Unchained - The Chopping Block: Robinhood’s Vlad Tenev on Tokenized Privates, 24/7 Stocks & AI-Verified Code - Ep. 891

Starting point is 00:00:00 There's a lot of talk about investor protection and protecting these folks from making poor investments, losing money. But we are now in a world where you can invest in a meme coin to your heart's content, but OpenAI and SpaceX are deemed too risky, which I think is just a bizarre, perhaps unintended outcome. Not a dividend. It's a tale of two fun. Now, your losses are on someone else's balance. generally speaking, air drops are kind of pointless anyways. Unnamed trading firms who are very involved.

Starting point is 00:00:33 D5.E.5 protocols are the antidote to this problem. Hello, everybody. Welcome to the chopping block. Every couple weeks, the four of us get together and give the industry insider perspective on the crypto topics of the day. So quick intro,

Starting point is 00:00:45 first you got Tom, the defy maven and master of memes. Hello, everyone. Next to get Robert, the crypto connoisseur and czar of Superstate. Good morning, everybody. Next, we've got Tarun, the Gigabrain,

Starting point is 00:00:57 Gauntlet. Yo. Joining us today, we have special guest, Vlad, fintech futurist, and founder of Robin Hood. Greetings and salutations. And I am a Steve of the headhiped man of Dragonfly. We're early-stage investors in crypto, but I want to caveat that nothing we say here is investment advice, legal advice, or even life advice. Please see chopping block that XYZ for more disclosures.

Starting point is 00:01:18 So, Vlad, you've got a lot. We've been talking about the assent of Robin Hood over the last year and a half, quite a bit on the show, but we're very excited to have you here to. answer to the community what the hell is going on with Robin Hood chain. So we've had a lot of discussions about, I don't know if you're aware of how much the crypto world loves to argue about the minutia of little tiny decisions that were made by some of these individual companies that end up becoming these larger sort of melodramas within the cryptocurrency. I'm familiar, yes. You're familiar. Okay, good. Good to hear. So first question off the bat,

Starting point is 00:01:53 why did you build Robin Hood or decide to build Robin Hood chain as a layer two instead of as a layer one? I think this has been discussed on Twitter, as you mentioned, and someone put it well when they said the decisions really about being an owner versus a tenant. And we thought that having our own chain ultimately gave us control over all the details end to end of operating the entire platform. And perhaps that comes at a slight cost of interoperability up front. But I think over the long run, tokenized products, including stocks and other derivatives, are going to be available on all chains. So I think developers will build bridging capabilities and eventually the choice of primary chain will be less meaningful.

Starting point is 00:02:48 But we thought up front having one chain where we can, control the entire end-to-end experience from the smart contracts to sequencer mechanics would be important. So the question, though, is so we've heard now that Stripe is building their own chain. There's been some reporting in the press, although Stripe themselves haven't confirmed it. And we now know that Circle is launching their own chain. And these two chains are going to be layer one blockchain. So they're going to be like Ethereum or Solana, as opposed to a layer two, which is what bases and which is also what Robin Hood chain is. Does that distinction strike you as meaningful for Robin Hood's goals or it's like whatever, as long as it works,

Starting point is 00:03:26 let's just build something and get it out there for consumers to use? I think the benefit of a layer two chain is it's slightly easier to have interoperability, meaning there's a lot of infrastructure already, a lot of infrastructure even built by our friends over at off-chain labs who develop arbitrum that make it so that we don't have to reinvent the wheel. Now, the argument for a layer one is you can actually rebuild everything from the ground up. So I don't think that it's an unreasonable option. But when we looked at it, we asked ourselves, what do we want to do right away?

Starting point is 00:04:09 What are the things that maybe we want to do two to three years from now? We thought layer two met our requirements. Okay. So what do you see then? The story for Robin Hood's layer too right now is that there's going to be tokenized stocks that are going to be issued on this chain, and they're going to be tradable through Robin Hood, but also tradable directly on chain. You can sort of take your tokenized stocks from Robin Hood directly into the chain. Correct me if I'm getting anything wrong here. At the highest, so that's maybe what the V-0, V-1 of this chain looks like.

Starting point is 00:04:45 At its highest aspiration, let's say, you know, draw forward to the timeline five years, 10 years, 10 years. years. If Robin Hood chain is wildly successful, what does that look like? Does it become the next Ethereum, the next Alana? Are there more users on Robin Hood chain than there aren't any of these other chains? Paint a picture for me. I think first off, when we set to build the chain, we looked at the landscape and a lot of companies were aiming to create some variant of like the number one chain for DGens, right? The the chains for DGEN space seemed very crowd. And we didn't really think we had any competitive advantage there to make a product that was unique. Maybe this is debatable and we do have some advantage there to your point.

Starting point is 00:05:31 But I think where Robin Hood's advantage is, is in being the place where traditional finance and crypto merge and meet. So that's why when we looked at business to business opportunities in the past, we launched Robin Hood Connect, which is an on-and-off-ramp provider. So a lot of the leading non-custodial wallets and DAP providers integrate Robin Hood Connect as the best and lowest cost and most seamless way to convert your Fiat into crypto and also turn it back out.

