AI之后的金钱：认识为互联网打造的新数字美元“稳定币” | EP #200

There’s a lot of conversation going on around the world, a lot of chatter about the potential of the US losing its position as the global reserve currency. What can the United States do about it? And that’s where stablecoins come in.

Jeremy Allaire is here. He’s the co-founder and CEO of Circle.

We think that the world needs a full-reserve banking system. We think that the world needs much safer base-layer money, and that’s what stablecoins represent.

Money’s about to go from static to supercharged. You’ve talked to lots of central bankers. Are they even aware of what’s coming?

The really smart people are like, “Oh my God.”

Everybody, welcome to Moonshots. I’m here with my Moonshot mates, Salim Ismail, Emad Mostaque, and Peter Diamandis. Today, we have a special guest, a new friend, someone I hope to have in my life for many decades to come: Jeremy Allaire.

He’s best known as the co-founder, chairman, and CEO of Circle, the company behind the stablecoin USDC. By the numbers, USDC has a market cap of $76 billion, over 90% year-over-year growth, and this past summer it did an IPO and raised $1 billion.

Welcome, Jeremy.

It’s a pleasure. Thank you, Peter. I’m psyched to hang out with you guys.

I like to call what you’ve done at Circle an overnight success after 12 years of hard work.

Yes.

Let me set the goal here for our listeners. The way our banking system works today is kind of insane when you look at it, and I hope to dissect that with you.

At the end of the podcast today, my hope is that our listeners are going to understand why stablecoins on the blockchain are the future of money, the future of payments, and the future of transactions. While stablecoins are safer, they’re also more ethical and operate at internet speeds. Finally, I want to discuss why America needs stablecoins.

Does that sound good to you?

Absolutely. I think we can bridge off of that into a lot of adjacencies as well that hopefully build on that, too.

I think we should begin with a definition of stablecoins and USDC. Do you want to take a shot, or should I?

I’m happy to. I’ve been thinking about this a long time, including the semantics of all this.

There are a lot of things that call themselves stablecoins. What I don’t want to do is try to describe the entire topology of things that are stable in name only, or whatnot. I’m going to use a narrower definition. That narrower definition is now really being enshrined in law all around the world. The major governments of the world are enshrining this definition of stablecoins in law as they codify them as a form of legal electronic money in the global financial system.

In our conception, a stablecoin is a representation of a fiat-denominated currency, such as a dollar, a euro, RMB, or yen, as a cryptocurrency operating on the internet and on public networks—these public internet networks colloquially known as blockchain networks.

They’re notable in that they’re not just denominated in fiat. They’re fully backed and fully reserved by that fiat currency. There’s a lot of nuance in there, which we can get into, but fundamentally, they’re backed one-for-one. When you have one of these stablecoins, they’re stable, meaning that you can always create them or redeem them for the unit—a dollar, a euro, and so forth. You always have that one-for-one creation and redemption.

Therefore, the mechanism by which you get the stability is that full reserve. They’re designed to be very safe, safer than commercial-bank forms of digital currency, but they operate on the public internet and inherit all of the public internet’s superpowers, which we’ve all come to love: openness, interoperability, global reach, programmability, marginal costs of moving things that approach zero, and movement at the speed of the internet.

That is what stablecoins are. The legal term in the United States under recently passed federal law is what is called a payment stablecoin, meaning it’s good for payment. It’s money that’s good to settle a transaction, basically.

I appreciate that. You probably did a better job than I could have.

I’d like to open with this question. There’s a lot of conversation going on around the world, a lot of chatter about the potential of the US losing its position as the global reserve currency, especially as China, India, and Russia are getting together.

Do you think that stablecoins could help play a role in maintaining the US’s dominance as a global reserve currency, stabilizing that position, and backing it in some fashion?

I think that currencies have network effects from their adoption and utility, and they have network effects from their liquidity.

The dollar basically got plugged into the payment systems of the world after World War II and became a required settlement asset for trade. Therefore, it gained enormous liquidity. That liquidity and network utility have been very strong, and that’s been the basis of the dollar for a long time.

But then, after 1971, something profound happened, which I think we all know about. The fiat currency—the dollar—had a global consensus behind it. It was somewhat imposed because the US won World War II. The dollar depegged from gold, which is one way to think about it.

The dollar had been a gold-backed currency, and it was considered incredibly neutral as a result. You had gold there, and the dollar was this sort of unit of account. Governments maintained their own pegs to the dollar, which implicitly meant they were pegged to some amount of gold.

Because of ballooning deficits in the 1960s and the war being waged in Vietnam, Richard Nixon said, “We no longer have the money, so we’re going to break from that. We’re going to create government-issued money by fiat.” It was just: we say it’s the money.

Fiat currency goes way back historically—the concept of fiat—but fiat currency is a relatively recent phenomenon. Most of monetary history was commodity-backed in some way. There are lots of examples of detethering from that, no pun intended.

You have this break, and then the currency basically became about the full faith and credit. That’s the phrase we know: full faith and credit.

And battleships and armed forces.

Yes, full faith and credit, and the sovereign power and sovereign creditworthiness. For most countries, it’s sovereign creditworthiness. Whether you buy an Argentinian bond or a South African bond, it doesn’t have to do with their military. It has to do with the creditworthiness of the government.

There have been these shifts globally, as you referenced, in geoeconomic powers and, arguably, rising hard power as well. We’ve seen that hard power exercised, like with the Russian invasion of Ukraine. We see hard power exercised in other places.

You’ve had this question for some time, which has been on people’s minds, and it became particularly exacerbated after Russia invaded Ukraine. The United States government basically went to all the utilities that are part of the dollar network.

These are software utilities. SWIFT is a software utility. It’s a messaging standard that broadcasts messages around the world, and who has access to that messaging protocol is who has access to settle money.

The US government didn’t just intervene in that. It also grabbed things from people—basically putting a lock on database records. If someone holds Treasury bills or holds something, your read access has been blocked. Your read access was blocked for Russia to a whole bunch of stuff, and that freaked people out.

It really freaked people out. It was like, wait a minute: this dependable, consensus-based international system, where full faith and credit is okay—there are some question marks about that. At the same time, we've been in a world where there's—you talk about exponential debt; we could talk about exponential debt. We've had exponentiating debt.

Exponentiating debt is a real issue, and so the Bitcoin adherents—and I happen to be a Bitcoin adherent as well—are looking at sovereign debt as a real issue. The full faith and credit part becomes challenging. All of this is combined into asking this bigger question: Is the role of the dollar waning? If you read Ray Dalio's 500-year history, he'll tell you it is, et cetera. You can look at this from some of these angles: debt supercycles, the geo-economic and geopolitical environment, et cetera.

All of this comes back to your question: What can the United States do about it? There's the obvious stuff, which is to become more neutral or less intentional in terms of control. That's a policy choice. It can have less debt, and that does not seem to be a politically viable choice in the United States right now.

