Mathematical superintelligence represents a paradigm shift in problem-solving, where AI systems can solve complex mathematical problems at superhuman levels by combining formal verification with creative insight. Unlike traditional AI that may hallucinate or make errors in reasoning, mathematical superintelligence requires rigorous proof verification to ensure every step logically follows from axioms. This approach, exemplified by Harmonic's Aristotle platform, aims to solve Millennium Prize problems and formalize unsolved research-level problems like the Erdős problems. The key insight is that mathematical understanding involves not just computation but discovering non-obvious insights that unlock problems, similar to how a mathematician creates auxiliary constructions to solve geometry problems. This represents a new era where AI can contribute meaningfully to mathematical research, democratizing access to solving problems that previously required years of human effort.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
Vlad Tenev Pt. 2Added:
Tetro Grammaton.
Tetrogrammaton.
>> I think that mathematics, hard mathematics, isn't computational at all. It's almost like art. It's really about figuring out an insight, a nonobvious insight that unlocks a problem. Yeah.
And and and I think the reason that I like it is because it's it's all about figuring things out.
And I think that insight, there's a lot of it that's kind of mechanical. It's like once you have it, you just turn the crank and you get the you get the proof or you get the answer.
Yeah, that special insight can be really really magical. And to me, it's just about figuring things out. So, the reason I actually got into mathematics, if you remember, it's because of physics. When I was a kid, one of the first books that my dad sort of encouraged me to read was Stephen Hawkings Brief History of Time. And that book deals with kind of the big questions, right? What happened before the big bang? Why is the universe the way it is? you have this constant that measures the flatness of the universe, right? And it if it's like basically like too high, the universe is going to expand at an increasing rate and everything is just going to just die out in this like expansionary death. If it's too low, the universe is going to recolapse again. We would have never had galaxies and stars. And it turns out if we measure it, ours is exactly one. It's like immeasurably close to one. And this equilibrium between like death by expansion and death by contraction, which is crazy. If it's a little bit off, we would have never existed, right?
If the laws of physics weren't exactly what they are, we probably would have never formed. Yeah. Because and and also the universe has to be the exact age where one set of stars would have had to form and die and collapse and we would have been because the our bodies are made of supernova debris. So there has to have been at least we have to be second generation in a sense. So th those were the questions that kind of motivated me like where did we come from? Why are the laws of physics the way they are? Is there one fundamental force? Like we know that electricity and magnetism we we found out that those were one force. Radioactive decay then we figured out was part of that. Then we figured out the strong force. Then the next one is well if you follow that train all of the forces must have been one to begin with. So there must have just been like one thing. What is that?
That's what really I think motivated me just figuring that stuff out. I think one of the things that I often think about from physics is the double slit experiment.
>> Particle or the wave.
>> Yeah.
Yeah. the the double slit experiment is something that just sort of simple to understand, but I never really have been able to come up with like actually the the explanations that I can think of for it are kind of scary to me.
>> Scary in what way?
So the double slit experiment basically shows that by nature of us observing something or it even being possible to be observed, its properties change. So I think if you if you study quantum mechanics, which which I did, they sort of say, well, everything's consistent, right? The math just sort of explains it. There's no spooky retrocausality where you go back in time and change things, but it's just that the observer and the thing being observed, the subject are part of one system and the observer can kind of impact the subject.
Yeah. And so what that makes me think of is is this some kind of like lazy evaluation, right? In programming, you have this concept of lazy evaluation.
>> I don't know what that is. So if you have a computer program, that computer program has a bunch of memory that it can access. But to be efficient, you don't want to claim all the memory right away. So when the program gets to a particular point, then it claims the memory only when it's needed. Let's say your program was simulation of the earth and you were just in the middle of the Pacific Ocean. It would only load sea. So you would only see blue ocean around you and it wouldn't load America yet. But then maybe you get closer and okay, now you're within 15 miles, you see some coastline on the horizon. then America appears in the program and that memory is is claimed.
So that's kind of how people program computers because it's much more efficient. I don't want to claim all the memory for the whole earth if you're only here and you only need all of this circle around you to experience the program. So yeah, what I think of in the back of my mind when I think of the double slit experiment is does the universe work that way? Yeah.
Right. And is it is it okay, we're being efficient here, and if nobody's actually observing these particles, they're just a probabilistic haze that just looks like a big wave. And then only only when you look at them do they exist.
>> And I think now it's even changed to where you can predict in advance whether it's going to be a wave or a particle. It doesn't have to be in time.
It can be pre >> Yeah. Yeah. It's called the delayed choice quantum eraser.
>> Yeah.
>> Yeah. Where they actually create two particles, right? One of them goes through the standard path and the other one goes on a very very long journey where they can erase the history. They've done this actually over in the Canary Islands, a big multi- island quantum eraser. Yeah. Yeah. And they they kind of end up at the same place. If it's possible through data analysis after the fact, >> to reverse engineer which way the thing went, then there will be no interference. But if you erase it and you make it impossible, then the interference shows up again. It's pretty crazy.
>> It's crazy. There's so much we don't understand. Do you think things like that are solvable or No, >> I think it's worth trying. Yeah, I think so.
>> Yeah, >> I mean that's really the >> the point of physics and why I'm interested in AI models, too.
>> I think they could be solvable.
>> Are you into puzzles?
>> I like to play around with puzzles from time to time. Yeah, but >> math scratches the puzzle itch for you, >> I think. So, I mean, puzzles, they're a good way to pass time for sure.
>> Definitely. But I don't think I think it's different than true understanding. It's almost like a game.
>> Yeah.
>> How um mathematically interested are you?
>> I'm interested, but I don't know anything about it.
>> I can tell you my favorite um simple math problem.
I think it illustrates this sort of difference between math being sort of like a computational thing and more artistic conceptual So this is a triangle right we call it ABC the three vertices. So let's say it's isosesles triangle which means that these two sides have the same length. So one kind of interesting fact about isocesles triangles is that it turns out because these two sides have the same length these angles are also the same.
So what that means the the way you kind of write that is AB equals BC implies that angle B A equals angle BCA.
>> Isn't it >> right? And if if you look at this I think the interesting thing is that this is kind of a self-contained fact. You've got these three points. The only thing you're using in this statement is those points. You have AB equals BC. So B A C which is this angle is equal to BCA.
>> Mhm.
>> But then the question is how do you prove it? Like why is this true? And the reason it's true requires you to do something that I can this is like the there's one creative step here and you actually have to construct a new point from scratch.
So you say okay I can create a a point in between A and C and I'll call it D and this point is the midpoint right it's exactly I can create a midpoint if these two points aren't the same there's a line connecting them that has some length and if I have the midpoint I can connect it to B and then I get two triangles triang triangle A B D and triangle B D C. Now these triangles are actually the same triangle because by nature of this being the midpoint, this side is the same as this side, right?
>> Mhm.
>> And then they're sharing this side. And so you have two triangles that are congruent. They have the same sides. And what that means is that all of the angles are the same, too. This angle is the same as that angle.
>> Mhm. This angle is the same as that angle. And there you get the fact that these two angles are the same. And this problem, by the way, is not just difficult for some humans, but it was somewhat difficult for AIs, too. And it's because the first versions of AI geometry solvers would kind of work with the points. And I'd say we have three points. You just have A, B, and BC. and you're not using anything else for the statement. So where does this point D come from? And that's kind of the the genius of it. You have to know that by creating this new point which is a midpoint, >> you can test >> you can you can kind of like create this new additional structure that helps you unlock it. Even though this point D doesn't actually appear >> Mhm.