Starting point is 00:06:04 So if you have crypto non-custodia, you wanted to get fiat, we provide a really easy mechanism to do that. So it's sort of like what capabilities have we built by virtue of having one of the largest consumer retail platforms. We've gotten really good at onboarding account seamlessly. We've gotten really good at preventing fraud and doing those types of mechanisms. We have a wide diversity of different assets and products that we offer on the platform. And really, over time, as I look at it,

Starting point is 00:06:38 the goal is for any financial market, any asset, whether you want a custody it or transactant to be available in Robin Hood in the easiest to use way. And so, When we look at the chain, the thesis is really, can we make this the best chain for real world assets? And I think we can start with the assets that we already have. You know, U.S. stocks, you can imagine any asset that we offer on Robin Hood pending regulatory maturity to eventually be available in an on-chain form for distribution predominantly outside the U.S., but I think there are advantages within the U.S. as well.

Starting point is 00:07:20 And then there are certain things that actually aren't easy or really feasible today, for example, private markets that we can make available as well. And that's kind of what we demonstrated at our event in Kahn. We wanted to show that, okay, here's something that maybe you did expect, which is stock tokens for public equities. but then something you didn't expect, which is that the same technology can be used for private companies, which have thus far been really inaccessible to retail. So you infamously were tokenizing, I believe, Open AI and SpaceX.

Starting point is 00:08:02 And there's a little bit of kerfuffle online after you guys announced this. So I remember everybody being super bulled up. Everyone is very excited that, oh, my God, Robin Hood is tokenizing stocks and bringing them on chain. And now it's going to even include private companies. and then Open AI, they tweeted something, basically like, bros, this wasn't us, I don't know who owns his equity. We didn't agree with Robin Hood to do this. And it's almost like, I'm curious what, because I don't know if Robin had an official response to Open AI kind of raising their hand and saying, hey, we didn't know that this was even happening. Although it does speak to the structure that you guys are using, being able to basically take any financial asset, regardless of what it is or whether or not, you know, the, you know, the,

Starting point is 00:08:44 the company itself is privy to it, of taking it and making it tradable in real time and ownable in real time on chain. Talk us through that episode and, like, how should we understand? Is this a sign that, hey, private companies want to be private and, like, they're going to fight back against attempts to tokenize their stocks? Do you feel like other private companies are going to be more amenable than opening AI was? Help us understand what that looks like.

Starting point is 00:09:07 Yeah. I'll see a couple of things that I think are relevant to this. Number one, when people have tried to do this in the past, you know, other companies try to give exposure or there's even, you know, Reg A Plus in the U.S. that allows limited forms of crowdfunding. Basically, the problem has been that there's an adverse selection issue. So the companies that opt in to making their shares available to retail, typically are companies that don't have a ton of other options. Now, there are exceptions. Some companies really care about it and they want retail to have access, particularly at

Starting point is 00:09:51 later stages. That's been a bigger thing recently. And I could talk a little bit about what we've been seeing on our IPO access offering as well. But IPO access, you know, we allow companies to access retail during the IPO, great for retail customers, I think great for the companies as well. And when we launched this in 2021, we had to do a lot of work to communicate with companies, the value of that. And nowadays, you know, pretty much the best companies are coming to us and actually giving retail high allocations. So now companies are starting to see the power of having a diversified and large

Starting point is 00:10:29 retail shareholder base as a constituency. So it's been an evolution. But when it comes to private companies, the adverse selection is real. I mean, if you are a company like OpenAI and SpaceX, and you basically have the freedom to raise tens of billions from private market institutional investors, they don't want to deal with the complexity a lot of times of doing anything different, particularly if it's not established or if there's not a lot of precedent. So tokenization of private companies is a new thing. Retail access for private companies is a new thing. And when you cross that with companies that have lots of options, you know, you talk to these companies and their core business is serving their customers and creating products.

Starting point is 00:11:19 they don't want to be innovators in capital markets. But we do, and we think it's a huge problem that, you know, you've got these private companies that are literally changing the world and leading their industries. It's looking like retail might not have access to them until they hit the public markets, if they do, add valuations in the hundreds of billions or maybe even trillions. So I think there's a big risk, actually, if we don't solve this problem. There's a big risk in the normal people just not accepting these technologies.

Starting point is 00:11:54 And that could lead to a whole lot of societal problems that could actually thwart progress. If you feel like AI is going to change the world and possibly lead to labor market disruption and take your job, there's one world where you don't actually own any of that. And there can be a tendency to kind of push back against it. But if you own the AI, suddenly you're kind of like rooting for progress there. And I think we've seen that play out in the public markets. And I think it's going to play out in privates as well. That's why we care so much about retail having access.

Starting point is 00:12:32 Well, actually, one interesting thing that's kind of related, right, is Anthropic actually is doing a new round and they are forcing all of their new investors to not use SPVs. So forget about pure retail investor, like even accredited. but not, you know, not a fund is kind of being kept out. So I guess like one question I have for you is how do you think about kind of companies not disclosing as much and that being priced differently between the private rounds and sort of like the public proxies?

Starting point is 00:13:06 Because like there is sort of this weird thing, right, where like the retail investor's disclosures for the companies are actually much less than like the private investor who's getting preferred shares. And so there's kind of always some information asymmetry. How do you think like that kind of gets resolves itself? Because I do feel like that always seems to me like one of the biggest problems in like the derivatives for private companies type of market. Yeah.