As you guys talk about all the time, we're going to have insane levels of GDP output from AI, and we're just going to grow our way out of it. That's one philosophy: We're going to grow our way out of it, and therefore don't be worried about the debt. Keep buying US Treasuries because they're going to be good in 10 years, because we're going to be so productive in 10 years.

But then there's this other question: Wait a minute, there isn't an alternative. There's a lot of noise about alternatives—noise about BRICS and BRICS currencies—but trade settlement in dollars is still 60-some percent, maybe even higher, as high as 80%. The question is: Is there something the United States can do that can support demand for US Treasuries and continue to strengthen the network effects that it already has?

That's where the internet comes in. That's where stablecoins come in. The United States famously liberalized and commercialized the internet and exported software utilities, made principally by American companies, everywhere in the world on the public internet.

I think the question now is similar: With AI foundation models, or with internet financial infrastructure like stablecoin financial infrastructure and blockchain infrastructure, can the United States create a liberal, free-market, competitive regulatory regime and export all of this globally, cementing its role not just in the AI arms race but in this financial utility arms race? That's effectively what the policy position of the US government is right now.

That is what the GENIUS Act represents. The fundamental argument is, let's just make dollars have higher utility by allowing them to compete on the basis of free-circulating digital currency on the internet. Let's let technology-driven companies be the primary IP generators and innovators, continuing to evolve these open software stacks and lead this. That would essentially continue to protect and bolster the US dollar.

That's a long-winded answer, but I think it's important framing. Salim, do you want to jump in? I have a specific question for you. I'm seeing a factor here that could be very exciting: Is USDC exclusively backed by T-bills, or are there other assets? How do you back it today?

Yeah. USDC is, I would argue, one of the most transparent financial institutions that's ever been created, and Circle itself is one of the most transparent financial institutions that's ever been created. If you go to—I don't know what bank you use; you don't need to tell me—but let's say you bank with Chase or Wells Fargo, or I don't know who you bank with. If you went and said, “I want a real-time view or a daily view into every instrument that backs the fractional-reserve loans that relate to the dollar obligations that you have to meet,” you're not going to get that.

In fact, if the Fed and Treasury said to the bank, “We want to see all that,” they couldn't put that together for you. The books and records are hard, and auditors are just sampling data to get the audits done.

What we've done is construct a model where essentially 90% of what backs USDC at any given time—sometimes it's 85%, sometimes it's 93%, but let's call it approximately 90%—is in a structure called the Circle Reserve Fund, which we created in collaboration with BlackRock. Effectively, it creates a publicly listed structure called USDXX. You can do a little Google search, and it'll bring you to a page on BlackRock where you can look inside that daily and see every single instrument that's there.

What you'll find in that 90% is that it's primarily short-duration US government Treasury bonds. They're all 90 days or less, but the average duration—you could probably tell me, because it tells you in real time on there; I haven't looked at it—but sometimes it's 10 days, 13 days, 14 days. They're very, very short-duration averages.

It's Treasury over-collateralized repo. For those who aren't familiar, giant banks take cash from us—they borrow cash from us overnight. So we give them our cash, and then they give us more than that amount of cash in T-bills. So we have over-collateralized overnight obligations from banks. If the bank can't pay us back, we have the T-bills; we're good, and it's over-collateralized. It's fundamentally short-term government-obligation risk. And then there's cash, and you can see where the cash is. All the cash in that fund is with Bank of New York Mellon. Bank of New York Mellon is the biggest custodian of cash. They have this incredible infrastructure. They custody $44 trillion of assets.

They're known as the bankers' bank.

So we bank with the bankers' bank. The other 10%—about 98% of that 10%—is held with bankers' banks, meaning the BNY Mellons and State Streets of the world, companies like that. We also have a little bit positioned around the world in high-quality banks, but also a little bit with fintech banks, to provide immediate liquidity. So if I'm in Singapore or Brazil and I want to create USDC from a local bank, I can do that basically 24/7.

That's what's actually there. The short answer is that it's T-bills, what's called Treasury repo, and cash. Despite people saying it's unregulated and all that sort of stuff, obviously there's the GENIUS Act, but we're also under a framework imposed by the most intense regulator on Wall Street, which is the New York Department of Financial Services. It has a set of specified assets that we're only allowed to hold, so we have to hold ourselves to the Wall Street regulator.

I've been watching Circle since its founding, and I have to hand it to you.

I've invested as a founder. The sheer perseverance you've had to navigate all of this regulatory bureaucracy is Nobel Prize-winning stuff.

Well, when you put it that way, that was what made this exciting to me.

Jeremy, what did you put the chances at on founding day that you'd get to today, honestly? Would you say 100%?

I don't think probabilistically, because you can't do that as a founder. It's all or nothing. You have this conviction, you have a vision, you see what you want to do, and you know it's possible. Then you just fight like hell to realize it. You create the future. That's the mission.

But you built the tech. You built the tech first and didn't change the policy first. I think that is something critically important.

Well, a little bit of both. A little bit of both, right? I was just doing a session at the annual meeting of the World Bank and the IMF today.

I'm in DC. Twelve years ago, almost to the day—it was November—I was asked, I was called to testify to the U.S. Senate about this issue: Can virtual currencies, as they were called back then, exist? What's the innovation potential? And so on. I was advocating hard for policy at that point.

While we built the tech, before I put a single dollar in the company, before my investors put a single dollar in the company, I hired the very best regulatory policy advisers in the world to basically help me figure out how we were going to do this. What was the thread we could pull to do this in a legal, compliant way? And we figured that out.

What that meant, though, is that while maybe there weren't stablecoin regulations, because stablecoins didn't really exist, there was a body of law around payment systems, electronic money, and all these things. To do what we needed to do, we had to be constantly working with policymakers, working with regulators, and collaborating with them at every moment. I think that posture has never changed, and it continues to this day.

But you made it otherwise, right? Because otherwise, it's so easy for them to go, “Wait a minute, he's trying to act like a central bank. Enough of this,” and slam the door on you. Emad, do you want to jump in?

Yeah, no, I think it's amazing what you've done over the years—the perseverance, especially through the dark times. Now it's really interesting because the regulatory tailwind is there. But then also, you've got AI agents literally hitting right now, who are possibly the biggest users of AI and stablecoins ever.

You guys have been doing great work with x402 and SDKs and things like that. I'm really curious: how much of your volume do you reckon in 5 years is going to be AI versus humans?

Yeah. Yeah, I think about this a lot. I'll just give a little bit of historical context. The really inspiring thing when Sean and I were co-founding the company back in 2013—Sean Neville is a co-founder—was this idea of programmable money. It was just an idea on napkins, so to speak, back then. There were some papers that some really incredible people had put out about smart contracts.