>> in the statement anywhere, right?
There's no D in here or D in there. So that's called um an auxiliary construction. And the first AI models that did really well at solving geometry problems, the only thing they did was figure out what the right auxiliary constructions to make would be and then the rest of it is just all computational.
I mean obviously this this is a pretty simple example >> in the abstract. the idea of adding a new piece of information that unlocks the picture even though that new piece of information isn't really part of the equation.
>> Yeah, >> that's beautiful.
>> Yeah, I think there's a certain elegance to it for sure. And I think there's crazy things in physics that still blow my mind that nobody can explain. Yeah. And if I had to start another company, I mean, I think I think Harmonic scratches the itch. There's like a very deep well of if you can just keep going and and figuring out all this stuff, I think it I I think if you just keep going forever, but there's also things in the real world that we can't understand and we can't explain. And I think that hearkens back to figuring out why the laws of the universe are the way they are.
>> Yeah.
Tell me about your spiritual life.
>> Yeah, I'm I'm trying to to figure that out.
I think a lot of my daytoday is sort of figuring things out. I mean, the way that this kind of I think you asked math and and physics, how do they influence what I do? I mean, a lot of what I do is really thinking of products, figuring out how people will use those products to have them have as much success as possible. Like what can get someone to really value using Robin Hood? And rolling out these products, combining them together just gives you a so many things to learn, right? There are a lot of things that we don't that that we don't understand. I think a big part of me is just being curious about it and sometimes just letting randomness take over. So, I know you're into astrology. I have astrologer friends and I listen to them.
If they tell me it's a bad day to launch a product >> Yeah.
>> or that it's going to be a bad day for me to initiate something, I try not to initiate anything on that day.
>> Yeah. And and I think that there's it it just shocks me that as a rational person, there's a lot of times where they've just been totally right and I can't explain it. But >> And you think it goes beyond superstition? It's something else.
>> I think so. I think it it it goes back to something in the physical world that we don't understand.
>> Mhm.
>> Like why do people have out-of body experiences? Nobody can explain it. I've seen all the cases where someone's literally clinically dead, no heart rate, and then they just remember everything that's happening to them in that state.
Yeah. So I I make my my decisions assuming that I have agency and of course I I can control things but I think I also am somewhat superstitious in the sense that even if I can't explain some something if someone has really high conviction I'll go with it.
>> Tell me the whole GameStop story and for the story take Robin Hood out of it. I know that there's a Robin Hood component, but that's tangential to that whole story.
>> So, I I think the the news story of GameStop started with this idea of retail investors banding together on Robin Hood to like take down the hedge funds, right? Solely on Robin Hood.
>> I mean, by and large on Robin Hood. I think the other brokers had some of the spillover like when we had to shut down the buying particularly at that point people were like okay I want to buy GameStop and they they started downloading other apps as alternatives but I think it started off being Robin Hood being the the lion share and certainly we were the ones associated with the positives and then later the negatives of shutting it down.
>> Still take Robin Hood out of it for a minute. So there were retail buyers >> trying to take down hedge funds.
>> Explain that.
>> I'd say there were a couple of things going on at the time. Some of which were COVID related. Right. One of the things was I think this nostalgia and sympathy for companies that were kind of unfairly harmed by CO. So GameStop and AMC probably the the two biggest examples, but there were also others like Nokia. A few months earlier, there was a big thing about Hertz and American Airlines, right? Which were you imagine people aren't renting cars, people aren't getting on planes, nobody's going to a movie theater, you're not going to a brickandmortar, you know, video game retailer to do your gaming.
And unlike the global financial crisis, the government wasn't like stepping in to help these businesses and bail them out. So I think there was a little bit of we're rooting for these businesses.
Yes.
>> I remember shopping at GameStop and buying video games and selling my old games when I was a kid. So I kind of want them to succeed.
>> So I think there was just an environment where retail had to some degree extra money from the stimulus and they were like >> around the time of the stimulus. Yeah.
Oh, >> big time.
>> So, that also plays into the story. I didn't know that.
>> Yeah. And the other thing that happened that not a lot of people realize was in 2020, Robin Hood actually had a big free stock referral program. So, anytime you joined, you would get a free stock.
Think of it as a referral incentive. If you refer a friend to open up a Robin Hood account, you'll each get one. If you refer seven people, you'll get seven, right? GameStop, I think it was a fairly low price stock at the time, was one of the more common ones. So, a lot of Robin Hood customers >> I see.
>> Which I mean, the customer base grew very quickly during 2020 had GameStop as their first stock on the platform. So, there was like an added level of you know, >> yeah, your first stock on the platform is everyone remembers to some degree their first stock, right? At least I do.
And so then in December of 2020, I started noticing on Reddit chatter about GameStop. And I didn't think that it was particularly notable at the time because you've seen chatter about so many different stocks and so many things happening. But yeah, there there was uh this guy, Deep [ __ ] Value was his handle. So later it turned out that this guy was a financial professional. He worked for some company in Massachusetts. But he started making videos, videos analyzing GameStop. And one of the things that he noticed was Ryan Cohen, who was the founder of Chewy. It's a a pet food company that did very very well. I think it sold to Amazon for a large amount of money. And that Ryan Cohen had then bought GameStop shares, right? And this was part of the fundamental analysis that he did. The the thesis was something like Ryan Bowen, smart guy, very businesssavvy. He's getting more involved in this company. He's seeing some value here. So maybe we should all be watching. And then he bought some shares and he started building up a following. And what happened was GameStop started ticking up, started building more of a fan base. And then January hit and January was a huge inflection point in retail activity. And so what happened was it it wasn't just GameStop, but GameStop was a big story.
Dogecoin, which is one of the cryptocurrencies, started going gang busters. And what happened was, you know, GameStop would tick up. Every time it ticked up, he would get more and more fans. So it created this self-referential viral loop almost where you know it would go up. So he would do a video, he would get more fans, people said, "Well, this is amazing." And then it kept going and as the price went higher and higher, it got more attention and more people would come in, more retail customers. By January 28th, it was all over the news, right? So, so the news was like helping amplify it, >> but the news was that this stock that was historically not performing very well, all of a sudden because of someone on Twitter is getting a lot of attention.
>> Yeah. He built an army, right?
>> Mhm.
>> And then I think a lot of stories sort of formed around that. Like they looked at all the people that were shorting the stock and it was hedge funds.
>> Explain that whole system. So the way short selling works is most people when they want to trade are long stock.
>> They're buying it to succeed.
>> Yeah. They want the stock to go up. So they buy it and they hold it long, right? What you can also do is to take the contrary view and you short a stock.
>> So you're betting against the stock.
>> You're betting against the stock. And the way that that works mechanically is you go and borrow the stock from someone who holds it. There's a marketplace for this. People who are long it are holding it could choose to release their stock into this marketplace and and they earn some more income for this. The short seller >> selling it or not selling it.
>> They're lending it.
>> Lending it >> and then they can call it back. They can recall it. Right? So let's say you want to short a stock and I own it. I can lend it to you through the stock lending marketplace.
>> What are the rules around that? Are there any?
>> There's a lot of rules around it. I think the most important one is that in most situations, I mean there are some exceptions. You can't sell a stock short if you haven't secured a borrow. And that's called naked shorting.
>> I don't know what that means. Explain that. So, in order to sell it short, I have to have borrowed it from someone.
>> You can't short it if you don't have it.
>> Exactly.
>> I see.
