Vlad Tenev: What It's Like to Need $3 Billion by Morning

It was a situation which had no precedent.

A juicy falsehood is more powerful than a boring truth.

I didn't see the entire thing, but I did see the clips that I was in.

I have very fuzzy recollection of that time. You know, sometimes you talk to trauma victims or like I've heard a lot from people that have many children, including myself. I have more than one. It's like my wife says, you know, the pregnancy and the childbirth are very, very painful, but for some reason I don't remember it. And then that sort of evolutionarily gives you the signal that you should do it again. And I think we're doing a... Yeah, I think I was at this IPO roundtable at the SEC about making the IPOs great again and with Robinhood Ventures we're taking that fund public. So, we're doing another IPO and I joked that yeah, it's kind of like that. The IPO process, I remember being painful and I didn't like it, but I don't really exactly remember why and now I kind of want to do it again. But yeah, that's a big aside that has nothing to do with GameStop. Other than yeah, that was a very challenging time. It was towards the end of COVID and I felt like everyone was going a little bit crazy. They'd been cooped up at home for about a year without much human-to-human interaction. I think from a crisis management standpoint, it was very difficult to deal with because we're doing these like conference video calls with all these different stakeholders. The regulators weren't in office and so yeah, basically what happened was we got this like automated file in the middle of the night and it had big numbers on it, right? It had big numbers that kept changing and it was a situation which had no precedent and you know, we had to make a tough call to put GameStop and a bunch of other companies on position closing only, which basically meant you couldn't open up new positions, take on more risk for a period of about one day. So, it wasn't even that long. But because there was this narrative that had taken over social media, this like viral narrative that it was the retail investors taking down the hedge funds and we were kind of the tool, people wanted to put it as a good versus evil thing. And I think what was just... what would have at any other time for any other stock been kind of an innocuous risk management decision to control our internal risk turned into Robinhood's on the side of the hedge funds colluding against the retail investor. And I think the fact that the name of the company was Robinhood made this like a juicy false narrative that continued to go viral. So, I remember in the beginning of the... in the middle of the night when I woke up, my phone was basically unusable because it was like those videos you see of what happens if a Kardashian turns off do not disturb on their phone. It just is a constant buzzing thing. So that was my morning. I was like, the phone is completely unusable. I can't even get on a Zoom call because there's just random people calling me telling me to turn it back on, you know.

It's never happened before.

That's absolutely right. So, you can make the argument that we kind of started the whole thing. You know, everyone, if you were joining Robinhood in the year 2020 leading up to the whole GameStop thing, which by the way, a lot of people joined Robinhood that year, GameStop was one of the collection of free stocks that was given to customers. So, yeah, a lot of people, you know, just came in, got their free GameStop share, and maybe they weren't engaged that much, but then when they saw GameStop going up, suddenly those shares were worth a lot of money.

I mean, I think the major thing is just that Robinhood colluded with hedge funds to shut down trading. There was also another false narrative that I think is funnier. But yeah, Sequoia had to refute this. Someone put on the internet that the White House actually called Sequoia Capital, one of our venture capital investors, and got them to pressure us to shut down GameStop. So, yeah, that was a particularly funny one. But if you look on Reddit, that had like thousands of reposts. So yeah, I think that... I mean, we don't really have any business with hedge funds. So the idea that somehow a hedge fund would collude to have us shut down trading of a stock, I mean, always seems silly to me, but I think what we learned is a juicy falsehood is more powerful than a boring truth.

Yeah.

Oh, yeah. Actually I didn't talk to Daniel at that point, but I did call him when he was dealing with his Joe Rogan thing.

Oh, totally.

So yeah, I called him to pay it forward to offer whatever support I could have and hopefully that was helpful to him, but he's actually yeah, probably wiser and better than me in these things. So I don't know if he needed it. But yeah, at that time Mark Zuckerberg called me and Elon Musk called me, Mark Benioff called me and a nice positive side effect was these people probably wouldn't have cared about little Robinhood at the time and suddenly I think I got to talk with these business magnates that have built massive companies and they're giving me their perspective on the whole situation.