I got very excited that these blockchain networks could become distributed compute engines. I had worked on virtual machines and programming languages; I'd worked on application infrastructure and built some of the most popular app servers in the early days of the internet. I was very excited about this idea of these trust machines—these distributed compute engines that could provide cryptographically verifiable compute outputs and inputs, data, and transactions. That was the bomb.

It was like, “Oh, man, if we have that, we can actually move economic activity in a broader-based sense—not just moving value from point A to point B, but economic activity more broadly. We could actually move that onto the internet, and we could reconceptualize all the building blocks of the economic system.” That's what motivated us.

The first thing we thought was, “Okay, well, we've got to get a protocol for dollars on the internet going, because we've got to get that going. We need these networks to become compute engines where you could actually build a protocol layer that's like an application utility layer.” We got that 5 years in with Ethereum.

We got this going, we got the legal going, got the technology going, and wired it up to the financial system. But those were the super-primitive first ideas that we had. Those are coming around now. To answer your question, I think the vast majority of stablecoin transactions are going to be AI-intermediated in 5 years.

I think we're entering a period where economic operating systems are being created. These layer-1 blockchain networks are becoming economic operating systems. They're designed to contain economic activity, provable data, transactions, and compute that are necessary for entities that face each other in a trustless way.

That could be AI agents that face each other and need proving systems, basically, to prove identity, yes, but also to prove inputs and outputs and other things. They need to be globally interoperable. If there's an AI agent spun up somewhere in Asia and an AI agent spun up somewhere else—I mean, “where” is all relative anyway—these agents are increasingly agents with capital, conducting work.

Maybe they're bringing in humans. Maybe humans are bringing them in. It's both ways. The intermediation has to happen on an infrastructure like this. There's no other infrastructure. It's not happening over credit card networks, right? It's happening on this kind of infrastructure because it needs to be highly scalable, from microtransactions—which is why we just released a toolkit for microtransactions with x402.

By the way, I love microtransactions. Talk about the number of times I've wanted to just donate small amounts.

Yeah. Yeah, I mean, yeah, but it's scalable. I use the metaphor that SMTP does not care about the payload, right? If I send you a picture in an email, with an attachment of a picture of my breakfast, and you send me an email back with a copy of a CIA dossier, the payload is clearly different, but they're both 100 megabytes or whatever it is, right? So the payload is the same.

It's the same thing with stablecoin money. I can settle a 5-cent transaction on a high-performance layer-1 blockchain in a fraction of a second. That 5-cent transaction might be very useful for paying for a few AI tokens or whatever it is. On the other hand, I can settle a $1 billion transaction—like, I just bought whatever gazillion barrels of oil—and I'm settling that transaction in a safe way on the internet. So, very scalable models.

I'm a big believer in this convergence between AI agents. This is the missing financial layer that the internet never had.

Exactly. Data, images, and video were on the internet, but it was missing the whole financial layer. We sort of patched together these maybe-trustable credit-card capabilities. Is my credit card safe? But this is—I mean, I don't know. It feels like we're about to see hyper-exponential growth across the board.

Emad, you were going to ask a second part of your question, I think.

No, I was just thinking: you actually launched Circle before Ethereum, and all of this came together even then. Now it's kind of moved on, and it feels like we're going beyond smart contracts because money will actually get intelligent. Smart contracts can be a bit dumb sometimes.

Yeah, yeah. Well, there's off-chain and on-chain, right? Off-chain compute, which arguably will be a lot of AI, still needs proving grounds, right? You need these proving grounds for the off-chain compute. That's where I think the blockchain networks, as economic coordination layers, can be very, very powerful, and that's certainly how we think about it.

When we founded the company, the concept of the Internet of Things, AI, and blockchains was talked about, but AI was nowhere. It was moving along, just like blockchain has been moving along, and people were like, “Isn't that just for speculative Bitcoin and shitcoins or whatever?” Yeah, we had that.

I'm a big believer, and I know you are too, Peter, in the idea that when you have multiple exponential technologies that compound each other—

Convergence.

Yeah, the converging forces. I think that's one of the most important things that a tech entrepreneur can do: see this point.

Business models have come out of that. Yes, right.

100%.

So, Jeremy, I've heard you talk about 2 scenarios that show how absolutely inane and antiquated our financial system is. I would love you to do it briefly for everybody—all of our listeners—to really get this.

When you look at how a brokerage works—how you invest in stocks through a brokerage—compare that today with what it could be, and then how banks actually work when you deposit $1 million into a bank. What happens? Can you cover those 2 and talk about how, when you stop and realize, “Really, that's the way it works? That's insane”?

I think that a lot of people don't understand. They do viscerally, because they get nervous when they give their money to anyone. But we've heard about bank runs. What is a bank run? A bank run happens because the business model of a bank is to take a dollar—to borrow a dollar from you. They're borrowing a dollar from you, and then they're permitted to lend it out 12 times. That's insane. That's what fractional-reserve banking is.

I've been an adherent since the founding of this company, and I continue to be an adherent today, of the idea that you can have a separation of money and credit, and that you can have full-reserve money, where if I give you a dollar, I'm not allowed to do anything with it.

It's a dollar. And I can build a super-hyper-efficient payment system on that extremely low-risk form of money. Then, if you want to borrow money, you've got to borrow on that full-reserve money.

Can I drill down on that just for a second? This, I think, is one of the most admirable aspects of Circle: each dollar is fully collateralized, right?

Yeah.

That's not the case with, say, Tether or some of the other coins. What made you—how did you make that choice? Because that's a very principled choice that costs a lot. Something had you make that kind of principled choice and go, "We're sticking by that." I'd love to hear the history of that.

Yeah. There's a short, easy answer, and then there's a slightly longer, fuller answer that expresses the real thing. The short, easy answer is we're required to do it by law. When we built USDC, we're regulated under what are called money transmission statutes. Money transmission statutes are federal law, which guides how electronic money transmission works.

Then each state in the United States has a license that you get for being what is called a money service business. It's essentially a bank that takes money and moves money but can't lend money, so it is a narrow focus. Historically, that was Western Union, and then that later became PayPal and Apple Pay and all these things. Lots of fintechs are actually money transmitters.

We were regulated under money transmission law. Let's say we go to the state of—I don't know. Where do you live, Salim? New York, New Jersey. Wherever. You're in a place. Okay, you're in a place. You're in a state. The state has a banking regulator, and we had to go state by state and say, "Okay, we'd like a license to operate in your state." They say, "Okay, here it is. You've got to post some bonds that are like collateral."

Then you have to back your electronic money instrument with a permissible set of investments called the permissible investment clauses, which are super narrow. They're designed to be hyper-liquid assets, so that instrument is as safe as possible. That's the short legal answer.

