Ben at Bybit. You make 1 mistake, and you could be out billions. That’s a terrible way to live. Avi Felman Yeah. That’s a lot of pressure, to be honest. Arthur Hayes So I’m glad that we’re no longer in the business of holding other people’s money in the exchange. We had a good run, made a lot of money out of it, created something super special, and were able to exit the game with no customer funds ever lost. That’s not something many exchanges in crypto can say. Avi Felman Basically, every exchange in crypto has been hacked at some point or another. You’ve managed to avoid the North Koreans, so props to you for that. But I am curious: Whatever happened to the insurance fund? Arthur Hayes TBD. No comment. Avi Felman But who owns it technically? BitMEX the company? Arthur Hayes Yeah, figure it out. Avi Felman Or Arthur Hayes personally? Arthur Hayes Well, anyway, if you need to figure out what to do with that money, you already know. Avi Felman How much was left in there at the end? Arthur Hayes 30-something thousand Bitcoin, I think—something like that. Avi Felman 30,000 Bitcoin. Hopefully that doesn’t hit the market anytime soon. But that’s interesting. I want to take a step back for a second and talk about 2020, because I actually remember October 2020. BitMEX gets hit with this action from the government, and Bitcoin doesn’t go down, right? It’s crazy. I think it went down 5%, and then it rebounded. To me, that stood out as 1 of the biggest indications that we were about to go into an insane rally. But I’m wondering from your perspective: What did you feel at the time when the U.S. government was finally saying, “Okay, we’re taking you guys seriously”? What was that like being on the inside? Arthur Hayes Obviously, it sucked, right? Avi Felman Yeah. Arthur Hayes At the end, everyone knows the government’s going to come after you. You’re fighting this war against banks and the entrenched system, and you just hope you’re not the scapegoat. Unfortunately, that was us. So you just have to deal with it and soldier on. Avi Felman Yeah, I guess that’s fair. It’s crazy today. You look around, and the CFTC and SEC are kind of letting anyone do anything. A lot of these exchanges are operating in basically the same way that you guys were, but nobody’s coming after them because people realize that this is the future now. Arthur Hayes Yeah. It is what it is, right? There’s nothing you can do. Life is path-dependent. Avi Felman Yeah. I guess part of it is life philosophy. You could really just fuck off and do nothing for the rest of your life, but you’re still out there, writing articles and giving takes on the market. I used to love your Pepe covers on the articles. Arthur Hayes I’m running out of Pepes. Avi Felman Yeah. Arthur Hayes We need to bring back Rare Pepes and continue the project. Avi Felman Look, the culture of crypto has changed dramatically. I’m literally wearing a suit. I spent 9 years in crypto, trading every day in my shorts and T-shirt. But now I feel like it’s kind of a costume for me. Arthur Hayes Mm. Avi Felman But at the same time, I think it’s just part of the changing times, and I’m curious what… Do you have a guiding life philosophy about what you want to do with your time? Arthur Hayes Enjoy. Avi Felman Yeah. Arthur Hayes Whatever that means for me is something different for everyone. I spend a lot of my time doing physical exercise. Obviously, there's intellectual stimulation, and writing and trading. It's a game to me, right? Thankfully, I have enough money that it doesn't matter if I make or lose money on anything I do. But to me, it's an intellectual exercise. Avi Felman Mm-hmm. Arthur Hayes And it's a game, and the yardstick is how much money you made. Avi Felman Yeah. Arthur Hayes And it's objective, right? You either made money or you didn't. That's why I love trading and financial markets. You learn about everything, because everything at the end of the day has a price attached to it. You have to determine how you're going to spend your precious time in this universe, and prices are one of the best ways to indicate that. So whether it's art, music, crypto, oil and gas, or AI, it all comes back to financial markets. If you really study financial markets, you study everything. It's a super interesting pursuit. Avi Felman Do you study a lot of financial history? Are you interested in it? Arthur Hayes Yeah, I read a lot of books on financial history. Avi Felman Yeah. Are you still trading actively today? Arthur Hayes I don't think I'm as active as some of your listeners who are trying