>> Yeah. And actually, there's some exceptions to this. Some people are allowed to short it without having it, but by and large for the normal person, you can't do that. So, you have to secure the borrow. And you can imagine for financial stability and things like this, you don't want to just have unconstrained shorting.
>> Yeah. The reason that rule was put in place was because that happened to some companies. Actually, I might have this wrong, but I think it might have happened to either Lehman Brothers or Bear Sterns and been a part of the global financial crisis. So, >> for that reason, what's called naked shorting is like severely clamped down.
>> But yeah, that's what shorting is. You can sell it and then at some point you have to pay back your borrow. Can you find out who's shorting stocks?
>> So, if you're a hedge fund, you publish.
>> I see.
>> You have to publish your positions and and that includes your your shorts if you're kind of past a certain level. If you're a normal retail customer, not really. And interesting thing, at that time, Robin Hood didn't permit shorting.
>> We have since rolled out the ability to short. We did that a couple months ago, but Robin Hood was long only equities platform.
So was part of the reason that GameStop stock was selling was because they were aware of the short positions against it or was that not part of the story?
>> So there's been studies about this and by and large the conclusion is the reason the GameStop stock went up was cuz more and more people were buying it. So the short covering so to speak I think was something like 10% of it or less because you can imagine this whole thing was investigated the SEC came out with a report. I think the conclusion was something like there was a big narrative around the hedges reloading their shorts and you know doing all this but by and large the reason for it was just more retail buying because it went viral on social media. Was the story for the retail buying to happen that it was against the hedge funds or No, >> the story morphed into that.
>> I see.
>> Yeah. I think in the days before >> there was an aspect to the story. Again, I don't really understand this stuff very well, but it was almost like a continuation of the Occupy movement in some way.
>> Yeah, I think it probably to some people was a continuation of the Occupy movement. Was it like an activistbased thing or was it just we like the story of GameStop so we're going to invest in GameStop or was it more than that?
>> Yeah, I mean there's there's a lot of different people involved and a lot of it's anonymous and you know not all of them want to sit for interviews. This guy uh Keith Gil was his name, Roaring Kitty or Deep [ __ ] Value which was the the first one that I remembered. Yeah, I remember Rory Kitty.
>> Yeah, he essentially disappeared, right?
Completely off the radar. And there was a movie about this.
>> Wow.
>> Dumb money. I don't know if you saw it.
>> I did not.
>> Sebastian Stan played me. So, I felt good about that. They picked a very good-looking guy to portray Vlad Tenev.
>> But yeah, there there was certainly a narrative around that, like we're an army. We're going to go take down the the hedges, right? They're coming after us. They're reloading their positions. I can't tell you how much people actually believe that. Some people probably did.
>> Well, did it have any impact on the hedge funds or No.
>> Well, a yeah, a lot of the hedge funds that were short while this thing was riding up and got out of their positions for huge losses. Some of the hedge funds failed in fact.
>> So then it was a successful >> Yeah.
>> activity.
>> Yeah. From that standpoint for sure.
Yeah. Yeah. And then, you know, the hedge funds will tell you, well, my LPs are pension funds, so pensioners lost a lot of money. So, I I can see the the argument on both sides, but certainly some hedge funds lost a lot of money for that. Yeah.
>> Okay. So, now the Robin Hood piece is that the buyers were Robin Hood customers.
>> Yes.
>> So, a lot of the transactions happened on Robin Hood.
>> Yeah.
>> But I don't feel like Robin Hood's the story. It seems like you're fairly passive in this story.
>> Yeah. I mean, we weren't doing videos and saying, "Come on, you know, we're we're coming after the we weren't the generals in this fight."
>> You were the service provider.
>> Yeah. I think there was some kind of poetic irony in the name of the product being Robin Hood, it being free and the retail investors using Robin Hood to take down the the hedge funds. But yeah, no, I think the reason people were using Robin Hood was Robin Hood had become synonymous with investing during that time period. Roaring Kitty wasn't using Robin Hood. I think he had an Morgan Stanley account. Yeah. But we we were pretty much the the platform that a lot of the activity was happening on.
>> And then did you get shut down or did you choose to shut down? What happened?
So the dramatic story was on January 27th, you know, Robin Hood was kind of a hero in a sense, like we were people were using our our platform against the man.
>> Yeah.
>> And then what happened was we basically had to pause, not even for a crazy long period of time, but for something like 12 hours buying a >> not even one day.
>> It was pretty much one trading day, but uh >> day.
>> Yeah. Not not a full 24 hours.
>> Was it clear that it would be short or No, >> it was not clear.
>> It was just shut off.
>> Yeah. Yeah. Yeah.
>> Was there an announcement or no?
>> There were some announcements, but there were very sort of corporate legal ease sort of thing.
>> And also, you don't know what's going to happen next.
>> Yeah. I mean, from my perspective as the person running the platform, there were a few things going on in my mind, right?
Number one, we had a a sizable PR issue in March of the previous year, which was at the beginning of the whole COVID shutdown, and Robin Hood was growing like crazy, where it was like a total platform outage. So, the whole platform, every stock was out for >> How long was that for?
>> It was for one full trading day at first, but then there were kind of aftershocks where there was another outage for a couple of hours. So, >> do you know what caused those >> intense load and our systems not really being being ready for it because our load probably went like 10xed in the span of a pretty short period of time.
>> So, in the back of my mind as this whole GameStop situation was unfolding and we were climbing the app store ranks at one point on January 27th we were the number one downloaded app in the country.
>> Wow.
>> Not even financial app. We were ahead of like Tik Tok and Instagram.
>> And you think that was because of the GameStop story?
>> Yeah. Yeah. Everyone was downloading Robin Hood.
>> Yeah.
>> And so in the back of my mind, my fear was >> total platform shutdown because of load, right? And I was like, okay, let's get the engineers to make sure like we were having daily war rooms, code red situation. What are all the bottlenecks?
But yeah, what what we didn't see was this clearing house deposit situation which kind of came out of nowhere.
>> Yeah. So so basically as a broker we have to put up firm capital to back the the positions that our customers make from trading. And the formulas for how much cash from our balance sheet we have to put up kind of depend on how one-sided the flow is. So typically under normal circumstances, people are buying and selling. So you kind of end up being flat. But in this situation when something's going up and all the retail are pretty much buying, it was very one-sided. These rules reared their ugly head and basically made our deposit requirements balloon. They went from sub 100 million a couple of days before GameStop to in the billions within a span of something like 24 hours. And then how did all this play out? Right? I was sleeping and at this point I was trying this thing. So you can imagine the depths of COVID. I was just online all the time, probably working 20 hours a day. My laptop was usually connected to my bed. And I was kind of going a little loony. So, I was trying this thing where I slept with my phone in another room because I just thought this is bad for my mental health. I need some kind of reset. and my my folks on the east coast who were doing like the morning open preparations were trying to get a hold of me and eventually they call my wife.
My wife does sleep with her phone next to her. So they they find out how to get to her. She wakes me up and she is obviously perturbed by this whole situation to begin with because nobody ever calls her. Like [ __ ] hasn't hit the fan to that degree. Yeah.
>> She's like, "Oh my god. Oh my god, Vlad.
There's something going on. I can't really understand it, >> but they need you.
>> They need you." So, I go, >> I pick up my phone. My phone is unusable. It's like I don't know if you've ever seen these clips of Kim Kardashian getting like uh having her notifications on and posting an Instagram post and then you see all the notifications just churning from the millions of followers. It was kind of like that, right? The phone was unusable from text messages and calls and I would have these messages like, "What the [ __ ] are you doing? How could you side with them?"