But the other answer is more philosophical, which relates to the discussion we were just having. One of the motivations for starting Circle was that I got very interested in the nature of money, the nature of banking, and the nature of the monetary system after the Great Financial Crisis. I studied it. There was stuff I already knew, like if you'd ever watched It's a Wonderful Life and thought, "Wow, these bank runs," but the GFC was a totally different scale.

It was all this opacity in all these leveraged-up instruments. I looked at that, and there were ideas bouncing around, as there have been from time to time: Is there a way to have full-reserve payment-system money separated from lending money? That idea came up again after the Great Financial Crisis, and it stuck with me.

It really occurred to me that a more sound-money basis for the banking system makes a lot of sense. When I started Circle with Sean, I thought, "Well, actually, we could construct that on the internet." In fact, it's totally necessary on the internet. You don't want these IOUs floating around, free-floating, moving around on the internet. That's a recipe for total disaster.

It's almost like, "Okay, well, we have to design this way." As we move from the early regulations, like the money transmission laws, into stablecoin laws that deal with this at a deeper level—as a defined form of money in the financial system—that kind of reserve model became paramount. It went from an operating reality for us into federal law as well.

I need a list of all the people you must have threatened along the way, right? Competing stablecoin folks, the big tech and financial payment systems like PayPal, the banks, and especially some of the smaller banks, the regulators and monetary authorities, the Fed, the payment intermediaries like Visa and others, firms giving credit and lending, incumbents in the fiat banking system. I mean, you had to navigate all of that.

I'm still navigating it.

Doesn't the Federal Reserve kind of go, "Wait a minute, he's acting like a central bank. Shut him down"? Don't you get that response? And how have you navigated that?

So, a couple of things on that. I've had a chance to spend a lot of time with the leaders of many of the biggest central banks in the world just in the past couple of weeks. I shared the stage with the head of the World Bank and the IMF. They're not central bankers, but they look over big, big, big things.

I think one thing that's really important is that a stablecoin—a regulated stablecoin—inherits the monetary policy of the central bank. It does not replace it. We don't set interest rates on money. We don't establish the price of money, nor do we create money. A central bank can create money.

As I like to say, they have the SQL INSERT statement capability in their database. You can do it once. Well, yes, but not even once. I can copy once, basically, right? We do not create money, nor do we set the price of money.

Nor do we replace the central bank's core settlement ledger for that fiat currency at an absolute level. We augment that with an internet infrastructure that makes that particular currency super useful, more useful than it has been before. We inherit the monetary policy.

If the price of money is increasing or the price of money is decreasing, the behavior of markets adjusts accordingly, but we don't have any role in that. We are what I call a kind of macrocyclical, macro-sensitive business, like banks in that sense. But we're not involved. I don't go to the rate-setting meetings.

I think probably some listeners are wondering, why don't the central banks issue their own digital currencies? The Trump administration said that's a bad idea. Why? And the other part is, what happens with the GENIUS Act when JPMorgan and others say, "Hey, we can do a stablecoin, too"? How do you see all that playing out?

Yeah. It's a great question. If you remember back in 2018 or 2019, there was a proposal for something that was colloquially referred to as "Zuck Bucks." We launched USDC in 2018. We published our white paper in 2017, and that was kind of conceptually introduced in 2019.

We were a little guy, right? There was 500 million USDC in circulation. But Facebook was like, "Oh my God, they have 3 billion users." All the governments in the world were freaking out, and there was probably good reason.

The other challenge was that Zuck bucks were actually synthetic money that was going to be based on a whole bunch of currencies. They thought, "Oh, the world's ready for that," and it turned out a lot of governments were very uncomfortable with the idea.

That was a wake-up call. The result of that wake-up call was that certain governments, notably the Chinese, said, "We cannot allow private money to come and sweep the world. We're going to respond by building a central bank digital currency." Research into central bank digital currencies skyrocketed. The Atlantic Council, I think, talked about 122 governments researching central bank digital currencies.

China had actually begun work on this, and I met the founders of it back in 2014. I met them in Beijing. Then that project, which was a sort of laboratory project, was, "Okay, we've got to ship." When the Chinese decided to ship, they shipped. That launched what is now known as e-CNY.

It's a really interesting lesson because the Chinese government is a very powerful government. I think we all think about the Chinese government as a centrally run government that can just tell people what to do and they'll do it. Well, in fact, that's not the case.

That product launched in 2020, over 2 years ago, and no one used it. No one wants to use it. They made every payment app, every bank, everyone have e-CNY. Why didn't people use it?

It's because, in my view, people want the most innovative product. People want to use the thing that has the most utility for them. Alipay and WeChat Pay just have more utility.

This is private-sector innovation. I can walk into a store and go up to the counter, and they can say, "Do you want to buy this?" I say, "Yeah, I want to buy this." It voice-authenticates me, and I walk out. That's a feature of Alipay and WeChat Pay.

Is that a feature that the central government is going to build? Private-sector innovation, technology, and software innovation are fundamentally entrepreneurship. You may have government funding, but it's fundamentally entrepreneurship. That's why we're sitting here on Moonshots, right? Entrepreneurship is so key.

I think even in a place where the government can say, “You have to have this. You have to support this,” people didn’t want to use it. And so, what’s actually happened is this has become somewhat of a philosophical debate. It’s a debate between the government being in your pocketbook directly, having direct access to your money and your behavior and the choices you make—like I’m piped right into the government’s database—or whether there’s an air gap between me and the government.

Today, 97% or so of the financial transactions we have are intermediated by private third-party intermediaries. That is the way it works today. If you go back over, say, even 75 years in the West, in the liberal market order, every major technology innovation in money has come from the private sector. That was checks and check clearing, ATMs, credit cards, debit cards, PayPal, Apple Pay, stablecoins—all this innovation comes from many times, consortiums of actors getting together and saying, “Hey, we’re going to build this new technical utility, we’re going to agree on it, and people are going to use it.”

I think in the United States, there’s hostility—outright hostility—to this concept that the government is going to build all this for us. I think the Trump administration essentially banned it, so there’s nothing going on here. In Europe, the central bank is pushing forward with a digital euro project, but the commercial banks are very resistant to it because they’re saying, “Hey, you’re going to disintermediate me.” At the same time, the Europeans have passed stablecoin laws, and now the banks and others, including Circle, are doing euro stablecoins.

The estimated launch date for a CBDC in Europe is 2029. I don’t know if that’s going to slip or what, but if you’re saying it’s 2029, by the time you get to 2029, it’s putting your card on the website. I think this sort of bet on open-source, open-internet software innovation, entrepreneurship, competitive markets, and a fair playing field is a winning bet. I think the U.S. has the winning bet right now.

But it does come to the other part of your question, Emad, which is, okay, well, JPMorgan and other guys are going to come in. I think a couple of things: for a commercial bank, creating a new stablecoin under the GENIUS Act is actually banned. Why is that? Because you wouldn’t want a stablecoin backed by risky deposits and risky loans.