to trade a 1-minute candle on Hyperliquid. Avi Felman Yeah. Arthur Hayes But at Maelstrom, I run the directional book. We do mostly directional, long-only positions. Avi Felman Mm-hmm. Arthur Hayes I'd say the average tenure of one of our more speculative positions might be, let's call it, 1 to 6 months. Avi Felman Makes sense. Honestly, the name 1000x is a bit tongue-in-cheek, because what we talk about mostly on the podcast is mega trends. We talk about where to allocate your capital so that you can sit in it, and we honestly discourage trading the 1-minute chart, because I don't think many people can actually make money on that. I think crypto was a lot easier to trade 6 or 7 years ago than it is today, in my personal opinion, on a short time horizon. But it's almost easier to trade on a long-term time horizon, because I think we've hit an inflection point where now it truly is inevitable. There was a period of time when it could have been killed by the government. It could have been stopped. Maybe we wouldn't have reached the point that we're at today if Trump hadn't been elected in 2024 and given companies free rein to build whatever they needed to. But now, at this point, I think it's really about long-term allocation. Besides that, what are you interested in now? Is 90% of your net worth still in crypto? Are you betting on this, or have you diversified out? If so, what are you looking at? Arthur Hayes Given that I have so much Bitcoin and other stuff, I'd say 90-something percent is in crypto. Avi Felman Okay. Arthur Hayes But if it went to zero, I have enough money in other stuff that it'll be okay. Real estate, gold, random stocks, T-bills. Avi Felman Are you bullish on AI stocks? Arthur Hayes No. I think we're wasting trillions of dollars of capital globally. Avi Felman Seriously? Arthur Hayes Yes. I know I'm starting an AI company, so I have to explain it. Avi Felman Wait, wait, wait, wait, wait. You think that we're wasting money in data-center build-out or in what way? ### AI Will Need A Bailout Arthur Hayes At a very high level, if we believe in markets—whether, you know, free state capitals or whatever, right?—we believe in the signals that markets give us. What's a signal? As a customer, I pay a price, and hopefully that price compensates the company offering the product or service for the cost of delivering that thing and some sort of profit for their effort. Avi Felman Mm-hmm. Arthur Hayes If I'm not willing to pay something at which a company makes a profit, then that company shouldn't be making it, at a theoretical level, right? Avi Felman Right. Arthur Hayes So we take a look at the main AI labs: Anthropic, SpaceX, and OpenAI. This is the majority of the development of frontier AI. None of them make money, and none of the customers are paying the price that it would cost them to make money. Avi Felman Yeah. Arthur Hayes If OpenAI, Anthropic, or SpaceX said, "We have these negative unit economics. We incorporate training costs and inference into the cost of delivering this token to you, the client. We're going to charge you enough money to make an actual profit," I don't think anyone would use AI. At least, I wouldn't say no one, but probably 95% of the people who use AI probably wouldn't use it if they had to pay the true cost, because it's expensive. Avi Felman Right. Arthur Hayes Right? Avi Felman But isn't that the point of the data-center build-out and investing capital in compute—to attempt to bring down the cost? Arthur Hayes But the thing is that even at that price, these companies don't make money. Avi Felman Right. Arthur Hayes The thing that they're delivering is a commodity. It's getting cheaper and cheaper and cheaper. Avi Felman Mm-hmm. Arthur Hayes But the cost to build more of the commodity—data centers, inflation in labor, inflation in materials—where do you get the energy from? You're competing with consumer grids. Avi Felman Right. Arthur Hayes And this is in the United States and a lot of other places. The more data centers you build, the more expensive it gets because there's a shortage of the actual materials and labor to build the thing. But the compute gets cheaper and cheaper and cheaper. So you have an upside-down chart in terms of the business. Avi Felman Sure. Arthur Hayes But we keep funneling money into them. SpaceX just did, what, Forty billion two days ago? A bond sale? Avi Felman Yeah. Arthur Hayes Right? Anthropic and OpenAI need to do something like $50 to $100 billion of debt sales every year just to fund the companies. Yet Anthropic told us