Starting point is 00:13:29 I mean, I think that the way it typically works in private markets with these rounds that are 30 to 40x oversubscribed is if you're a SPV or some kind of institutional invest, or private wealth, or, you know, you're part of the private wealth groups at these big banks and you end up getting an allocation, you're not doing a ton of diligence, right? And there can be, you know, you imagine a company that's 40x oversubscribed and you've got people clamoring to get into the deal. They're not going to spend time allowing each SPV participant to do deep diligence for the company. And I think for a lot of these investors, that's okay. They look at the fact that soft bank may be investing or Altimeter or one of these top funds, and they're happy

Starting point is 00:14:27 piggybacking off of that diligence. So I'm generally big proponent of disclosure, but I think there's a lot of people that would say, I understand the risk. I'm willing to take it. I use chat, GPT as a product or I'm aware of the space industry. I see the publicly available information. And I fully understand that, you know, it's theoretically possible that this investment can go to zero. And I want to make it anyway. And I think as a believer in free markets, I think people should be allowed to make that

Starting point is 00:15:04 decision for themselves. And, you know, in the U.S. where we have accreditation standards, we've been pushing for some reform there. And I think a lot of the basis for these types of accreditation rules are antiquated. They're pre-internet. It's really when there wasn't a ton of information available about these companies. Now you have AI and you have a ton of information out there easily accessible on social media. And we have the tools to digest it and make sense of it. And so I think they have to be revisit it and it should go much more toward a disclosure, self-certification regime. Because again, if you also have to impose restrictions on companies to do idiosyncratic things

Starting point is 00:15:52 for retail shareholders, then the risk is they don't want to do that. And they're happy just like dealing with very concentrated small set of large money managers. and you end up with the same problem that we set out trying to solve in the first place. Totally. So let's square that with the current rules because a lot of what you discussed before was that these products would be available X-U-S and not focused on U.S. You know, A, is that because the current rules in the U.S. are too limiting. And B, you know, as you look to launch this, how do you envision this sort of X-U-S offering and structure? You know, are you planning to like KYC and geo-fence this and say these products are

Starting point is 00:16:32 strictly not available to users in the U.S. Are you planning to make the blockchain itself permissionless and the assets on them permissionless? What do you envision the structure is based on the current rules? And obviously I know that you hope the current rules evolve. But what's the structure that you see this going live? Yeah, it depends on regulatory maturity in each market. High level, what we expect this to look like is Robin Hood would

Starting point is 00:17:02 get exposure or positions to a variety of private companies. And then we would package them up and make them available to retail either directly where applicable or in other vehicles that may be diversified, depending on what's permissible in different markets. So right now, I think there's a good analog with stable coins, right? Stable coins are sort of the original tokenized asset. And the way it's evolved is in the U.S. where capital markets and financial rails are relatively robust, people are using ACH debit credit to move money around. But stable coins have become a great

Starting point is 00:17:46 delivery mechanism for U.S. dollars outside the U.S. And I think you'll see the same sort of situation for tokenized versions of U.S. equities. In the U.S. where capital markets are pretty robust, the uptake might be a little bit slower, but outside of the U.S., where access is a little bit more challenging, the tokenized versions will develop into the, I think, the leading vehicle for exposure to U.S. assets. And, you know, it's hard to make predictions about how the U.S. exactly will shake out. You know, tokenization isn't broadly permissible as a retail vehicle today, but that might change. And I think if it changes the biggest and most immediate sort of like visible improvement and quality of life for retail investors will be exposure to private assets,

Starting point is 00:18:45 because tokenization solves the liquidity problem. It used to kind of be a problem, even if you were able to get your hands on private assets, you have restrictions for how to sell. Pricing was a little bit unclear. There have been market. places and different secondary markets that have tried to form, but they had to build their own market. And they had a hard time getting enough volume of buyers and sellers. But suddenly you plug in, you know, the global crypto networks and the liquidity problem becomes much less severe. So yeah, I think that's going to be the biggest change if, you know, regulations mature and it's allowed to happen. Yeah. I think, I mean, I hear.

Starting point is 00:19:29 you on tokenization being kind of an access story. I mean, stable coins, I think being a great example of that so far. I think I look at other companies in the crypto space that are also going on chain to sort of accelerate, go to market, and improve access for financial services that might otherwise don't be able to. Like I look at Coinbase basically offering Bitcoin lending through morpho, through their wallet product or, you know, Phantom offering hyperliquid perps through the product directly, and they are able to sort of accelerate their go-to-market through this kind of self-custody model where they're interacting with a defy application on chain, but they themselves are just a wallet or a provider. Is that something Robin Hood is looking at all? Or have you thought about?

Starting point is 00:20:12 I mean, how do you think about those companies kind of approach to this? Well, we have the Robin Hood wallet, which has been growing in an accelerated way, and we're continuing to invest there. The Robin Hood chain, obviously, is another component. And yeah, we think that, you know, with tokenization, particularly overseas, but, you know, hopefully eventually in the U.S., we can offer all kinds of financial products that have fundamental utility, starting with stocks and private markets, but eventually extending to the wide breadth of what we have available. We want to make those available on chain in a permissionless way. And I think we have an advantage there because we're kind of the only financial services platform that has large scale in both crypto and traditional assets. And if you look at tokenization, it's where those two intersect. So you still need a traditional financial service company to custody all of the TradFi assets in order to mint and burn them and deliver them on chain.