>> Wow.
>> Yeah. It was just people were upset.
They thought that we were siding with the hedge funds and we were colluding with them in some way.
>> Who decided to shut it down for that 12 hours?
>> It was the president of the brokerage of the clearing firm. And part of this is just we have a bunch of businesses and they have presidents and of course you know the presidents are responsible for all the regulatory matters and and running the business. So th this one the president you know made the call and it was a tough call for a variety of reasons right because you can kind of think of the counterfactuals what would have happened and there's some scary stuff that happens Lehman Brothers shut down they had to liquidate and these rules around capital requirements they're not optional so they can come in and shut you down >> you saw FTX a couple of years later just when the survival of the firm is is in question. I think the most important thing turns out to be you have to protect all the people's money, right? And kind of became a question of all right, there's a lot of people on the platform that aren't trading GameStop and there's some people that are trading GameStop for sure and they're upset. But >> was the whole system turned off or just GameStop trading?
>> It was GameStop but also a bunch of the other meme stocks.
>> I see.
>> There were handful of others. AMC and then I wake up and I'm hearing the story of basically sheets came in and a lot of this was automated of you can think of this as like a tax bill coming in with $3 billion >> and then we can't get on the phone with anyone to explain it to us and then another one comes in and it's lower $ 1.7 billion and >> so was it not clear to you what was going on at the time?
>> It was not clear. My first thought was actually maybe the central clearing system has some issue.
>> Like maybe they're just sometimes when things are overloaded just >> bogus numbers are being sent out.
>> And then there was just this this latent fear of if it's 3 billion today and let's say we managed to come up with this money and and we did. Knock on wood, right?
>> We came up with it.
>> What if it just keeps going? Yeah. Could it be 50 billion tomorrow?
>> Because the day before it was 100 million.
>> Yeah. It had gone up by over an order of magnitude. So yeah, we we made the tough call to the president made the tough call to shut it down.
>> Y >> and then what we did was we went out, we broke up into a couple of work streams.
We had the engineering one, making sure the systems were stable. Then we spun up a money raising one of we don't know exactly how much money we need, but it's pretty clear we need some more money because we don't know what's going to happen in the future. And then there was the public relations one of okay, I figured out how to make some calls and get through my phone, but the Robin Hood brand and what we stood for was kind of under attack. So we had to like communicate. And at the time the story was that we were colluding with the hedge funds. There was a narrative that the White House had called me to shut down GameStop trading. And so there was this fake collusion narrative that we felt like we had to dispel. Nobody knew about this deposit requirements thing.
Yeah. Yeah. And if you think about it, it's very technical and hard to explain and less compelling as a narrative than Robin Hood is turning their back on the poor siding with the rich and taking the poor's money. Hedge fund collusion simple delightful narrative. This like clearing house market plumbing thing.
Very hard to understand.
>> Were you able to solve that the next day when it got turned back on?
>> No. We thought we would be.
>> Yeah.
>> Yeah. And and from my perspective, you could imagine fresh in my mind was the full platform outage of March of the past year. And I was kind of like, what's the big deal here? At least we didn't shut down the platform to some degree. I was like, wow, the engineering held up and we were able to serve all the buy and hold customers. But no, you know, once this collusion narrative >> took off, it was like wildfire. I see.
>> And I gave some very bad press interviews.
>> Yeah. and and my instinct was look at all this noise on social media. We have to get out there right away. We have to communicate with the customers. We have to tell people we're not colluding. And so the first messages were like highly scrubbed generic legal ease messages like we did not do this at the direction of any other market participant or market center or we didn't want to say Citadel or hedge funds or anything. So there was these generic things and then of course the reaction was of course they're colluding. Look at this [ __ ] They sound just like what someone colluding and doing some stuff would say.
>> Hostage videos.
>> Exactly. Yeah. They're telling us they're not colluding. So obviously they're colluding and and then I did these interviews with the press with Andrew Ross Sorcin where I just kind of had my talking points and I was reading them and you know.
>> Were you scared? I mean, I wasn't scared. I was actually like, I want to go out there. This is my time to lead, you know? This is this is my time to to show what I what I'm made of as a leader. Because I'd been kind of a reclusive, not reclusive, but I hadn't really been out there very much. Kind of just building behind a keyboard and and not doing a lot of talking. So, I wanted to go out there. And actually after the interviews finished, I ended up thinking like, I thought that went pretty well, you know, felt like I got my points across. I read them, you know, throughout throughout the years since then.
>> I keep thinking, was there some magic collection of words that I could have said or interview I could have given that would have taken the temperature down and would have solved that issue.
and help the brand out cuz it's it's been a gradual rebuild in the years since then of the brand and the and and the public perception around that largely it's just time has healed the wounds people kind of move on >> but yeah I still play that back if those interviews had gone better and I'd given like a super charismatic response and said hey this was a really bad situation we weren't ready for it we didn't have $3 billion at the time Yeah, >> we didn't really know. But yeah, I'm not sure. Probably to some degree maybe it could have helped 20%. I don't know if those magic words existed. I think we were just in a really tough spot. I think the moment I picked up my phone and it was just buzzing with all the hate texts, it was probably too late to do much about it.
>> Do you feel like now it's completely in the past and the brand has recovered?
>> I think the brand recovered. There's still people that bring it up for sure though.
>> Wow. How many years ago was it?
>> Five years ago.
>> Yeah.
>> Yeah. So I It's not completely gone. We look at this thing called net promoter score.
>> What's that?
>> So it's basically a measure of product loyalty. You ask people how likely are you recommended something or there's another version of it like if this product disappeared today, how unhappy would you be? And if they say seven out of if they say 8 out of 10, 9 out of 10, or 10 out of 10, they're considered a promoter and they're like 100 points. If it's six or under, they're a detractor.
And if it's seven or eight, they're neutral. So it's a little bit of a contrived measure, but directionally it's it's pretty useful. And so NPS of 100, everyone loves you. NPS of negative 100, everyone hates you. NPS of zero, you've got as many promoters and detractors.
So, ours plummeted on that day.
>> What was it before? Do you remember?
>> It was pretty high. I don't know if we've ever shared it, but I mean, it was up there with some of the best companies and it actually >> Yeah. I mean, it it crashed. Yeah.
>> And so, we were able to measure this old adage that trust is slowly earned but can be lost very very quickly. We saw that in in the number. It just went down. Yeah.
>> And then it took many years, probably three years. 2024 was when it got back to like pregametop levels. And it was a combination of I think GameStop decreasing in importance and us improving the rest of it. But yeah, even today you see some people that that mention it.
>> You didn't make any mistakes. You were asleep. Well, >> there was no way you could have even been thinking that this is something that could happen.
>> No, I mean, yeah, of course, you could second guess and you could say, should we have been better prepared?
>> Prepared for what? Prepared for an earthquake >> or prepared for a fire?
>> How can you be prepared for any possible event? Yeah, I did think by and large we were in firefighter mode from the beginning and it was sort of like, okay, this thing happened. Let's clean it up.
I think I think a lot of the armchair quarterbacks were like, well, they should have been better prepared. It should have been capitalized with a lot of money. I was very happy that our system stayed up from an engineering standpoint. That was the known risk.
>> But yeah, if I had to look back, I think my interviews left something to be desired.
>> Mhm. It was funny like I did my press, everyone was watching those interviews.
I did one with Andrew Cuomo on CNN who I followed a lot during COVID because he did like all these COVID commentary.