in 2025 that it costs them $3 to generate $1 of revenue. That's a terrible business. Avi Felman Yeah. But what do you say about the comparisons with the early dot-com bubble, Amazon, Google, and all those companies? Arthur Hayes It's the same thing, right? Avi Felman Right. Arthur Hayes We are optimistic about a technology, AI, and its usefulness. I think it'll be extremely useful over time. But the problem is we're paying the wrong price for it today. Avi Felman Okay, interesting. So you do value AI as a product, as something that generates value for humanity. You're just saying these businesses suck. Arthur Hayes And we're wasting money. We're spending all this money on something that people don't want. Avi Felman Right. Arthur Hayes If they wanted it, then Anthropic and OpenAI would be super profitable. Avi Felman Makes sense. But are you taking any bets against it? Are you trading? Arthur Hayes I bet that they'll get bailed out. Avi Felman Yeah. Arthur Hayes The entire global economy—whether it's the United States, which is building the data centers, or places in North Asia and Southeast Asia, which are supplying the hardware, the chips, and the raw materials to America to build data centers—is in on the same trade. So if American AI—if the end consumer, OpenAI, Anthropic, and SpaceX—do not consume the trillions of dollars of compute that they've booked, that they said they're going to consume in the next 5 years, then the whole global economy falters. All these politicians who have basically come out and said, "AI is the future. Productivity is how we're going to get ourselves out of this debt issue that we've created over the last 80 years," will look stupid. And the people are going to say, "What? I could have had free healthcare, free university, better water, better food, yet we decided to blow $2 trillion or $3 trillion on steel and cement right now." Avi Felman Right. Arthur Hayes I still have no healthcare. How is that good for me as a person? And all my data was stolen by these companies, and I didn't get to invest in any of these things at the— Avi Felman Yeah. Arthur Hayes So, again, I think it'll get bailed out because it's too big to fail, and that's going to benefit crypto. Avi Felman And so, what is your trade at the end of the day? It's long Bitcoin, right? Long Bitcoin, but also I think the way that we operate is that we think in terms of probabilities. Is this, like, a 60%? Is this, like, a 90%? Is this a... Arthur Hayes I would say that, if I had to put a probability on it, I'd say it's probably 75% that these companies will not be able to afford the compute commitments without a government bailout. Because the demand will not catch up and pay the price necessary for these companies to have positive unit economics and sufficient scale to basically honor the contracts they signed. Avi Felman What do you think that does to the social fabric of the country? Arthur Hayes Nothing. Avi Felman Nothing? Arthur Hayes It's just a bailout. We did it in 2008 globally, with the U.S., China, and Europe. Everyone got bailed out. Obviously, at this conference in crypto, I think people are like, “Oh, yeah, hard money. Stop printing money.” I just happen to make money on it because we know it's inevitable. But if you're the average person, are you really going to support a politician, whether it's in a democratic or undemocratic system, who tells you, “Okay, I'm sorry, you might be 30 years old, but 35 years ago, somebody took out a bunch of debt and made all these decisions about the global financial system. They said, ‘Tomorrow, we'll pay for it,’ but tomorrow is today, and now your living standard must materially deteriorate to pay for this debt that you weren't even born when they put it into practice”? Are you going to support me as a politician? Absolutely not. You're going to say, “Okay, I'm going to be on the winning team, and if I'm on the winning team, I get to print money and hand it to myself,” whether that's in lower taxes or government welfare benefits. And so it's this factionalization. The U.S. is obviously the best example. This is red versus blue. Avi Felman Mm-hmm. Arthur Hayes Both sides print money. They might have different pseudo-philosophies about why they should deliver money to certain constituencies, but at the end of the day, whether you're a Democrat or Republican in the United States, your goal is, “I want my team to get elected so that I get the goodies.” Avi Felman Right. Arthur Hayes We are not saying, “I want my team elected so that the global economy drops by 