Starting point is 00:21:19 So there's kind of two axis of vertical integration. There's the blockchain one where you can offer customers better pricing if you own the chain, the wallet, and kind of everything in between. But also if you own the custody layer of the traditional assets, it again gives you not just more control, but more ability to offer lower pricing and better economics to customers. So actually, I have a question for you. What do you think happens to like transfer agents, DTCC? etc in a world where like 10% of stocks move 25% 50 because like it's kind of an interesting question of like if this structure ever gets to higher volumes than the current underlying like does that does that like flip the kind of story on like how how you currently deal with kind of being a broker dealing

Starting point is 00:22:11 a transferage dealing with a whole like ecosystem of people whose websites have uh no 2FA yeah um My guess for how this shakes out will be that it'll be similar to an ADR or like an ETF custodian where you'll still have the traditional assets that are custody at a Tradfai custodian. And you'll have traditional trading of those assets on public exchanges whenever you can get a better price there. at least this is how our system is designed as we outlined in our event and can. But then there's also going to be secondary trading of the tokens as well. And at first, the value prop there will be the tokens can trade when the traditional markets aren't open.

Starting point is 00:23:07 So weekends, holidays. And over time, we'll see it expand from there. So, Vlad, the 24, going from 245 to 247, like that story about, okay, stocks can trade now on weekends and when markets are closed, totally makes sense to me. And I think that that feels kind of inevitable, especially given the way that Robin Hood's core products have always evolved in that direction of making everything easier for retail. This private market story is the one that I find really fascinating. And I heard from your answer that you think that the private market side is actually in the long run, the bigger story than just, okay, things are maybe more efficient on the public. public market side. What strikes me, you know, being ourselves investors into private companies, there's long been this kind of push and pull between private markets and public markets,

Starting point is 00:23:57 which is that, of course, companies are being private for longer and longer. You mentioned, you know, look, you might lose your job to Open AI. Maybe at least you can own a little bit of opening AI stock. But it's very clear private companies, despite the fact that they might want the rules of going public to be easier and for it not to be as much of a cost, they also really don't like. And I feel like they've become smart and smarter about this since the Facebook days. They've become smarter and smarter about not allowing secondary market transactions, not allowing SPVs to trade. So there's a lot of things that I'd imagine even just in response to this adversarial game,

Starting point is 00:24:32 essentially, that's being played with private companies of saying, hey, as long as there's enough SPVs out there, Robin Hood can find one of these SPVs to go take, tokenize its ownership in this private company, and then allow retail to get access and to get liquidity. What do you think is the end point of that game? Because at the end of the day, the company has the ability to say, hey, we don't like what you're doing, screw off or we're taking you off the cap table. Yeah, two things.

Starting point is 00:24:57 One, I think it's important for the tokenization mechanism to work well without necessarily the opt-in of the company. That way you... How can you do that? How can you do that? Well, I mean, we demonstrated that by these things that you've outlined, like if you get indirect exposure and an SPB as an institution, you can then have the traditional assets in a bucket essentially and tokenize them or otherwise make them available to retail by other vehicles.

Starting point is 00:25:32 So for example, when you're in the public markets, it would just not work if every company had to approve every shareholder or every transaction. So obviously, if the company had to approve that in the private markets for tokenization, tokenization wouldn't work. So I think it's important for the mechanism to work without having the companies actually be in the loop. And not just because they might not want it, but also because they probably don't want the extra work or the extra steps. I think a lot of it is that they actually don't want it.

Starting point is 00:26:12 I think they actually do not want a liquid price to be set every day. And I'm sure you probably understand the reason why, because before Robin Hood was public and post being public, I imagine you feel very much the pressure of public markets and the fact that there's a stock price that moves every single day in response to what you're tweeting, what you're saying, what you're doing. Private markets very much love the illiquidity of markups. And they love the opacity of not being.

Starting point is 00:26:37 not being able to say, oh, hey, actually the company's worth less than it was yesterday. Yeah, I mean, a lot of people bring this up as an argument. Like, oh, when you were private, you probably would have not wanted this. I think there's lots of private companies, each have slightly different things that they care about. For me in particular, I don't think I'm a great example because I would have been fine with it, clearly. But there's also lots of companies that are interested in doing this. And it's not simply companies that can't raise money other ways, but it's companies that understand the value of retail.

Starting point is 00:27:13 So the optimal scenario would be, you know, we work with companies and they understand the value of what we're doing. And they make it available to retail at earlier and earlier stages, culminating with retail actually being a vehicle for company formation. So you have companies being formed. at Seed with retail investors being a big component of the cap table from the earliest possible stage. It sounds very crypto-inspired. Yeah. Sounds like the SEOs. But in this model, Vlad, it's the company's opting in.

Starting point is 00:27:50 Yeah. They're giving you, because they're excited about it, because they think that having a ton of retail on the cap table is a good thing. You know, they're making themselves available while they're still public, private, versus companies that are being permissionlessly taken public. while they're still private against their will. So as I hear this, it's opt-in, not we can do it against- it. Well, it sounds like some are opt-in, some are, hey, we're doing this. Guess what?

Starting point is 00:28:14 Opening eye is too important for it not to be accessible to retail. I think there's a slight nuance here, which is it depends on who the customer issuing stock is. So I think if you're talking about early stage, the customer of what's really capital as a service is the entrepreneur. It's someone creating a new venture. they want to raise capital, and that capital is primary in nature. And obviously, if you're an entrepreneur raising capital on a primary basis, it's opt-in.