That one's viewed millions of times. And yeah, when I look at it, it was not very confidence inspiring. You know, I could have done a little better on the communication, >> but it was also the first time you're ever in that position.
>> It's like trial by fire, right?
>> Yeah. And then I did a clubhouse with Elon Musk, which Elon Musk interviewed me.
>> Great.
>> About what happened with GameStop. And I think that one went better.
>> Yeah.
>> Cuz that's the one where I I shared the details, the $3 billion capital call, the NSCC, the DTCC, all the all the players. And yeah, but that happened on Sunday. And I think what happened really was the the whole thing was over by Sunday.
because the market closed on Friday and then there was the weekend and the story shifted.
>> Yeah.
>> And I think people lost interest. So what what actually ended up I think there are some people that blame Robin Hood for ending the whole GameStop thing, but I think the reality is it probably would have ended anyway. What really ended it was the weekend. And these types of things, these like short squeezes, >> they're pretty episodic. And usually if something like spikes up a lot, it doesn't end up at like a permanent plateau of high price. Throughout history, a big spike up, usually followed by a big crash, >> especially considering it wasn't based on the quality of the GameStop business.
It wasn't Tesla shipping their new model. It wasn't that.
>> Yeah. I think certainly people criticized the fundamentals of it and and I think we got some blame for both sides, right? We got blame from the GameStop traders, the retail army for doing the position closing only trades and then also from the people that weren't doing it that were like, "Oh, Robin Hood facilitates this GameStop stuff and we don't like that." So you could imagine we were sort of sandwiched by both sides. And when I went to testify in front of Congress, it was like very very stark and it was mostly from the people that didn't believe that the GameStop activity was healthy market activity and they were just asking us about is your platform responsible for the whole thing to begin with.
>> Yeah. It's like did you instigate it?
>> Yeah. And I was like well it was all over the news. It was all on social media. the trading didn't originate like the roaring kitty was using Erade but everyone I think what I realized is when something like this happens people look for someone to blame.
>> Yeah.
>> And I think uh unfortunately and through lack of savvy we kind of became the scapegoat.
>> Tell me about your note takingaking habits.
Yeah, this is something that I started kind of recently and I am not as regular as I should be, but every night I kind of like to wind down and just take notes about my day. Just kind of like what's going through my mind, what bothers me.
And then I also like to look ahead to tomorrow and just think of like the the top three things that I need to get done. Usually the list is like 10 or 15, but I really focus on the top three.
>> Do you think of it as journaling?
>> Yeah, but it's actually it's a little bit more. It it always rears into >> to-do list.
>> To-do lists a little bit. Yeah. So, it's kind of like half journaling, half to-do lists. But I also try to do like on Sunday, I'll do it for the week.
At the beginning of the month, I'll do it for the whole month. And then I'll I'll also do it for the year. So, it's kind of like December, end of December, if I'm at home, I'll just think through my year and what what I want the year to be. And I kind of say, "All right, if if it's December 29th, 2016, and I'm kind of looking back and what what would it take me to feel good about what just happened?"
>> Do you ever look back?
>> Yeah, >> you do.
>> Yeah.
>> Great.
>> I look back. Yeah. And that's actually some of the most fun stuff. And I notice, wow, I'm thinking about a lot of the same things.
>> Yeah.
>> Yeah. Basically, I'm thinking about half the things that I was thinking about a year ago. So, those are kind of unresolved. and maybe that tells me something. And then there's just a whole bunch of new stuff.
>> Mhm. When you're thinking about a new idea, do you write about it?
>> That's changed a little bit. A lot of my ideas that I have are software ideas, >> but there's this new product or new feature, new piece of software that I could build. Lately, I've just been building them.
>> Oh, great.
>> And that that's been kind of fun. Yeah.
Or I go on my chalkboard and I sketch stuff out. That's like if I really want to get my hands dirty and feel it. Yeah.
>> Do you build stuff coding in a traditional way or vibe coding?
>> Vibe coding.
>> Mhm.
>> I think now it's just >> that's coding now.
>> Now that's coding. Yeah.
>> I remember when I first started coding.
Yeah. It always blew my mind how the vast majority of software engineers didn't really know the internals of the systems. like they didn't learn assembly language or memory management or pointers. And I think there's two ways to look at it, right? Some people would say, well, that's a great thing because it democratized coding. You didn't have to a lot of people don't want to deal with all that stuff. They get bored of it and so it opened up coding to a lot more people. But I always thought you have to really understand all the stuff to really know what you're doing and to get the most out of it. I think now you're seeing it again. So I I still think it's important to understand what the code does and how it works, but the vast majority of people that are coding now probably didn't code a year ago.
>> Yeah.
>> You know about the other business I started, right?
>> No. Tell me.
>> I was a mathematician or an aspiring mathematician.
So a couple years ago, I found >> You have a PhD in math. You're not an aspiring mathematician.
>> Well, I dropped out.
>> So I just left with my masters after one year. Yeah. So, basically, Harmonics goal is to build what we call mathematical super intelligence, which is an AI that can solve math problems at a superhuman level.
>> How is that different than what an LLM does?
>> Well, LLMs can certainly solve math problems, but when we started this, which was 2023, they made mistakes with addition, right?
They would just add things and subtract things.
>> I read an article last week >> Yeah.
>> about that. It's not solved.
>> No, no, it still happens. But you could see a lot of the ingredients being there. Like the substrate existing for creating AI that's superhuman at solving ma mathematics, but it's just it wasn't there at the time. It's still not quite there. I think we're harmonic, which is the company has a product called Aristotle. Aristotle is very much at the frontier and the big thing that is getting in the way, the big obstacle that we have to solve is the hallucinations.
>> Mhm.
>> So you can imagine if you're writing a history essay or just doing research, hallucinations are not that big of a deal. I mean annoying, but you can generally one hallucination, one mistake doesn't destroy the entire argument. But if you have a complex mathematical theorem or something that you need to prove, it could be thousands of steps and a hallucination in one of the steps essentially invalidates the whole thing.
So the whole thing is just garbage. So in order to to solve mathematical super intelligence, you have to solve this hallucination problem and make sure that every step of reasoning is logically sound and follows from the axioms and that you can trust the conclusions. And if you do that, you solve this like subtraction being wrong problem.
>> Isn't that the case though with AI writing code?
>> It's 100% the case with AI writing code.
These things are very very connected. So the issue with vibe coding is that you just create a lot of slop code. You can think of this mountain of of code, right? And you just anytime you send another message saying, "Okay, add this feature. Fix this bug. Oh, this thing screwed up." the code just increases which I think works in some cases but it creates this problem and changes the job of the software engineer from someone writing the code to someone like trimming this bonsai tree or like dealing with this conflration of of nonsense and and clamping it down. So there is a problem with vibe coding too which is there's just hallucinations that you also have to solve. And I think the bet that we're making with with mathematical super intelligence with Aristotle is that these things actually have the same solution.
And that solution is something called formal verification.
So what formal verification is basically you write a spec. The spec is essentially you can think of as the list of properties and rules that your system has to follow. And then you formally and mathematically prove that any code that you generate follows these rules. And now formal verification existed before AI. The French have deployed it like in in a lot of places like the Paris train stations are formally verified and you can mathematically prove that the trains aren't going to hit each other and I think there's something like 30 years with no accidents. But these are humans handcrafting very artisal work of just making sure that all of the mathematical properties are proved by hand. So the bet that we're making is there's actually going to be a new paradigm of vibe coding and vibe proving where it's not just like the wild west but things are constrained by these theorems that that you have to follow. And if if I think about harmonic, the sort of like northstar goal is to solve a millennium prize worthy problem. So like one of the big problems of mathematics and I think that with with a fairly high degree of confidence, these don't get solved very often. The last one was about 20 years ago by this Russian guy Gregori Perilman. He solved the point conjecture. I think the next one very likely will be either significant AI assist or maybe even largely solved by AI. So that's kind of the long-term goal, but of course we have to break that down into into shorter term goals.