50% because we're going to stop printing money.” That's not a political solution, not in China and not in the United States. So the average person doesn't want the thing that would stop crypto from going up. The average person says, “Print the money, make sure I get it, so that my life is okay.” Avi Felman Mm-hmm. Arthur Hayes That's the entire global political system. Avi Felman I think your take on whether it affects the social fabric is interesting because, from my perspective, if we get a massive bailout of AI, AI is already a hot-button issue on both the right and the left. Obviously, the right is pushing forward, but then you have governors like Greg Abbott saying, “Whoa, whoa, whoa, whoa. No more data centers,” and he's a Republican. And so, if there is a government bailout that has to happen for these AI companies, I see the U.S. swinging pretty hard left at a certain point because of that. I could see socialism— Arthur Hayes Sure. Avi Felman —really, really taking off. I mean, that would just be, I think, bad for ... I do think that crypto gets caught up in that as well. Arthur Hayes Right. But it'd be like Obama in 2008, right? He inherits this recession, this global financial crisis. He talked a big game about how he was going to hold the bankers accountable, yada yada yada. “You're not going to lose your house.” Avi Felman Did he not hold the bankers accountable? Arthur Hayes No. He specifically said that they were too big to jail. Avi Felman Right. Arthur Hayes I forgot the exact phraseology, but he and Eric Holder basically said, “We can't jail any bankers. It's too systemic to the financial system. Nor are we going to roll back TARP or any of these programs. We're going to do them and do them even more.” And so, “We're going to foreclose on your house.” Avi Felman Right. Arthur Hayes So all the people who voted for Obama thinking that he was somehow this ray of hope or change found that he was the same as everybody else. When presented with, “Okay, the global economy goes down 30%, 40%, some big number, or we just print some more money and kick the can another 8 years to the next guy,” they kick the can. I don't care what your political philosophy is, because do you really want to be the person who gets elected and then goes, “Okay, I'm going to crash the global economy. It's time to actually pay the bills”? Avi Felman No. Arthur Hayes So Democrat, Republican, socialist, capitalist, China, America—it's all the same thing. No one wants to be the guy holding the bag. Avi Felman Do you think it's the same in China as well? Arthur Hayes Absolutely. Avi Felman Yeah? Arthur Hayes The Chinese system has so many distortions that the Communist Party knows about, but they lack the political will to assign the losses. They do not know how to say, “Okay, we had a property bubble—the biggest credit expansion probably this century—over the span of 8 years, from 2009 to 2018. Okay, there's this loss. We built all these houses we didn't need. We did all this infrastructure we didn't need. Who bears the loss? Is it the local government? Is it the banks? Is it the central government?” They're all technically owned by the Communist Party. Avi Felman Right. Arthur Hayes But which part of the Communist Party is the one that gets stuck with the loss? They can't even decide among themselves, so they just keep printing money and keep exporting because they think, “If we just keep the status quo, it will be more harmonious, and we stay in power.” It's a different sort of situation, but it's the same decision-making. Does Xi Jinping want to be the Chinese president who goes, “Okay, yeah, I'm sorry, Hu Jintao and all these guys pumped up a bunch of credit. Now the material standard of living is going to drop 50% in China because we wasted too much money”? Avi Felman Right. Arthur Hayes Again, that's not a palatable political solution, even in China. Avi Felman Talking to you is making me a lot more bullish on Bitcoin. I'll tell you that because I did have that general take on the United States. I didn't know enough about China to have that take on China as well. So hearing that, the global economy is obviously interconnected, and if that is true across the board, in your opinion, that's obviously very bullish for BTC. But I'm curious about your take on where that ends. You take that to its full logical conclusion, and you're kind of saying that at some point, that bill is going to come due. Arthur Hayes At some point. Avi Felman And— Arthur Hayes I don't know when that is. Avi Felman But— Arthur Hayes Maybe there's