Starting point is 00:28:46 As you get later and later stage, the customer-based shifts, and it becomes senior executives and employees and early investors seeking secondary liquidity. And I think it's worth debate how much freedom employees and early investors. investors should be given to sell their stock, right? I mean, people have different views in them. Obviously, some companies lock that down quite acutely. Others don't. And in the cases that they don't, investors can sell their shares without the company

Starting point is 00:29:24 being in the loop, right, without the company opting in. So every case is a little bit different. Hopefully, you guys can appreciate the nuance. But do I believe that contractually every employee and early shareholders should be, should require the companies opt in? I wouldn't say that. I mean, that's sort of a legal decision that the company can make. Yeah, but I, so I'm less focused on, okay, who's you getting this from?

Starting point is 00:29:50 Is it an employee? Is it some early investor? I trust that at any cap table, there's somebody who the founder doesn't really care if they get liquidated because, you know, whatever, it's fine. But there's also people they don't, they do care about. There's people they do care. And I'm sure Robin Hood is not probably getting their arms around those shares anyway. The question I want to get at, I think is more interesting.

Starting point is 00:30:09 And it's kind of very much what you pointed at, which is Robin Hood's story has always been this kind of Promethean story, which is that we kind of bring the flame from on high down to give retail investors access to any financial asset, any stock, any financial instrument that normally you're not allowed to touch. Now you're allowed to touch them. And part of what this story does, I think I appreciate the virtue of the story. and it's very aligned with the kind of crypto ethos. But at the same time, it also collapses the distinction between public and private companies.

Starting point is 00:30:39 Because the whole point of the creating this entire regime of all the shit that you have to do in your public company is to make it so that there's this big boundary between public and private companies. And presumably it's all grounded in investor protection and clarity in markets and blah, blah, blah. And the idea is that, okay, if the company's private, we don't really have to worry about that. We can more or less let them kind of run amok and just get away with murder, relative to if they're a public company, then okay, everything has to get pretty cleared. If retail is getting access to all these stocks, like not of them, obviously, but like attractive companies that people are interested in, and you can buy it alike for public and private

Starting point is 00:31:14 companies, from a social perspective, doesn't it kind of erode the entire meaning of the regime that we've built around public and private companies? What do you think the regime ought to be instead in the world that you're trying to enact? I don't think it erodes it. I don't think anyone sat down and constructed a system where, you know, 80% of individuals in the U.S. can't have access to these types of appreciation opportunities proactively. I think it's kind of evolved incidentally, right? It started as investor protection.

Starting point is 00:31:54 It started as removing, making it easier so that companies could operate. and get started without disclosure environment, without disclosure environment requirements, sure. But what's happened was what's happened in the past few decades is it's gotten harder and harder for companies to go public. There's just more onerous requirements on that process. Simultaneously, private markets have become easier to access for a lot of companies. There's more VC firms, more private equity. So access to capital has become easier. way. And the combination of those two things has led to less companies going public or when they do go public, it's sort of like at very high valuations. And I think collateral damage of these things

Starting point is 00:32:44 were that retail investors don't have access to these opportunities. So I think we can probably address some of the concerns while also giving retail investors access to these great opportunities. Actually, so to what you're saying, there's actually this very funny thing, crypto microstructure that's happened repeatedly that sort of reminds me of this. So like, I sort of view what you're describing as like this like inevitable tug of war between the private and public markets over like who gets to like have a first earlier allocation. And in crypto, oftentimes what you see is like private rounds in crypto will have some retail syndication when there's sort of like a bull market. So we saw these platforms like Echo or Legion

Starting point is 00:33:30 where like you could companies would raise private rounds and then like syndicate some fraction like Angelist SPV style to purchasers. And then somehow that stuff grows too quickly relative to like the true growth rate of the companies. And then you see this flip where everyone tries to go fully private and like stop syndicating. So I think the interesting thing is crypto sort of replays this type of thing at like a faster time speed, right? It's like you can see 20 public market cycles in like an hour of crypto sometimes. And there's kind of this interesting question of like, if you cross the chasm of this, will there kind of be like a back, you know, like to Haseet's point?

Starting point is 00:34:12 Like what is sort of the reversal? Or do you think this is like a true one-way function? Like there's no, there's no inverting it. Yeah, it's really hard to predict how this will shake out. What I'll tell you is in crypto, I think the difference is that we haven't been allowed to connect crypto assets to companies, to company equity and things that have fundamental utility, as I like to call it. And what you see happening is as a result of that, downstream of that, you have a lot of

Starting point is 00:34:44 activity in meme coins because, and exactly because they're disconnected from fundamental. utility because that's what's allowed. And, you know, there's a lot of talk about investor protection and protecting these folks from making poor investments, losing money. But we are now in a world where you can invest in a meme coin to your heart's content, but OpenAI and SpaceX are deemed too risky, which I think is just a bizarre, you know, perhaps unintended outcome of the current rate. Absolutely it is. Absolutely it is.