No, no, I thought that was probably the best media appearance of the week for sure. Not saying much because I had some bad ones, but...

I didn't see the entire thing, but I did see the clips that I was in. And...

Well, I did see the parts that I was in, which was about six minutes. And some of the rest. I actually thought, you know, I know it didn't do very well, but I found it more or less entertaining. My favorite part, which maybe wouldn't be other people's favorite part or maybe it would for some ladies there, was that my character was played by a very good-looking actor, Sebastian Stan, which, you know, I wasn't displeased about. I would have thought Adam Driver would also be good, but Sebastian Stan maybe a slightly less good actor, but probably easier on the eyes, right? And the thing that I enjoyed most is that in every single scene he was shirtless. So it was like in the kitchen grinding a smoothie talking about, you know, GameStop or in the bathroom shaving. Just the idea of me being shirtless dealing with all these complicated business situations just made me laugh a little bit.

I mean, I see what they were doing. I'm flattered, but yeah, I wasn't solving all these business problems, you know, grinding my smoothie, finishing my workout.

I think 2022 was harder in the sense that it was sort of like a gradual, slower burn. I mean, GameStop was acute and very painful and stressful for a short period of time, but once we resolved the situation, unlocked the shares of GameStop so they could start being purchased again, and you know, I basically like did my congressional hearing and did my round of different podcast appearances, the acute part of it was over. Like the acute part of GameStop was really one day. Whereas in 2022, it was a gradual shift of all of the economic trends that had been tailwinds for the business during COVID reversing rapidly into headwinds. So for example, I mean first the COVID relief stimulus checks stopped, right? And then it became clear inflation was ticking up and that was having a big impact on people's discretionary spending and investing. And then layer on to that the interest rates. The government went from a long period of rock bottom interest rates to the highest interest rates in over 30 years. They went to, you know, four or 5%. And when that happens, actually investing becomes less attractive because you can get your average 7% rate of return from the stock market after inflation, or you could get 5% just sitting in cash. So people naturally reallocate a little bit and start holding more cash and buying less stocks. And our business was first timers getting into the stock market, buying stocks. So all of the tailwinds for our business turned into headwinds. And of course, when that happened, it was obvious to the market as well. And to be fair, not just Robinhood, but our entire sector got hit hard. We went from IPOing at about a $32 billion valuation in 2021 to in 2022 we were trading at like $6 and change. So we lost 80% plus of our market value since IPO, which you know, a lot of time people were calling Robinhood a broken IPO, right? And I was getting advice that maybe I should try to figure out how to do a buyout and go private or something. So yeah, all of these things start coming up that don't actually have much to do with running your business and building products. They become kind of distracting. So yeah, that was very, very tough to navigate. And I think I took some solace in the fact that it wasn't just a Robinhood specific issue, but nonetheless, I mean, we, as an entrepreneur, yeah, it's very competitive and you really just want to win. And so I said, you know, I'm not going to be one of these people that either gives up because a lot of founders, unfortunately, we lost a lot of great founders in that time where they just left their companies, right? But so I'm not going to give up. I'm also not going to batten down the hatches and say we're just going to ride this out and hope for interest rates to go to zero in the future and just like not do anything and turn into an ostrich or a turtle. A lot of businesses were doing that. For example, the mortgage companies, a lot of the mortgage companies, if you listen to their public statements, they were like, 'Well, the market's going to improve at some point and people will want mortgages again.' We said, 'What can we give to our customers that will actually let them thrive in this particular market environment where you see high rates and cash is attractive?' And that led to the birth of or the revival of Robinhood Gold which started with how do we give customers the absolute highest yield on their uninvested cash so they can put that onto the platform. We followed that up with Robinhood Retirement which has over one and a half million accounts. I think it's the best retirement product on the market by a wide margin. We give everyone a 3% match if you're a gold member for making contributions into retirement. And so we really started to think about, okay, how can we diversify the business away from trading, but also away from being sort of like a zero-interest business that thrives in that environment, but maybe doesn't do as well when we're in a high interest rate. And so we ended up doing that. We diversified the business tremendously. And then quicker than I imagined, you know, we become a business that has 11 business lines now with over 100 million in annual revenue.