And the first accomplishment we had was to get a gold medal or gold medal level performance at the International Math Olympiad, which is like the most prestigious math competition, but sort of high school level problems.
>> Mhm.
>> Still highly creative. like these are smart high schoolers. Like I never even went to the IMO as an aspiring mathematician. I was sort of like three rungs below it. So I think even that was a big accomplishment. But then we went from that to solving like small unsolved research level problems. I think now we're getting to the point where we're used to solve and formalize more meaningful results that people actually have tried to solve by hand but couldn't. not because of lack of attention, but because they're just hard. So, I think there's this rising capability that I think will culminate into true research level math results.
>> That's really exciting.
>> Yeah. Vibe proving I call it. The fun thing is like you have these Erdos problems. Have you heard of Paul Erdos?
>> Mm-m.
>> Yeah. There's actually this picture of Terry Tao when he was a child. I think he was 7 years old sitting with an old Paul Erish I think who was probably in his 70s or 80s at that time. He was a Hungarian mathematician. He was kind of a traveling mathematician. So he would go and stay at people's houses. He would just come up with with his suitcase and be like, "I'm going to stay with you for 2 weeks. What are you working on? I'll help you with whatever math you're working on." Yeah, I >> think he was like hopped up on amphetamines the whole time. You would travel the world, stay in people's houses. And so he became the most prolific mathematician. Yeah. There's actually uh an aired number kind of like the Bacon number which is how many degrees away are you from Kevin Bacon, right? Kevin Bacon zero. If you're in a movie with Kevin Bacon, you're one. If you're in a movie with someone that was in a movie with Kevin Bacon, you're you're two. So there's a airish number as well.
>> Yeah. Paul Erish is zero. If you if you've published a paper with him, it's one. There's also an Eridish bacon number, which is you add the two. And there's actually someone with I think an Erish bacon number of two. So someone from Goodwill Hunting who had co-authored a paper with Paul Erish. So anyway, this guy uh during his whole lifetime, there's like,00 Erdish problems that have been collected, half of them unsolved.
And you know, people collect them. Now there's a library of them. So now Aristotle is being used to solve and formalize all of them. And they're just they're dropping like flies.
>> That's great.
>> Like every week you have and sometimes it's by amateurs. It's someone who's not a mathematician but they're essentially vibe proving like they're a AI savvy student and they're just making contributions to the field. And that that would have been inconceivable 5 years ago.
>> That's amazing.
>> Yeah. like both because it's hard but mostly because you have all these like crank amateur mathematicians that are like I've proved this big theorem but nobody's even going to pay attention to it right because to get a mathematician to like concentrate and most most of the time there's some mistake but you have to like really concentrate to find it but what formal verification results in is the machine can check the proof.
Yeah. So you don't have to convince anyone. If it compiles, it's guaranteed to be accurate.
>> That's great.
>> I gave a TED talk about the future of AI. Okay? And there's a lot of dumerism out there where people think nobody's going to have any jobs because the AI models are going to do everything that we can do, but they'll do it a thousand times faster and better.
I'm probably a little bit more optimistic for a couple of reasons. One is that there's so many of these things that we don't understand. Like we don't understand what's going on with the fundamental fabric of reality, which you can see with simple things like the double slit experiment. We don't know what happened at the beginning of the the universe. We don't understand what happens with out-of- body experiences.
So that tells me with a fairly high degree of confidence that there's some aspects of what we consider intelligence that are quite possibly outside of your own brain. It's probably not that hey, if you understand just how the brain works and image it and duplicate it and copy it and move it into the cloud that you can replicate what's going on here.
which makes me think, well, there's probably some aspect of our intelligence that can't be replicated by just AI models getting better and better and faster and faster. I think you'll also see it, you know, AI capability is growing, but it's not like clearly overshooting human capability. It's like we saw a big jump from generation three of the Transformers to generation 4, but then it's kind of like at some level where it's always questionable whether it's smarter than humans or not. It's like converging to some level of ambiguity of whether it's like much better than us or not rather than just it shooting far ahead and becoming incomprehensible.
And then also the fact that well if you look at our jobs and what we do probably 50 years ago they would have looked at what we do and thought that we were just having fun and we're not doing real work. We're not in a factory. We're not >> doing manual labor. So probably we would conform to our ancestors idea of technological unemployment. They would say these people aren't working. They're having meetings all day. They're on Zoom calls.
are managing companies. So yeah, I think I'm I'm a little bit more optimistic in that sense. I don't think that we're likely to see a big AI doom and gloom scenario.
>> Do you think that the influence of your parents being involved in finance impacted you?
>> Yeah, maybe in a roundabout way. When I was growing up, I was really interested in math and physics and sciences. And for a while, I just didn't really think or care about doing anything commercial in nature, right? I just wanted to make discoveries and and do cool stuff. And my impression growing up was my parents were very like professional, stable existence. It was really about having comfortable benefits, being able to provide a good life for me. It was sort of like they took this giant risk in moving the entire family to the US and it was really all about stability.
You know, I think I was rebelling against that a little bit and I just really wanted to be a a scientist or a mathematician or or some kind of inventor. that that was what I what I kind of gravitated toward.
>> Do you live in the Bay Area now? No.
>> Yeah.
>> How long has it been?
>> Uh, a while. Yeah. I mean, >> before the company >> before. So, I moved around quite a bit.
Uh, I have lived in LA as well, but I was born in Eastern Europe.
>> Mhm.
>> And then moved with my parents to the east coast of the US and then I went to Stanford >> five, I think. Yeah.
>> Do you remember life before at all?
>> Yeah, for sure.
>> Really?
>> Yeah. I mean, I don't always remember like, oh, when I was three this happened, or when I was four, but >> you remember the general feeling of it?
>> Yeah, because it was just a very big discontinuity, right? So, I can basically say, okay, well, this was all the stuff before I moved to the US and this was all the stuff after. And it it was so different that you know that bucketing makes it very easy to bucket my memories.
>> Why did your parents decide to move?
>> You know at around that time, this was in the early 90s, there was a huge transition in Eastern Europe. Basically, the Berlin Wall fell. You saw the the Iron Curtain lifted.
>> Yeah. And the late 80s in Bulgaria was a time of crazy optimism. Bulgaria was doing very very well. Population was growing. People had a pretty good standard of living. We were winning Olympic gold medals. So everything by all measures was really good. But my grandfather, I remember when he was still alive, he'd tell me the story.
He's like, I think this empire is going to collapse being the Soviet Union and all the satellite states. So, he kind of pushed my dad to strongly consider coming to America. And then an opportunity came up. So right after the Iron Curtain lifted and people from outside Bulgaria were allowed to come in to the country, this group of folks from the University of Delaware came to visit. And my dad was a professor at the time at um the economic university of Varna. Varna is like the summer capital they call it in Bulgaria. It's on the beach, so it's it's kind of nice. The the biggest industry is tourism.
What was his subject?
>> He studied tourism. That was >> He was a professor of tourism.
>> He was a professor of tourism.