A bank holding company—a company that owns a bank—can also create a stablecoin subsidiary and issue a stablecoin. The guys running that are going to say, “Okay, I have this money. I can get it over here, where I can lend it out 12 times and capture more margin, or I have this thing I can’t do anything with, and its ultimate promise is the commoditization of payments, where the payments are priced at zero.” Well, that was totally not what you were saying, right?

There’s a structural thing there, but it still may be that banks actually want the payment-system utility benefits. They want the programmability of money. They want to offer their corporate and institutional clients this innovation. We actually see huge opportunities to partner with banks, and we are. You’ll see more and more of that.

We’re partnering with banks where they’re saying, “Well, we can just convert in and out of the stablecoin for its utility on the internet, but then preserve our deposits and lending.” You’re seeing some hybridization happen. Nonetheless, Emad, there’s going to be a lot more competition in this space.

My best reference is that stablecoins are internet platform utilities and networks. They have developer-driven flywheels and network effects, and they have liquidity-driven flywheels and network effects. Those are significant competitive moats. We have those moats, but we’re not sitting still because there are so many people who see this as a huge, multitrillion-dollar opportunity.

So, Jeremy, this has been an incredible moonshot. You’ve effectively achieved your first goals from a decade ago. Looking 10 years out and speaking to the entrepreneurs who are subscribers on Moonshots, what should they be thinking about? Where are the massive untapped opportunities that you’re excited about seeing?

Yeah. It touches on something we were talking about earlier. First of all, I think we’re in the super-early stages of all this. This is super early stage.

I like to think about it through a couple of lenses. The total addressable market of the global financial system is hundreds of trillions of dollars. It’s this enormous thing, and the revenue streams of all the intermediaries are $10 trillion. It’s a huge thing.

Just like earlier epochs of the internet, where open-internet infrastructure and open software collided with industries, restructured the product utility, restructured the unit economics, built a better way to approach it, and unlocked completely new things that weren’t possible before, those kinds of things happen over 10 or 20 years. They take longer than people think. We might be in an acceleration environment because of AI. That may come in, but these things still take these long arcs.

Even after these long arcs—even after 10 or 20 years—the platform businesses that get built are trillion-dollar companies. But they still represent a small percentage of the equivalent prior regime. The viewing hours of broadcast, terrestrial, cable, and satellite are still far in excess of people who watch streaming.

You were in the digital media world, right? There are so many great analogies between old media, today’s media world, and stablecoins.

Yeah. I like to say “over-the-top internet money.” It’s like over-the-top video.

You were going down—you and I have talked about the idea of the future of the corporation.

Yes.

I think this is very cool for entrepreneurs to be thinking about. I will be talking more about that in the weeks ahead, but do you mind giving a vision statement there?

Yeah. So, I wanted to share that. My first point is: it’s early days. For entrepreneurs and people who are thinking about what they can do with this technology, it’s super early.

I think the second point is that we’re at the front end of a new operating-system paradigm. We go through these platform shifts every so often. The web was a platform shift away from personal computers; it was a platform shift in terms of information. We’ve had platform shifts in terms of communications utilities, social media, social messaging, going from desktop to mobile, and from data centers to cloud—blah, blah, blah. We have these platform shifts.

I think there are 2 operating-system platform shifts happening right now. One is AI foundation models, which are effectively new operating systems. They are compute engines: you write apps that they run, and they execute tasks. They’re far more capable operating systems than we’ve ever had before.

The other is these economic operating systems, which are purpose-built blockchain networks. These economic operating systems are another platform shift. With that, just like with other platforms, you get a new set of materials to work with.

With the iPhone, you had GPS, a touchscreen, a camera, and other capabilities. You had new facilities to work with. You’re getting these new facilities to work with these economic operating systems, these blockchain networks. You’re getting the ability to have provable, executable contracts that can be deployed and interacted with on the internet.

You’re getting the ability to have provable, tamper-resistant data structures that can be interacted with. It creates a substrate where you could, for example, take what is conceptually a corporation and implement that entirely in software as a mixture of humans and agents.

You could form the capital by selling tokens. The token capital can be stored in a treasury that’s on-chain, and that treasury could be mostly stablecoins. You can automate all the monetary flows. You can have governance, which is stakeholder participation: who decides what to do with the treasury, who decides what contracts to enter into or manifest. That can be the governance layer, with provable voting entirely done on-chain.

The entire audit of the books, records, and everything that happens is real-time auditable. You can use view keys and other things to create selective-disclosure models. You can create on-chain governance and on-chain decision-making. You can have that entity delegate work to AI, and you can have it delegate work to humans.

You can have the actual contract between the AI and the entity manifest in code. You can have the contract between the humans manifest in code. You can build all that. That’s available.

Salim, that’s almost—yes, it’s coming, Salim. It’s like the promise of DAOs without the governance issues.

Well, yeah. DAOs are fascinating because they’re an experimental thing, but I’ve studied roughly 400 years of the development of the joint-stock corporation, which then became modern capital markets, which then became modern banking. The emergence of corporations as organizing entities is tied up in—I think about this a lot—theories of capital.

The Industrial Revolution was all about capital and labor. It was Adam Smith versus Karl Marx, and what is the ultimate relationship of labor to capital? But now that’s being disrupted, because capital can have a relationship to machines and machine labor, not human labor. These are deeper philosophical questions for society, but they’re real things.

In this world that we’re moving into, I think it’s the first time since the development of joint-stock corporations and the governance models that we have that we can revisit all of that. We can actually revisit all of that.

Jeremy, you’re describing the first fully on-chain corporation, right? Contracts, payments, treasury, governance, agents, and even robots are on the blockchain, developing products and services and delivering them.

Yeah. 100%. Yes.

And the velocity of those companies is going to be—

We actually have to have enforcement. That’s the other thing. We still need prisons for the humans who do bad things. We still have people who don’t meet their commercial obligations, and you need to be able to sue them and go after their assets. There are certain things we absolutely need.

We’re not at the point where there’s an AI court, which is probably a good thing right now. But the velocity of those companies is going to be so explosive that I’m not sure any corporation can compete.

Is there a hard takeoff in the future of corporations in that capacity?

Yeah, that’s what I think. I don’t know what the time frame is, but given exponential improvements in AI and this technology substrate, I think we already have the single-founder company phenomenon, which is an AI-driven phenomenon. You also have on-chain organizations.

The best recent example is something called Hyperliquid. Hyperliquid is a software protocol. It’s just a set of smart contracts that exist on the public internet. It’s open source, it’s a protocol, and it has a token.

There are stakeholder economics for the people who created it, the people who govern it, and the people who make choices and improvement decisions on the protocol itself. The protocol generates fee revenue from its use. It’s a specific protocol focused on structuring futures markets.

A perpetual DEX, if I remember right?