Yeah.

I don't know if I would characterize that. I still think I'm very nice. Basically what it was was I had to spend a lot of time thinking about how to fix things. And I think that when we went through COVID, there was a lot of pressure not just from what other companies were doing but also just because we were doing very well during COVID and growing very, very quickly, right? We went from I'd say end of 2019 we had 700 people and something like 200 and change million in annual revenue. To in the next year end of 2020 we had thousands of people and close to a billion in revenue. So it's like a 3x plus in growth. And you look around us and you know during COVID a lot of our contemporary companies were struggling. Airbnb, which we always to some degree grew alongside, we started at similar times, they were a little bit before. So what was swirling in the air was we've got to batten down the hatches, COVID, you know, but our business was booming and in fact what we were hearing was customer support is getting strained, our engineering systems were strained, we ended up having to hire a bunch of people during COVID just to keep up. I think we continued that hiring and became a big company very quickly. Things didn't really work very well together. A lot of the people that joined since we were remote had never met in person. And so it wasn't just that we were being nice or we were like coddling employees. I think that the inputs that led to that tremendous acceleration in the business and the headcount growth were not sustainable inputs like you only have one COVID. And so when that reversed, we actually it gave us an opportunity to rethink all the changes that we made to keep up and reset a little bit, which I think at the time was painful, but it led to a much healthier company. It's like, not to use another weightlifting analogy, but they're so good. If you want to get really, really strong, there's one way where you just like gradually build up muscle mass and keep your fat low over long periods of time. But what Robinhood did was we bulked up gigantically and also gained a lot of fat in the process and then we did like a massive leaning out. I think that works for a lot of people and you end up getting to the same place. And in retrospect, I'd say maybe it's too much to say that I would have done it differently. I think we had to adapt to what our reality was at the time. So I don't know if we had... I probably would have done things differently around the edges with the culture and thought about the values and really sort of like enshrined them earlier. You know, the values I mentioned earlier: high performance, safety always, lean and disciplined, one Robinhood. But in practice, fixing things... everybody says that but where it sort of meets difficulty is that means undoing something I've already decided and that means admitting that I was wrong and so people sort of like tend in general, not everybody, they tend to like slowly sort of like, oh, I'll undo a little bit of it and then I don't get any of the results.

Got rid of Sichuan food catering on Wednesdays and then you know, see how it goes and then you realize well maybe it's not so scary then you can do it for a serious thing and then maybe eventually you can do it for three or four things simultaneously. I'll give you an example. During COVID we introduced a lot of... it was this thing that from the beginning I just like didn't really like the idea of but a lot of people felt very strongly and there was just too many things happening but yeah we were like we had these wellness days where entire teams would like take a wellness day after working particularly hard. And at one point I was just like, 'Let's kill these wellness days.' And there was just a lot of fear. I mean, some people actually legitimately like the idea of wellness days. I was like, I don't want to be a wellness day company. We have generous PTO. Take a PTO. And this whole idea that an entire org or team would take the same day off, that seems like a problem. What if there's an issue and we need someone and the entire org is taking a day off? Doesn't make any sense. I think this was like one of those COVID era things that some companies probably still have. But yeah, anyway, and there was fear like we were taking back a perk from employees. They like their wellness days. They'll complain. Well, we took off the wellness days. There was complaining for one day and then we never heard about it again. So I think when once you start doing these things and you realize our deepest fears about the consequences were wrong and yeah, you learn some things.