Starting point is 00:35:26 Well, actually, speaking of like, kind of you were talking about like the idea that access is sort of limited. How do you feel about, you know, like the digital asset treasury companies? Because they're sort of, they're sort of like this perfect hybrid of like all the things you said where it's like, they're vehicles that are encapsulating assets that are not in a company format, but now you can buy them as a stock. I'm kind of curious like what's your, you're sort of like. how do you feel about these kind of weird rappers? Because there are also actually for private stocks. There have been a few ETFs that like, or sorry, closed-end funds that are like,

Starting point is 00:36:01 oh, we own a bunch of striped stock. If you buy our ETF, you basically own prorata share and like whatever closed-end stripe stock we bought. So like, how do you view these kind of weird vehicles that exist and like what do you think about the fact that they've grown so much in last year? Yeah, I mean, without actually talking too much specifically about different. investments, I tell you that, you know, I believe in free markets so that if someone lists an equity or other type of instrument or even a cryptocurrency and it's compliant with the rules of the law, it should generally be accessible for retail. And I think that the beauty of free markets is

Starting point is 00:36:45 that you let the market decide, right? Like market, if a product is attractive to consumers, over time, the fundamentals do play out and you end up sort of like getting to a market solution to some of these questions. So I wouldn't necessarily give you like a bull or bear analysis of every treasury company or closed end open end. I mean, if you did, it would be hilarious because I would restrict you to one emoji only for your review. Yeah, but I think that, you know, it's from my perspective, we want to allow access to all types of financial assets and transactions. And we think Robin Hood can be the place to do that safely in a compliant manner, also with a great user experience and at low cost. So, you know, we have the opportunity to list stuff.

Starting point is 00:37:43 Yeah. I'm realizing that we can't get you to actually talk shit about digital asset treasury, so that's fine. So an interesting thing here is that, So Robin Hood traditionally was basically a place to trade mostly U.S. equities. And over the last couple of years, more and more of the business, the revenues and the earnings have been attributed to crypto. And at the same time, one of your biggest competitors is Coinbase. Coinbase has been a traditionally crypto business that now is talking about starting to add equities, almost like this sort of convergent evolution, is that there maybe is some

Starting point is 00:38:19 kind of deeper structural shift taking place where the newer generation, sort of, you know, Gen, Millennials, Gen Z and then Gen Alpha, that the places in which they're going to trade are ultimately going to look more and more like the union of crypto and traditional finance. How have you thought about the way in which Robin Hood is a business? I have to imagine that you probably weren't expecting crypto to be as explosive as it ended up being in the growth of your business. How have you changed your perception of what Robin Hood is supposed to be with how? much crypto has grown as a portion of your business within the last few years?

Starting point is 00:38:53 Yeah, I think the underlying thesis for Robin Hood has remained consistent, right? We want to be a financial super app that helps you handle all of your financial needs. So custodying any asset, conducting any financial transaction. And, you know, do you point out crypto as a percentage of our business has become significant. I think last quarter we had just under a billion in revenue, Crypto was, you know, 20 to 30% of that. As crypto technology becomes more and more infrastructural and Robin Hood becomes more global, I think the delineation between the two will blur. So, for example, if you have a tokenized stock transaction or custodying stock

Starting point is 00:39:38 tokens or private tokens, do you put that in the crypto bucket or kind of the Tradfly bucket? And I think the, anytime you categorize stuff in this way, you end up having to deal with these edge cases. For example, technology as a sector is becoming less and less relevant as every company now. I mean, it's kind of silly to think about non-technology companies because even they have to use technology. So I think over time, the distinction will become less material. And if you're a customer of Robin Hood, you know, you'll probably be using crypto more and more without it being, front and center in your experience. Of course, if you want to buy Bitcoin, it's very clear that you're investing in crypto as an asset. But if you're getting a tokenized asset, crypto is really

Starting point is 00:40:25 an infrastructure technology. And it's sort of a delivery mechanism to the instrument that you end up wanting to get. Actually, and sort of related to this question about the juxtaposition with kind of Robinhood starting with stocks and adding crypto, Coinbase going the other way, is, you know, I think the other difference, I guess, between two businesses has been, like, the institutional slash infrastructure side of, like, owning the full stack of custody and staking and other aspects in order to support institutional usage versus retail usage. Do you view that as things become the everything market for both crypto and non-crypto, that you'll have to own more of the infrastructure directly than kind of like public markets,

Starting point is 00:41:17 you can never really own the full stack. But in crypto, it's kind of weird. Single entity can kind of own the full stack for their end user. And so, like, do you see yourself owning more of that? How do you, like, view that? Because, like, I do feel like that's a big shift, right? If, say, tokenized stocks dwarfs the volume of your traditional stocks? Yeah.

Starting point is 00:41:37 Yeah, I think that I see our business evolving with three arcs, right? One arc is sort of like short term, then we have a medium term one, and then sort of a 10-year long-term arc. The first one is really just being the number one platform for active traders across all assets, crypto, stocks, options, futures, every asset that we offer. The second arc, which is sort of like building right now, is us being number one financial super app for our customers. So all of your wealth is in Robinhood and all of your transactions go through Robinhood. So the first two arcs are like retail only. Basically, we think we can do that without actually vertically integrating completely and serving businesses and other institutions.

Starting point is 00:42:29 But what happens is by virtue of us building these capabilities, things like 24-hour trading, rock-bottom margin rates, tokenization. We build products that are attractive to businesses and institutions. If you're an institutional investor, you also want 24-hour access to markets. You want crypto and traditional assets in one place. You also want the lowest possible margin rates. And so that plants the seeds for compelling B to B, and, you know, and institutional offerings.