Yes, it’s a perpetual derivatives exchange protocol. But it’s now an underlying infrastructure where you can instantiate it for other things. You could create a perpetual for sports betting. You can create a perpetual for anything that you want a perpetual future on. You could have a perpetual futures market on AI compute—whatever you can imagine.

They’ve created this thing that you can instantiate on, and that instantiation derives revenue. That entire protocol and infrastructure has been built, delivered, and operated by 11 people. It’s doing well over $1 billion in revenue right now.

That revenue is being returned to the token holders and stakeholders, which is getting more people to pile in and build on it. So it’s creating compounding network effects for the protocol itself.

I think these kinds of digital, token-based incentive systems, combined with AI workers and human workers and these new systems of governance that are possible on highly globalized networks, will create super-predator corporations, if you will. They’ll be hybrids of humans and AIs, and they’ll make things. They’ll make physical things and services.

They’ll drive the bankruptcy lawyers’ business.

It’s definitely going to keep the lawyers, which will also be AIs, by the way.

I have a quick thought. I know you want to hear Emad’s views on these things, but go ahead, Salim. We’ll go to Emad next.

No, no, go ahead. Let’s hear from Emad. I’ve got a thought.

I don’t know. I just think it’s going to be fantastic. The U.S. hasn’t recovered its monetary velocity since COVID, right? Basically, what you’re describing—and I think this is why you’re building Arc, your own L1, and things like that as well—is that money’s about to go from static to supercharged.

The thing I’m most fascinated about is that I’ve been working on new economic theory and things like that. You’ve talked to lots of central bankers. Are they even ready for that pickup in monetary velocity? Because that impacts inflation. If you don’t have output, it impacts inflation.

Question, by the way, from Emad: are they even aware of what’s coming?

I’m telling them. The really smart people—and I’m not going to name names—the really smart people are like, “Oh, my God.”

You’re exactly right, Emad. Monetary theory, money velocity, and monetary theory are going to get upended. One of the risks, of course, is that people throw up the gates and say no.

I think the fact that, currently, the United States is not throwing up the gates is important. The United States is saying, “No, we’re going to let this grow.” It’s a similar situation with AI. People can throw up the gates, or you can have prudent regulation and what I call agile policymaking, which is an oxymoron.

There are going to be things that are profound. The concern, of course, for people who have studied recent financial crises is that it was the liberalization of derivatives regimes that arguably allowed toxic balance sheets to develop, which then led to the blowups behind the Great Financial Crisis.

There’s an argument to be made that this also had to do with inherent opacity in the system. One of my first principles—and blockchains are really good at this—is radical transparency. Cryptographic proving is another part of the new toolbox.

What’s different is that you have this new material to work with: cryptographic proofs. That allows for real-time auditability and a lot of things that weren’t possible before. Governments are probably going to need technology people to write software to deal with some of this stuff as well.

One of my community members, Jerry Michalski, is famous for saying, “Abundance equals scarcity minus trust.” Something you’ve done profoundly well is create structures for scaling trust. This is going to help us get to where we want to go, so I think that’s really powerful.

Just a comment back to the core future of the corporation: when I’m advising CEOs—and Peter and I are both doing a lot of this right now—what we’re saying is basically, create a digital twin, with everything being AI-native and digital-native, with smart contracts running most of your operations, and put the entire company on-chain, driven by AI.

Essentially, that’s going to replace the mothership with all these people at headquarters: bean counters, policymakers, and product-strategy people. If you’re not doing that, somebody else is doing that to you, and you have a choice.

And it’s a really simple discussion.

I think that’s right. I would say 100%. Now I’m a New York Stock Exchange-listed public company, and so I have a lot of embedded infrastructure obligations that are there for a good reason. I’m not that agile startup; I can’t just say, “Poof, I’m on-chain.” So I think the road for larger companies that have these larger systems of regulation on them is a different road.

But I agree with the basic advice. There are still open legal questions on a lot of this stuff, and the tooling isn’t quite there yet on some of it. To do everything—to put all your operations in smart contracts—it’s still hard. But if I look at the progression of things, it’s going to become easier and easier in the coming years.

Now people are starting to conceptualize this, and there are going to be people really going after it. There’s a tools market to develop here as well.

Huge, Emad.

Yeah, I think so. I saw you launching your own L1, Arc, and there were 2 really interesting things. I’m picking up on Salim’s point about transparency and trust. One is that, on the one hand, it’s going to be super fast, like subsecond finality. On the other side, there’s refundability.

I just got some stablecoins today, and I was like, “Where are they?” And they’re like, “Oh, we sent USDT instead of USDC.” And I was like, “USDC all the time, obviously.” So I just wondered: how is that balanced? How did you come up with the requirement for finality, but then the ability to reverse transactions as well? What was the thinking behind that?

So, actually, there’s a little bit of misreporting that you probably picked up on, which is that fundamentally, it’s sort of deterministic settlement finality as fast as possible, as cheap as possible. That’s a fundamental thing, and it’s core. It’s important not just for moving a dollar; it’s important for who owns the house, who owns the stock, or whether the contract completed its execution. A provable final state is essential, and so that’s there.

Now, you have to think about, okay, now I have other things. I have an issued asset that’s a protocol, USDC. I know that Circle is a regulated, centralized issuer, and it sort of works on the public internet. But because Circle is a regulated, centralized issuer, it has to follow sanctions law. So if an entity is sanctioned, we actually have to freeze that account. We do that on 28 blockchains, not just Arc, but on 28 blockchains. That is a hard legal requirement.

But then you say, okay, well, I want to start using this in retail transactions. I want to buy a cup of coffee with this, or more realistically, I want to buy a product on the internet that someone is going to send to me. All of a sudden, there are other things you need. You need metadata and other data alongside the cash. That data could be the invoice or the receipt; it could be other relevant metadata.

For example, we’re adding a native protocol called Receivo, is what we call it. It’s essentially an ISO 20022 data-messaging protocol that’s useful for metadata alongside the actual bearer asset—the monetary asset. Then what we’ve also done—and this is just in R&D right now; we haven’t published it—is we’ve created something called the refund protocol.

The idea is not that the blockchain itself is reversible. The blockchain is not reversible. It’s not that USDC is reversible. USDC transactions are final unless they’re sanctioned. So that’s just a thing. Above that, you can create a payment protocol that supports a refund.

That’s experimental right now, but the concept is that if I buy the product and it’s fraudulent, or if I buy the product and I don’t like it—it’s the wrong product, it’s this shitty thing I bought and it’s broken, it came with whatever it was—you need a way to return value on the basis of whether it’s fraud or just a customer choice.

Refund protocols are just a proposed protocol; they’re not actually built into the blockchain. It’s a higher-level protocol. We’ve proposed an insurance pool model so that people can stand in on the transactions and effectively buy the insurance and sell the risk. You effectively have a market for the risk of refunds that anyone can participate in. So you create a global, decentralized market for risk.