>> Is there such a thing?
>> I mean, it was the biggest booming sector. So, >> wow. I've never heard that before.
>> Yeah. He was a professor of tourism and he had a business on the side which was essentially a tourism business. So, he was an entrepreneur too where they would basically take the French tourists.
There was an influx of people from France that were visiting Bulgaria and they had all these buses and they would take them around the country. which was funny cuz I went with my wife last time I was in Bulgaria was 2016 and my parents came with us. My dad was like I remember when I would take the the tourists and I would take them around and he took us through all these back alleys in some of the cities.
>> Do you think your dad still thinks of that as home?
>> I think so. Yeah. So he was 31 when he came over to the US. So, you know, his formative years, all of his friends from high school, you know, they're all still in Bulgaria. So, when he goes back, that's when he's hanging out with people, staying up late at night. You know, in America, I think it's really just the family and, you know, he's got a couple of friends from work, but Bulgaria, I think he still thinks of as home.
So, these folks from University of Delaware came. They were impressed by by my father because at the time in Bulgaria, if you were a professor of economics, there was a lot of pressure institutionally to teach like Marxist theories like Adam Smith, free markets, those were all taboo. So they had all of the Western literature locked up in the basement. It was like banned literature.
And my dad would would tell me the story about how he would sort of like sweet talk the librarian essentially and she would let him go down there and he would read all this stuff and it would make its way into his thesis.
But his thesis was about a taboo subject and so they made it hard for him to actually defend his dissertation and and finish. So they were kind of holding it up. But then everything shifted, right?
And when these folks from University of Delaware came, that weakness was like his biggest strength. They were like, "Okay, well, there's there's people here that actually aren't Marxists and and and who understand this stuff." And so, he got an opportunity to to get a scholarship to do a master's at University of Delaware. That was 1991.
He moved over there and he had to go by himself cuz he wasn't sure whether we had a future in the country, you know.
And then my mom came about a year later.
And then I came about 6 months after her in summer of 1992. Basically, I finished kindergarten in Bulgaria. I stayed with my grandparents. But yeah, I >> What do you remember about your first trip?
>> That was my first trip out of the country. The first trip on a plane.
>> Yeah. And I remember, you know, one of my memories was like my weekly phone call with my dad because back then, of course, it was very expensive to call between US and Bulgaria. So, it was once a week for basically 10 minutes we would get on.
>> And how long hadn't you seen him?
>> A year and a half. There's still pictures of me uh from when I landed at JFK. My parents took these pictures of me and I I remember my aunt brought me over. She was like a punk rock teenager at the time. She was like 17 years old.
And I was shy to like see my parents. I was like hiding behind her and kind of peeking out cuz I I'd basically forgotten them, you know.
>> Yeah. Did you go into New York?
>> Went into New York.
>> How different was that than anything you'd seen before?
>> That blew my mind. Yeah, that blew my mind. Just like I never >> speak any English.
>> No English. No. and they dropped me into school right away.
>> Wow.
>> So, it's like, >> yeah, >> here's kindergarten, right? I did kindergarten again in the US and it was just no English. So, I had to kind of learn it from scratch basically.
>> Would you say you learned it more from watching television movies or from speaking to people?
>> Speaking to people because we didn't have television.
>> Really?
>> Yeah. My parents got a television. I remember it was about six months after we moved for the second time. First we were in Delaware. Then we moved to College Park, Maryland. And my dad continued with his studies. He got a PhD, his second PhD at University of Maryland. And our first TV, which was 6 months after we moved in there, someone had thrown out this like shitty black and white TV. It was probably from like the ' 60s or ' 70s. You know those TVs with the two knobs?
>> The big knob was like it changes >> channels by the 10ens and then there was the little knob which was missing that lets you change channel by the ones.
Then we figured out okay if you get a wrench you can just like kind of adjust it and and get to channel five or something. And of course no cable. We couldn't really afford that. So was it frustrating not being able to talk to people? What was that experience like?
Yeah. So, kindergarten, I think it was pretty frustrating.
I found in an old briefcase when I visited my parents in DC a couple months ago. There was my old report card from kindergarten, right, where the teacher writes notes. And I remember January of 1993, Vladimir is adjusting to kindergarten.
He's adjusting to the new school. He's not talking to the other children yet, but he's at least not crying and he enjoys playing with these toys and drawing or whatnot. That made me kind of sad. It's like, oh, I wasn't talking to the other children for probably four or five months.
>> Yeah.
>> But yeah, I remember being frustrated because I wanted to play with a certain thing or a toy and I didn't know how to express that to the teacher and she wouldn't understand me when I tried.
That was probably my my memory from kindergarten and just seeing all the different types of kids. It was a very cosmopolitan people from University of Maryland would send their kids to the schools like the local school. So it was just smart kids, right? Professors, kids, grad students, postocs. So that kind of blew my mind and and that was really cool. How do you think English being your second language and being thrown into it that way impacted your life? Is that one of the pieces of why you are who you are?
>> I think absolutely. I think about it now in the context of having my own children, right? I have three kids who are roughly that age. You know, I have a 8-year-old, a seven-year-old, and a three-year-old. So, right around this time period. And they have everything, right? They can get the best tutors, got a chess teacher, someone teaching them piano.
Yeah. When I was that age, my parents were they they each had two jobs. So I basically they would drop me off at the bus stop in the morning. I would come back from school at 3 p.m. I'd have to walk back as a 5-year-old from the bus stop to my apartment.
At first, my parents would say, "Okay, go to the our neighbor upstairs has your key, so just go and she'll give you the key and let you in and whatnot and make sure you're all right." Then they realized, "Well, we're asking a lot of this kid. He's just already walking a mile from the bus stop and everything and navigating all this complex logistics. We might as well just have him keep the key in his backpack." So, I'd let myself in, do my homework, call my mom to check in, and then if I finished my homework, I could go play outside with my friends, and then my parents would come back at 8:00 p.m., sometimes 8:30. We'd have dinner together, and I would go to sleep, and then my dad usually would be working late into the night anyway. So, I kind of just saw him on weekends.
>> Yeah. So, it probably made it okay for you to be by yourself. and good work ethic.
>> A lot of time by myself. There was always this like threat that my parents would not be able to pay for my college, so I'd better get a scholarship because finances were were always tight. And there was always this subtle threat that our status in the country was not very secure because we were just like student visa holders, right? And we could go back to to Bulgaria at any moment.
>> Did you look back on Bulgari? cuz I'm glad I got out of there or No, >> no, because well, I did like America, but I I didn't necessarily fear it. I loved my grandparents.
>> I didn't particularly love school. I think I enjoyed school in America much more.
>> What was different about that?
>> You know, I just kept moving between different schools. I had like this one school when my parents were there and then when I moved with my grandparents, I went to a second one and then I had to go to a third one. So, I just remember moving schools and never really being able to settle in and have a group of friends.
>> In the US, oddly enough, first grade I started making friends. At the beginning of the year, my English was pretty weak, pretty rudimentary, but by the end of the year, I was actually like in advanced English, right? So, I had kind of like I started taking English and math with the people from higher grades and I had my group of friends and then I stayed in that school for 4 years. So, it was like kindergarten through third grade, which at that time was the longest I'd been at a place. So the the longer it went, the more acutely I felt that I didn't really want to go back to to Bulgaria, the more settled and kind of American I felt. And then I remember my parents 1996 was when they started doing a little bit better. My dad had finished his PhD. He had gotten a full-time job at the World Bank.