Starting point is 00:43:00 And I think if you look over a 10-year horizon, our business X-U-S could be even bigger than U.S. And business and institutional could also be even bigger than retail, which I think makes the opportunity. I mean, I get excited when I think about what we can do in the long term because we're just like at the very, very beginning. And I see multiple ways that we could 10x this business. and we can kind of pursue multiple of those vectors simultaneously.

Starting point is 00:43:34 So you guys are now in a position of both being a contributor to the crypto ecosystem with Robin Hood chain and all the stuff that you guys are doing there. You've also, in some sense, also been on the other side of like, okay, you're working with Arbitrum and you're getting a sense of what we as an industry have built. And I'm sure you probably have some sense of like, hey, guys, this part of what you guys built kind of sucks and really some improvement. We have a lot of founders who listen to the show, a lot of entrepreneurs, a lot of builders. I'd love to just get, you know, Vlad's feedback.

Starting point is 00:44:04 What do you feel like the crypto industry should be doing better? What have we underdelivered on and what would you like to see more of from builders in this space? Well, I think that I don't really view it as, oh, I'm a customer of the crypto industry. I do think that we're a very active participant. I mean, we're one of the largest companies in crypto. If you look at market share and sort of like volumes and revenues, particularly in the U.S., and the thing that I felt was missing was that crypto and the financial system have basically been two separate worlds with like minimal connections in between.

Starting point is 00:44:50 Stable coins being kind of a recent counter example of, okay, they are, they are starting. starting to merge. But my approach to that wouldn't be, hey, you guys should go fix that. Like, we're in the arena trying to improve that and, you know, leveraging our capabilities and the things that we're great at to make that better. Surely there is some, like, time component to that. Like, there was, okay, hey, Robin Hood, we're doing this arbitram L2 now. Maybe you wanted to do it several years ago.

Starting point is 00:45:22 I'm sure there are things where you're like, man, I would love to do something today, but the tech isn't there or this product suck or there's not enough liquidity. I mean, even, you know, kind of these stuff around these private market assets or we can also talk about some of the prediction market stuff that you guys are doing. It feels like maybe now is the time. What is the thing where you're like, I wish now we're the time, but we're not there yet. Part of the context for this is that we spend almost every episode bitching about the industry. So that's why we're trying to get a little bit of that from YouTube.

Starting point is 00:45:51 Talk shit, yeah. Yeah, yeah. Glad unchanged. Yeah, I think that, I mean, there's been some talk about AI agents and whether AI agents are going to be paying each other with crypto to conduct transactions. I think there's also real-world AI, right? You're talking about robotics, medical devices, things of that nature, and are real world AI agents going to be paying each other in crypto and kind of what that looks like.

Starting point is 00:46:28 So I think we're probably one step away from both of those things and two steps away from sort of like a intersection of crypto being used as a tool of agentic commerce. Okay, so taking the AI point that you just made, one of your, I don't know, side quests is you're working on this company called Harmonic, which is, if I understand correctly, you're building a foundation model that can basically verify mathematics using lean and theorem proving in order to verify the answers that it's generating something like this. That's right. What's the, first of all, how do you have the time to do that as you're running company? What's your, what's your level of involvement there? And tell us the story of like, why this? Why that?

Starting point is 00:47:14 Yeah, so I'm chairman of the company. Again, this is an entirely separate company from Robin Hood. So I don't have an operating role, but I am the founder and I care about it quite a bit. And actually, a couple of weeks ago, Harmonic announced that our model called Aristotle reached gold medal level performance on the International Mathematical Olympiad. So if you guys were familiar, this is a very, very hard mathematics exam. And off the shelf models like GPT5, GROC, the others, get like barely even one complete question, right? Like they do very, very poorly.

Starting point is 00:47:57 And part of the issue is that, you know, usually these problems require a spark of creative genius. And if you don't find it, you're not going to solve the problem. The other is that they require, like 10 logical steps. And if one of those steps is wrong, then your proof is wrong and you don't get the answer either. So you have to be both creative and you have to make sure that every step is correct because hallucinations in this case kind of compound. So it's a great demonstration of the technology. And in fact, we were, we did this in a formal way. So you mentioned lean,

Starting point is 00:48:37 the theorem prover. So we produce formal proofs of all these results, which means a human doesn't have to validate it. You can pass it into the lean kernel, and the lean kernel checks that every step of the proof, and the result is correct. And now you might ask, like, what's the application of this? We call it mathematical superintelligence. The application of this is that you can not only use AI to just create output and answer your questions, but you can use AI to validate that the output is correct up to a very, very high standard. So you think about all this computer code that AI is generating right now. It's kind of turned the job of an advanced software engineer from writing code to verifying AI generated code. And if it's like a run-end UI, it's pretty easy to

Starting point is 00:49:32 verify. You could just look at it and make sure that it looks like the spec and what you intend. But if it's a back-end system or if it's a smart contract, particularly if the consequences of failure are losing hundreds of millions of dollars, humans have to have to verify it. And I think that kind of creates the market for formal verification. So yeah, we're excited to build that. I think the applications aren't just math, but it's really also all software. and even hardware, too. So you think this is a better path to building software engineering agents,

Starting point is 00:50:09 like what Anthropic is doing? You're like, hey, having this foundation of theory improving and this mathematical undergirding is necessary to get to real AGI. I think it is a better path. And also it's inevitable, right? Because if you're in a world where AI is producing voluminous content, that's too, that's just too much for any one human to read. We need another way to actually make sure that it's correct.