Makes sense. Makes sense.

That’s actually what we are proposing. We are not proposing a blockchain that reverses transactions. We are not proposing that USDC reverses transactions. We’re working on these higher-level things that I think are important at an application layer for ultimate commerce, right?

Jerry, take it home for me one second. When am I buying things on Amazon or at Starbucks with USDC? Where is it right now? What is the average user listening today using USDC for, and what will they be using it for a year from now?

Yeah, it’s interesting. USDC is widely used around the world. It’s used by people and businesses in hundreds of countries. Obviously, it got its start as digital cash for investing and trading and as working capital with the digital-asset markets. That was what I called the bootstrap utility. 24/7/365 markets need 24/7/365 money, and naturally that emerged.

That morphed into use by innovators building on-chain protocols for financial products: borrowing and lending protocols, things like these derivatives protocols. And so USDC has become capital in borrowing and lending. But what’s really happened, in particular over the last couple of years, is we’ve seen strong growth in its use in cross-border transactions and international transactions.

There are so many companies now building products for payroll, payouts, B2B payments, and more, adding the stablecoin rail because it’s just a more efficient medium to settle transactions into markets around the world. That relates to another major use case, which is as a store of value.

So this gets back to your dollar question: people want dollars. People want to hold dollars, and in enormous parts of the world, they’d rather hold a digital dollar that has internet utility than hold a local bank dollar or a local bank currency. This is sort of an over-the-top phenomenon again. It’s like when people figured out, “Oh, I can use WhatsApp instead of paying SMS fees with my local carrier.” So there’s this over-the-top phenomenon, and it’s proliferating.

You’re getting users, and that’s cascading into small businesses that are figuring it out because they don’t have big compliance departments. You’re seeing this happen around the world. By virtue of that, you’re not walking into Starbucks. There are plenty of what I call USDC debit cards. Visa and Mastercard have wired up to USDC.

If I’m an issuer and the money I have is actually USDC, you can use that USDC to pay at a merchant. But it’s the same payment rails. It’s the Visa rail; it’s Apple Pay. The actual settlement between the person who gave you that card and Visa is done using USDC, which is kind of cool.

But my view is that the retail commerce—I mean, Shopify just rolled out USDC for all their sellers, and it’s just there. It just shows up, and they’re incentivizing merchants. They’ll pay the merchant 50 basis points to accept a payment in USDC. That’s coming online this quarter. I don’t know what that’s going to do.

Stripe has rolled out USDC as a payment method that’s available out of the box to any merchant. But usage in e-commerce is still very, very small. I think we still have work to do on the user experience, work to do getting more and more mainstream wallets to have this integrated in a simple, seamless way.

We need things like the refund protocol; we need other pieces there so that some of the user expectations can be met. That’s why I think it’s still a couple of years away from widespread use.

You haven’t had your Pizza Day yet—the equivalent.

I mean, there are lots of transactions happening. There are lots of every kind of transaction happening: people buying digital goods, people buying other things. And actually, there are huge transactions happening, too.

The biggest electronic trading firms in the world use USDC to settle multihundred-million-dollar transactions with some frequency. So it has all that. But I think there are areas from a UX and ultimate distribution perspective where there’s still work to do.

So let’s move it up from the consumer to the CFO. We have a lot of corporate CFOs and CEOs who are on this podcast. 5 years from now, you’re running a multinational organization. Is USDC sort of the internet rails for all your transactions through all of your stores around the world?

Yeah, we believe that on-chain treasury management is going to become a huge thing.

There are also lots of companies, startups, and fintechs that are already in this space, baking USDC into what they do. It’s actually a faster-growing startup area, and established companies like Brex just launched USDC as a feature for people. You’re seeing spin-outs from the big ERP systems, like SAP, building on-chain treasury solutions.

I think the ability to move money from a tokenized money market into stablecoin cash to settle transactions programmatically and instantly anywhere in the world, across geographies, is very powerful. It’s time value of money, capital efficiency, and auditability. There are a lot of really powerful things there.

I think the regulatory clarity that’s coming online has sort of been one of the missing pieces. Until you have USDC treated as cash or a cash-equivalent instrument on a balance sheet, where an auditor can say what it is, and you have the regulatory backing that this is actually legal electronic money from a payment-system perspective, you don’t have the full picture. You also need the tooling and enterprise infrastructure, and that’s all kind of coming online right now.

If I look a year out, then 2 years out, and certainly 5 years out, that curve around what’s happening with on-chain treasury is one of the most exciting areas of this space.

Salim, what are you thinking? How do you compare with Tether?

I think Tether has actually built an extraordinary business. They’re larger than us—good bit larger than us. I think they really had a lot of their core strength come from the critical role they played in offshore crypto markets.

It was actually kind of birthed out of one of the offshore Asian exchanges, Bitfinex. Then, obviously, other big Asian exchanges adopted it because they didn’t have dollar banking. A lot of them were Chinese firms, and they needed to create a way to have a form of dollar banking. They got bootstrapped in that, and they’re still very strong there.

USDC has grown a lot in these digital-asset markets on a percentage basis, but they’re still super strong there. Their growth is strongly supported inside that space.

I would say that philosophically, our outlook may not be entirely different, but we’ve always been rooted in this idea that we’re building a new internet financial system. We need to do this in the regulated realm. We need to enshrine this in law.

We’ve been U.S.-first and regulatory-first, and we’ve built a very strong compliance apparatus. That’s allowed us to work with some of the biggest banks in the world, some of the biggest asset managers in the world, and some of the biggest governments in the world. Those are things that we’ve done.

I think our view is that when we talked about that TAM earlier—this multi-hundred-trillion-dollar TAM that’s out there—and the TAM in the revenue streams of what runs through all of that, it’s still on the come. It’s totally there. No one’s really gone into that.

I’m comfortable and confident that continuing to do what we do, the way we do it, is going to lead to continued growth. That’s what I believe. I also think that this is a market that will support many significant platforms.

You’ve definitely taken the road less traveled on that one. I’m going to guess that the GENIUS Act is generally a great tailwind for you. Am I right? My question is, what did you do to celebrate when the GENIUS Act got passed?

It’s pretty amazing. I had been working on getting a federal law for stablecoins for more than 4 years—5 years, explicitly, where there was actually a bill being discussed, and conceptually for much longer. It was a pretty extraordinary moment.

My kids were in summer camp, so my wife and I had fun. I was in D.C. and went to the signing ceremony, and that was pretty powerful.

To the question, it is a tailwind because it starts to build certainty around what this form of money is and how it’s treated. If I’m a mainstream corporation or another financial institution, I now have the confidence that I can start to use this and build on this. That’s huge.