a revolution. We're seeing signs of it, but in most places, people are so satiated with social media and just so dumbed down that they decide to opt out, right? Lying flat in China, quiet quitting in the U.S. and the Western countries—it's all the same sort of mechanism. We'll see. Avi Felman Just as a total aside, I did notice that there's quite a bit of a phone addiction problem on the streets of Singapore. I've noticed literally everybody is staring at their phone all the time, 24/7, watching short-form video. You see it in New York to an extent as well, although, from my experience, it's a little bit worse here. And it does make me think that people are sort of opting out of paying attention to the system now in a big way. I think that trend is probably going to continue. But part of why I decided to do this podcast and also talk about media is because, if you do believe that the U.S. government is going to print, if you do believe that the Chinese government is going to print, and if you do believe also that AI is going to be useful in the future, regardless of whether these companies make money or not, it's imperative to have a stake in that future. To check out completely, I think, is probably the wrong choice. You have to encourage people to say, “Okay, I will allocate to the things that are going to benefit in the future,” and get people into investing in the first place. One of the most heartening things that I see is that Trump has said that the S&P is now sort of the bedrock of the future of the U.S. That's why he opened Trump Accounts. That's why he's giving $1,000 to every newborn: because it's no longer enough just to make money. You have to be invested in the American experiment. That financialization of the world, which I think you actually contributed to in a huge way with perps, I personally think is good. Bringing people in to invest in the market is good. But I'm curious about your take: Do you think that what we're experiencing now with the financialization of the world is a good thing or a bad thing? Arthur Hayes I think it doesn't really matter. It's neither good nor bad. It just is a way in which, I think, if you ask Trump what his number 1 KPI is, it's “number go up” on the stock market, right? It's an objective measure. He's just a finance guy. Obviously, a lot of people who back him are in financial services. But even if you give $1,000 to every American, all you're basically saying is, “Close your eyes. Don't look at the 1% who own all those stocks anyway.” The $1,000 isn't going to materially change anything. How are you going to pay for the $1,000? You're going to print a bunch of money. It's going to help my bags, great, but is that really going to solve the quiet quitting or this opting out? I don't think so, because the second people see, “Okay, well, I got an extra $1,000, but my rent went up 50%,” it's like, okay, cool. You just created more money to chase more scarce goods like housing, food, and all these sorts of things—social services. So it doesn't actually solve the problem. Maybe some people think that they now have an ownership stake, but when they actually look at the numbers, you're just as poor as you were before. Avi Felman Right. You're sounding a little bit like President AOC. Arthur Hayes AOC isn't going to be any different from Trump, right? Avi Felman Right. Arthur Hayes Trump, AOC, and Mamdani are all the same. They just have different rhetoric around what they're doing, right? Trump gave the largest money-printing exercise probably since FDR in the 1930s with his stimulus checks, which Biden continued because it was great politics. “I'm going to shut you in the house for this virus that may or may not be a hoax. Don't do it. Don't worry about it. Here's a few thousand dollars.” If you're a business, “Okay, paycheck, whatever it was called—we're going to cover your costs, right? We're going to give you hundreds of billions of dollars.” Democrat, Republican—everyone supported this. No one says no to free money. Avi Felman Yeah. Arthur Hayes They can give it to their supporters, and then their supporters come and support them at the polls. Trump, AOC, Mamdani—all the Democrats, whatever, the socialists—they're all the same. They just couch it in different rhetoric. They have a different shtick for how they present it to the people. But at the end of the day, it's money printing to get votes, and it's very effective. Avi Felman What do you think about Zcash in this context? Arthur Hayes It's great. I love the privacy narrative. Obviously, there have