Starting point is 00:50:39 Two things, we need to make sure it's correct. Also, it doesn't have to be in languages that humans necessarily understand because humans aren't even reading it anymore. So I think the entire substrate changes and you have to think about how can humans really, really quickly verify that it's doing what it's doing. supposed to be doing without having to like go through a line by line. So I actually have two questions for you related to this, which is one of sort of the aesthetic criticisms of formal verification is like you know you dropped out a

Starting point is 00:51:12 math piece sheet. You probably got into it you know as a similar dropout. I like you got into it because there was some aesthetic beauty to some proofs right. You're like this particular area was just like such a beautiful theorem or like how concisely succinctly it was proved. And somehow math has always historically had this like tradeoff between aesthetics and computation. They've never kind of like coexisted. But oftentimes maybe the lean proofs that are generated might look like, you know, those, you know, the coloring theorem proofs that are like really ugly relative to like the like indirect ones that we have. So how do you think about like preserving aesthetics?

Starting point is 00:51:49 Have you have you seen Aristotle's lean proofs? I've seen ones. Like I haven't seen like the compendiums. Yeah. I mean, I find them very beautiful, actually. And so the thing is, converting from a lean proof to natural language is actually a pretty straightforward problem because you can almost even do it mechanically. Like, you could look at the functions and lean and their descriptions. And you can very easily go from something that has more detail to a more descriptive sort of like English exposition.

Starting point is 00:52:27 of it. I think it's particularly easy for lean relative to other formal languages in the past, which I think in part has led to being popular, not just for AI, but for mathematicians that are trying to formalize stuff. So there's that. The other way is the difficult part. Getting an informal proof and formalizing it is actually a lot of work. And there's a huge formalization initiative right now being run by this professor, Kevin Buzer, in the UK at Imperial College. She's trying to formalize Vermont's last theorem. And this is an initiative that's like gargantuan in scale.

Starting point is 00:53:07 It's sort of like hundreds of mathematicians, multiple years, to formalize this result by hand. So you can kind of tell the difference in difficulty that way. Well, I guess the last question that I wanted to kind of go to is like, what non-millennial prize problem would you want salt? Would you want kind of, you know, Aristotle or successors to solve? Like, like, Millennium Prize problem is like an easy answer, right? It's like, okay, yeah, obviously, like people have said these are, you know, Riemann hypothesis and P's not equal to NPR are worth thinking about.

Starting point is 00:53:42 But what is your sort of pet problem and like which, which thing do you think would be like, if you solve it, you're like, I've succeeded. Okay. So not the Riemann hypothesis. Don't, yeah, it's too obvious, right? No obvious. I feel like this too obvious. Because I do like that one.

Starting point is 00:53:59 I think like a benevolent Hal 9,000 would be pretty cool. If we want to, I would love to build like the logical core and the control center for a spacecraft. And obviously not have it go haywire and try to take over. I don't know if that's the example you want to give people. It's a proof of proof of spacecraft. A benevolent, a benevolent. Yeah, yeah, yeah. Obviously, we need good control systems that are reliable.

Starting point is 00:54:26 Ostensibly benevolent. Ostensibly benevolent. Okay. But will it be able to trade stocks on Robin Hood? Or will any of this outcome of this work? To the moon is going to be much more literal now. Well, I think that the initial goals were obviously quite separate. But I think there is overlap.

Starting point is 00:54:47 So, for example, smart contract verification. If you think about a smart contract, it's a relatively self-contained piece code running on solidity or rust or what have you. There are certain things that we want to make sure that, you know, hold like the smart contract doesn't halt and you can't double spend and all those sorts of things. And actually the consequences of that failing could be hundreds of millions of dollars, as we've seen. And right now, if you're a smart contract firm, you're basically forced to go through this

Starting point is 00:55:25 manual auditing process where you pay some firm, you know, hundreds of thousands, even millions of dollars to comb through it. So there are examples in the crypto space where I think formal verification, once it becomes practical, could be very, very powerful. You could look at any query where you actually have to do complex mathematics, you know, if you're computing margin or options pricing. So there are lots of examples in our industry of software where the consequences of failure can be quite severe. So I think financial services, along with healthcare, automotive, aerospace, robotics, they're all sort of like right for formal verification. Well, I'm sure we've now lost. The vast majority of our audience is totally lost.

Starting point is 00:56:17 but I will say, Vlad, I'm very excited to be able to buy tokenize your DTE options on Harmonic and bring them on chain as soon as they're ready. So we're up on time. No, no, no, no. You're not just buying the option. You're getting a proof that the Black Joel price was computed correctly. Oh, good, good, good. Yeah, that's right.

Starting point is 00:56:33 You get a lean proof alongside your purchase. It's like a happy military. And sharing your perspective with us. It's great to have you. Yeah, thanks for having me. Excellent. All right. Until next time.

Starting point is 00:56:43 Thanks, everyone. Thank you.

Unchained - The Chopping Block: Robinhood’s Vlad Tenev on Tokenized Privates, 24/7 Stocks & AI-Verified Code - Ep. 891

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.