It’s also recognized around the world. Right now, my team are the explainers-in-chief of the GENIUS Act to governments all around the world. They’re asking, “Okay, what is this? What does this mean? Should we be doing this?”

If you’re a GENIUS Act-compliant stablecoin, how do we treat you in our financial system? They all support dollars in their financial systems, so how do they treat this in their financial systems? It’s really causing that kind of conversation, and those are good backdrops.

But as Emad noted earlier, it’s invited so much more interest and so much more competition. I wouldn’t be surprised to see a new stablecoin every week that’s introduced and reportedly GENIUS Act-compliant, even though the GENIUS Act isn’t yet in statutory effect and won’t be for a while.

I’m going to create a competitor to you called Square. No, I’m just kidding. We’ve been taken, Salim, already. Did you hear Circle and Square merged and formed Hexagon? That was an April Fools’ joke, wasn’t it? Actually, yeah. Okay.

So, Emad and Jeremy, it’s 10 years from now. We’ve got digital superintelligence way beyond our current conception, and there are a trillion agents running around. These agents have buying authority and access to stablecoins. Is that a viable option a decade from now? How is that going to be even conceivable in today’s regulatory structures? What kind of hyper-exponential effect is that going to cause, and how exciting is that future? Emad, you first.

I think that’s a super-exciting future. With stablecoins and the way you’ve done it, Circle, it’s always, “Why would you want dumb money if you can have smart money?” The agents will want smart money, and they’ll direct it to where it’s most economically valuable.

The goal of regulation shouldn’t be to stifle innovation. Earlier, Jeremy was talking about the Chinese and DCEP, which turned into e-CNY. I was a hedge-fund manager back then, and Tencent and Alibaba had Alipay and WeChat Pay. They enforced reserve regulations on them because the money was going around. This was before the digital yuan.

They basically said, “You have to give us all the interest.” So there’s a trillion yuan there where, unlike Jeremy putting it into really high-quality U.S. Treasury bills and things like that, they just lost billions of dollars a year because of regulation. A lot of the Chinese payment regulation slowed things down.

If I look forward 10 years, what’s happening is that these systems are finding the most economically valuable things. Again, we need to define the value to end users. They’re using smart and then intelligent money on optimized rails to move that around, and then you have a self-balancing, self-driving economy if we get it right.

If we get the regulations wrong, there’s going to be competition to adopt it because money will flow where money is liquid. If Europe doesn’t adopt this and other countries don’t, it’ll come to the U.S., which has the regulation. I think that’s a world of abundance because we’ll finally be allocating things where they matter, as opposed to now, when most of the money just sits there, dumb.

I like that perspective, Emad. I definitely share that view. There is such a thing as a control function in corporations, and I think the interesting question is how we have provable controls in an AI-intermediated system.

Regulators mostly check that you have a control function that’s actually functioning. That’s what they do: “Show me your policy on this. Show me your policy on that.” They’ll get a third-party auditor to come in and perform a SOC 2 audit to make sure that what you said you do, you actually do.

Controls are really important. I think one of the things we need to work toward is how we have provable and auditable controls in an AI-intermediated system, both human-in-the-loop and agentic. We need to get policy and regulation to deal with these kinds of proving grounds for control mechanisms.

We just launched an open-source project called Secure Tool, which is a wrapper around OpenAI SDKs for agents. It specifically enables the ability to insert control functions into monetary transactions that agents would conduct, to deal with this issue of, “Wait a minute. I’m not just going to delegate this so the agent can party on with this wallet.”

I’m actually going to want the API to have a permissioning structure for the AI itself.

Now, that's just one example that we've crafted. And I think, again, it's a whole tool chain—entire tool chains can get built up in this area. And, again, I think it's probably the innovators, the builders, the tool developers, the software creators that are going to figure this out by necessity before the regulatory bodies.

But I do think, to achieve that hypervelocity—which I do think is an incredible world of output, capability building, wealth creation, and prosperity—we have to figure out what the control functions are. And that's obviously an issue with unbounded AI as well, right? Control functions, societal control functions, et cetera. So it's not unique to financial systems. It relates to interaction with critical systems and all that kind of stuff, too.

If we get it right, the future's going to be amazing. I remember Jeff Booth saying once that the reason he was excited about Bitcoin was that it gives you money velocity without debt. And I think what you've achieved with USDC is the same thing: You have money velocity without the extreme debt, which—

Exactly. It's a first principle. Yep.

Oh, Jeremy, I know that your 10-year vision, your founding vision, if I have it correctly, was to raise global prosperity through the frictionless exchange of value and build safer and more inclusive financial systems on a sort of internet financial system. And you've done that. I'd love to close with: What's your vision a decade from now? Do you have your founding mission a decade from now?

Well, I actually feel like we're still at the front end of realizing that mission. And I think—I talked to my team about this, and I just had a leadership gathering and talked to my team about this, and sort of looked out a little bit and described some of the things that we talked about here, which is really rewarding to talk about here as well. What is the nature of corporate forms? What is the impact of this societally?

Our secret is that we're maybe talking about an XPRIZE on this topic. It's an interesting idea, that's for sure. It's definitely an interesting idea.

I think—I mean, look, I think we're early, and ultimately I want to see—I mean, you could argue that data center, GPU, and energy spend has contributed all the GDP growth in the U.S. this year. So, arguably, at least GDP output has increased because of AI this year.

Okay. Yes, for sure.

And now converting that into other output is obviously the thing that everyone's focused on. I want to see, at some point, the economic infrastructure that we're helping create and the economic velocity that is actually measurable—because it's all measurable on-chain—lead to measurable increases in economic velocity that actually lead to measurable increases in global GDP and prosperity.

I'm not going to be happy until I can actually see that and measure that. And that's—we're not there. We're not there. We're not even close to there right now. And so that animates my work as I think about—

In the early days, when you digitize systems, there's disruptive growth that eventually becomes exponential growth, and it dematerializes, demonetizes, and democratizes the world. So I love it. Congratulations, Jeremy, on that. Where do we find you on the World Wide Web? What's your handle?

Yeah, so the best is @jerallaire on X.com. That's my public handle. And otherwise, you can check out what I'm working on at circle.com.

Amazing. Emad and Salim and Jeremy, we're going to all see each other in Riyadh in about 10 days' time at FIA. Yeah, it's going to be fun. Salim, how's your week ahead looking?

USDC for you to sign, Jeremy.

Okay. Okay. That's it. I can do a sign. I can use a digital signature.

Emad, how's your week ahead, buddy?

The work always continues, buddy.

You know, it's so funny. I get a text from Emad at, like, 6:00 at night, and I go, “Where are you?” He goes, “I'm in London.” And then I ask, “You're still awake?” He goes, “I'm always awake.” Oh, well. Amazing. That's good. All right, guys.

Congrats on everything. Just incredible.

Thank you. It's been a lot of fun. Thank you, guys, so much. Thank you so much.