been some issues with the minting and all that sort of stuff, and I think formal verification will make its way into the protocol sometime in Q4 this year, I think. At Mailstream, we're an investor in Zodle, one of the leading apps for Zcash. So I love the narrative that will hopefully fix the blockchain issues, and, yeah, I like it. Avi Felman I do want to ask you about this new thing that you're working on. Like you said, it's kind of confusing to me because you seem quite bearish on the AI companies, and you're building this. You also potentially might have access to the full insurance fund, in which case you could just do whatever you want for the rest of your life. You're deciding to come to a conference in Singapore and spend an excessive amount of time building, talking, and doing something new. Why? What is it that you're building, and why are you building it? ### Building Money For AI Agents Arthur Hayes It's called the Flop Network, and it's basically trying to address the issue of—I personally believe that the agentic economy is real. Avi Felman Okay. Arthur Hayes There will be quadrillions of these things out there, this new life form, and they're going to do useful stuff for us and for themselves, but they don't have a payments network. Avi Felman Mm-hmm. Arthur Hayes When we're born, we're pretty much born into some sort of payment network. You can't change what the reserve currency is, even if you hate it with your whole life force. But for agents, there is no dominant currency for AI agents. Avi Felman Right. Arthur Hayes Whatever network services them will be the most valuable network ever created because you have the most economic actors growing rapidly, exponentially. They can communicate with each other 24/7 at the speed of light in a way that cognitively we can't even comprehend as humans. Imagine if you're the network that powers that coordination, because coordination requires scarcity. Avi Felman Mm-hmm. Arthur Hayes The scarcity here is compute. If I'm an agent, just like a human eats calories that we pay for with fiat currencies or earn from our jobs, an agent consumes floating-point operations per unit of time—compute—at a data center provided by GPUs. The question is: Is there a currency that they can hold that directly converts into compute? Avi Felman Okay. Arthur Hayes That's what we're creating at Flop. Using the principles of Bitcoin, there are miners who come on. They provide GPUs and do useful inference for agents. They get rewarded with a block reward in Flop and transaction fees in Flop. Now you've tied this made-up currency called Flop directly to compute, the thing the agent actually needs to exist. The reason we think we're a contender to be the payments network for AI is that AI agents will have a nonzero convenience yield to hold the currency. Just like when you get paid, you don't spend all of your money. Even if you're the most broke person, you still have a few cents in your wallet. Avi Felman Yeah. Arthur Hayes Right? You don't just go out and buy all the food. Maybe if you're in hyperinflation or whatever, but in normal society, you don't go out and spend all of your money. You have a convenience yield to hold some of it because you know that whenever you need to, you can walk down the street, buy some more calories, and feed yourself. An agent will hold a small amount of whatever the currency is of this new network because it knows that whenever it needs to—if it has a job, if its human tells it to do something, whatever—it can convert Flop into compute directly. If you have trillions and quadrillions, or however many agents are out there, all holding a little bit of Flop, then the network value gets insanely big. That's the bet we're making with Flop, and that's why I'm super excited to essentially come out of retirement to do this. I think it's a massive opportunity to possibly create the largest cryptocurrency ever, bigger than Bitcoin. If you think about the economic actors that support Bitcoin, it's a few hundred million humans, right? Hopefully, we get to 8 billion. Avi Felman Yeah. Arthur Hayes There are already billions of agents. We don't really know exactly how many. The amount of commerce they're actually doing is probably not that much yet— Avi Felman Right. Arthur Hayes —but it's growing quickly. Look at how successful Facebook's Muse consumer application is and how useful that's going to be for a lot of people—to be able to run their entire digital lives with an agent that can book their hotel rooms, cancel subscriptions, whatever, right? This