>> What was he doing at the World Bank? At first he was a consultant but then he became an economist and then he paid for my mom to get her MBA at Georgetown and then she joined the World Bank as well. So they worked there together for their entire career you know 25 years plus for my dad.
>> How much do you know about the World Bank?
>> I know a little bit about it.
>> Tell me what you know. I know very little about it. Is it private?
technically private so that the different countries own shares in the World Bank Group. At least this is how it is for the IFC, the International Finance Corporation, which is part of the World Bank Group. It's like the IFC IMF and then the World Bank itself. And then there's another one called MIGA. I forget what it stands for.
>> Is the World Bank in competition with other banks or it's sort of different than other banks? It's kind of different than other banks. So the shareholders are countries and my understanding of it at least is so the the IFC you can think of as kind of a venture capital firm. So they make investments in private enterprise in >> of governments >> on on behalf of their shareholders.
Yeah. Which tend to be governments.
>> Is it only governments or can other people invest with them?
>> I don't think others can invest. So it's only governments.
>> Yeah, I think it's only governments. So yeah, you can think of it as a venture capital firm. So my mom my mom's job was to go and like like underwrite investments in you know it would be like a chemical plant or a fertilizer company or she did a university at one point and you know they invest for a return but different than a typical venture capital structure because there's really no carry that occurs to the the partners. I think all of it goes to to the shareholders which are the countries and then there's the IMF which sort of like provides monetary assistance to countries not to private enterprise. So you know if there's a currency crisis the IMF can step in and actually provide assistance to stabilize it. But yeah, my my impression of it, big company, you know, I would go with my parents to work when I was a a kid, and I just love that they had the the fast internet. So, that's kind of where I figured out how to use the internet and play computer games. I didn't take a particular interest in what they were actually doing, but I loved being able to like use the internet and use computers at their work. That was the first time I I kind of discovered that stuff.
>> What decisions have you made over the years that were considered questionable by experienced people around you but turned out well?
>> Mhm.
>> A lot of the consequential decisions I think in in Robin Hood's history launching with zero commissions, a big criticism or question we get early on is, well, why not some commission?
>> Yeah. How about like $3 or 99 cents? You know, why does it have to be zero? Doing the credit card with 3% cash back.
That was risky. Doing crypto as a regulated financial firm. I mean, that's like you're going into the fog at that point in 2017. Crypto was not an accepted ass dangerous.
>> Prediction markets recently, big one.
Like the big criticism we get now is how is it not sports betting? Does sports betting belong next to your retirement account? If I had to sum it up, yeah, I think that yeah, in some ways it's gotten easier over time. I think making some decisions and having some proof points that they work out even if people doubt them.
>> Yeah, >> that kind of like builds up my bank of credibility and then, you know, I can lose that. Some of it fluctuates with the stock price too. I notice when the stock is doing well, people get a little bit more not differential, but I think I earn a little bit more respect even though it's a little silly to have that be tied to the stock price.
>> Do you bet on prediction markets or do you use it more for ground truth understanding?
>> I use it more as understanding and a source of information. So, I prefer to browse them.
>> Mhm.
>> But I make trades from time to time. M.
And how's that experience been? I've never done that.
>> It's been really good for me. Yeah. And and I don't have a huge track record.
>> Mhm.
>> But I've personally done well, right?
And and I think it's because I've tended to focus on things where I have some conviction and kind of some understanding.
>> Have you learned anything about Wall Street through your Robin Hood adventure?
>> You know, I don't even really think of Robin Hood as a Wall Street company.
Wall Street now is like almost a media environment. If you go to the New York Stock Exchange and you see CNBC there and there's a couple of people trading, but mostly they're trading exotics. So, it's sort of like niche products. Mostly it's where you see Jim Kramer and Mad Money and they're like broadcasting. So, it has a ceremonial purpose. I think what's happened in the past 30 years and I think now it's accelerating is technology what used to be sort of considered a separate sector the technology sector started eating everything and so the most interesting Wall Street companies aren't really on on Wall Street anymore. It's just like the the platforms are in Silicon Valley, I think, increasingly, and it's sort of this long drawn out story of people's habits are hard to change. A lot of older people up until recently were still using dialup internet, right?
>> Yeah.
>> But with every new generation, I think it becomes easier and easier.
I think Wall Street's like primacy as the center of financial activity is sort of vestigial from this old like brickandmortar in-person way of doing finance. I mean look at um 24/7 trading.
So it used to be that even now markets are open according to East Coast Wall Street working hours, right? 9:30 to 4:00 p.m. Eastern. Mhm.
>> But, you know, we rolled out 24-hour market, we unshackled that, and now it's kind of like increasingly markets open Sunday evenings when our 24-hour market session starts. And and I think eventually it'll just be indistinguishable. So, we're we're kind of getting rid of the last vestigages of being tied to East Coast, US, New York working hours.
>> Tell me about your media diet. I try to ignore most things. The things that are hard for me to ignore are just X. I post on X and I use it a lot for work.
>> Okay.
>> And then my feed is kind of this sort of combination of AI related news. Probably half of it is AI related and the other half is like financial services, politics, and Robin Hood related.
>> And tell me something you believe now that you didn't believe when you were young.
I think now I believe that I can do anything and that all it takes is the desire to want to do it.
>> I think when I was young, I very much felt like I was like subject to forces outside of my control and a relatively small amount of things that I could do were under my control.
>> Yeah. like I could do well in school, I could do my homework, get good grades, but where I lived, what country I was in, what school I went into outside of my control. And the idea of actually creating a business or building something like a physical thing that mattered felt a little bit outside of the realm of possibility. I never felt like I could build something meaningful.
Ra.
Tetra grammatin is a podcast.
Tetra grammatin is a website.
Tetrogrammatin is a whole world of knowledge.
What may fall within the sphere of tetra grammatin?
Counterculture tetra grammatin. Sacred geometry tetra grammatin. The avantgard tetra grammatin. Generative art tetra grammatin. The tarot tetra grammatin out of print music tetra grammatin.
Biodnamics tetra grammatin. Graphic design tetra grammatin mythology and magic tetra grammatin. Obscure film tetra grammatin. Beach culture. Tetra grammatin. Essoteric lectures. Tetra grammatin. Off the grid living. Tetra grammatin. Alt spirituality. Tetra grammatin. The cannon of fine objects.
Tetra grammatin. Muscle cars. Tetra grammatin. Ancient wisdom for a new age.
Upon entering, experience the artwork of the day.
Take a breath and see where you are drawn.
tetraadin.com
Related Videos
OpenHuman VS Hermes AI: Who Wins?
JulianGoldieSEO
285 views•2026-05-29
Long-Running Agents — Build an Agent That Never Forgets with Google ADK
suryakunju
142 views•2026-05-30
This computer is made from real human brain cells. And you can buy it.
Talktmsmedia
3K views•2026-05-28
BREAKING: Microsoft’s New Image Generating Model Beat Out GPT 1.5 and Nano Banana 2
aimmediahouse
122 views•2026-06-03
I Made the Same Anime Fight Scene in Every AI Video Generator
NobleGooseAnime
295 views•2026-05-30
Nvidia Bets Big On AI PCs | New Chip To Power Windows Laptops | Technology | AI Updates | N18S
cnnnews18
3K views•2026-06-01
I Tested NEW Opus 4.8 on Four Projects (Updated LLM Leaderboard)
AICodingDaily
298 views•2026-05-29
3D Platformer Update - NO CAPES
SolarLune
294 views•2026-05-30