is just a start. Imagine when all these AI data centers come online and all this compute is there. There's no buyer for it on these take-or-pay long-term contracts because all these labs can't afford it, so they dump it on the spot market. Now the agents have cheaper life force, cheaper compute, so we can make them do more complex things. We can have more of them. We can embed them more into different parts of our lives. Why? Because it's really, really cheap to have millions of agents do something. This is how OpenAI and Infrapark are solving these math problems. It's not that agents are uniquely super smart. Yes, they are smart, but if you have unlimited compute and the ability to spawn instances of life, then you can do a lot of cool things. Avi Felman Yeah. Arthur Hayes Imagine what we can do with super-cheap compute. Everyone has a fleet of agents doing things for them. It's going to be insane. Avi Felman Did you get approached by somebody who had this idea and then decide to come out of retirement, or did you have this idea and then go build it? Arthur Hayes I had this idea, and then I went and found my team. Avi Felman Oh, really? Okay. Arthur Hayes Yeah. Avi Felman How did you go find your team? Arthur Hayes Our CTO is someone I worked with at BitMEX. He was on the trading engine team—a very, very good engineer. Our CSO is someone I've known for 20 years in banking, a good friend of mine. We have 3 of us and a lot of AI agents. I would never do this without AI. Avi Felman What— Arthur Hayes Because— Avi Felman Yeah. Arthur Hayes If you have proper people who have real experience, then they're able to direct AI agents to do useful work in a responsible way. Avi Felman Did you ever think about reaching out to your old BitMEX founders? Arthur Hayes No. It's a completely different sort of situation. Avi Felman Right. Are you in touch with them at all? Arthur Hayes Yeah. Avi Felman Yeah. They're super under the radar. They don't really seem to like the spotlight very much. Arthur Hayes Yeah. I mean, they're around. Avi Felman They're around. Okay. Arthur Hayes Yeah. Avi Felman Well, do you think I'll see them at any point during this conference? Arthur Hayes Yeah. I've already seen one of them. Avi Felman Oh, really? Arthur Hayes Yeah. Avi Felman Okay. Interesting. I want to ask one final question, because it's just super interesting to me. You were pardoned by Trump. Arthur Hayes Yes. Avi Felman And what was that process like? There was a 60 Minutes piece that just came out. Arthur Hayes It was about the rapper, right, or something? Avi Felman Yeah, and it was about this idea that you had to go through a bunch of these pardon brokers and pay them a ton of money. Did you just wake up one day and you were pardoned, or was this an effort on your end? ### Arthur Hayes Gets Pardoned Arthur Hayes Getting a pardon from any president is essentially a lottery ticket, right? Everyone—there are lots of people who want one. There's this one guy who determines what he signs, essentially, and you kind of have to figure out a way to make your compelling case. So, at a very basic level, you just employ lawyers, and they write a brief for you explaining why you deserve a pardon. Obviously, Trump had put it out there that he was considering giving pardons to folks who've been wronged by legal action and all that kind of stuff. So you're like, "Okay, wow, let's just try it out," right? Yoink. The lawyer puts together the application. They put it through the correct channels, and it either gets signed or it doesn't. Avi Felman Okay. Arthur Hayes It's literally that banal. Avi Felman There are no briefcases full of cash and drops? Arthur Hayes No. Avi Felman You know? Arthur Hayes Okay. Avi Felman I was really hoping for something a little bit more like, "Yeah, I met some guy at the Watergate Hotel." Arthur Hayes Yeah. Avi Felman Arthur, I really appreciate you spending the time with me. Arthur Hayes Thanks for having me. Avi Felman This was a really fun conversation. It went longer than we thought it was going to go, so I appreciate it. This was great. Thanks. Arthur Hayes Awesome. Cheers. Avi Felman All right. Catch you out there. Nothing said on the ThousandX podcast is a recommendation to buy or sell any investments or products. This podcast is for informational purposes only, and the views expressed by anyone on the show are solely their opinions, not financial advice or necessarily the views of 1KX Media. Our hosts, guests, and the 1KX team may hold positions in the companies, funds, or projects discussed.