Our values are high performance. So we make it clear to not just employees, but anyone applying for a job that this isn't meant to be like a cushy, chill job. Like this is for people that want to stretch and to do in one year what maybe another company would expect you to do in 10 years. So we keep a very, very high performance bar and that filters into how we compensate. Like we reward people disproportionately based on impact and we try to stay away from rewarding people on conventional things. Think about a typical company. People are paid generally proportional to the org size that they manage. And if you think about what that incentivizes, that incentivizes empire building. I want to have a big org with a big team because then according to traditional HR metrics, I'm more important. But really what we want to incentivize is the opposite. Can you have a lot of impact with the smallest possible team? So high performance is one of them. Safety always. We have a safety always value which means sure we're going to move fast. You know, sure we're holding ourselves to an absolute performance bar, but you can't use that as an excuse to cut corners or compromise on the security of customers and their money. So that's a very important one. You know, regulatory compliance is very, very important to our business as a trusted financial platform. So, we have that one. Lean and discipline. Another one I always ask, how can we do more with less? We scrutinize every dollar and every process as well. And yeah, I mean, I think that if I think about culture, I won't go through every single value, but those are the big ones. If I think about culture, it's a few things. It's how we hire the talent that we bring in, how we performance manage and reward people. And also the working environment, what's the environment like in the office? So, hiring, it's really top talent. I'd rather have a small team of the best people than a large team of mediocre people. And that goes along with the high performance, how we compensate and performance manage. We want to reward people disproportionately based on impact. And also if it's not working with someone, we want to make it as easy as possible process-wise for that person to go somewhere else, right? We don't spend too much time.

We want to make it... I mean, sometimes it's pretty obvious that you know, for whatever reason, we made a hiring mistake usually that someone's not a fit. And at that point, we want to make it as easy as possible. So yeah, if it's three weeks, doesn't make sense for us or for the person to continue. I think six months once you know it's not a fit is way too long. But of course, sometimes we've had people that have come in and for whatever reason they don't hit the ground running right away. But if we see potential in someone and they're extraordinarily good at a particular thing, you know, sometimes they actually get there and that could take six months. But yeah, generally speaking, I think what most companies suffer from is process getting in the way and actually making it very, very difficult for people to get rid of low performers. And I think we try to make that very, very easy.

Well, I should say I'm not super familiar with how every company does it, right? So, I'm sure some companies do many of these things, but yeah. Perhaps one thing that's unconventional at least in financial services is emphasis on early career people. I think from a very beginning we put our company next to Stanford University. We would spend a lot of time recruiting interns and engineers from there and that was my alma mater and Baiju, my co-founder, as well. So we would spend time actually when we were individually hiring everyone going to career fairs at the top tech universities. And you know, I still spend a lot of time with early career folks and interns and I think it's very, very good for the company because a lot of the companies in our space tend to get older as they get further along and you know, the folks working there can be more disconnected from the young people and then that puts you at risk of becoming more of a generational company in the sense that you know, Charles Schwab serves baby boomers very well. You had E-Trade coming later that was really a Gen X company, but I think they've struggled a little bit getting the younger generation excited. And I know, you know, they try very hard to get there, but I think the best way to do it is to actually make sure that the company itself has the point of view of young people and you're sitting with them, working with them, and learning from them at the same time.

You know, I always like to have some interns or early career people working on important things and I like to be around the people that are working on the key priorities as well. So, yeah, we don't just stick them in the basement and have them fetch coffee. The big thing is we actually want you to work on projects that ship to production and do meaningful work. And what I like to say is I started off as an intern at Robinhood. I mean, I didn't have any career experience. I went straight from school into becoming an entrepreneur. And so, I kind of have empathy for what it's like. And, you know, if I can succeed and become the CEO of a company from an internship, then everyone should have the ability to do that.

Yeah. So, weekly leadership meeting is a big meeting. There's a lot of people involved and I like having large groups of people involved because then everyone can sort of hear what's on everyone's mind. I don't like having a lot of one-on-ones. And some people are big one-on-one fans, but over time, my one-on-ones are basically on demand when you need something critical or some important decision needs to be made. So the leadership meetings are either important information that I want to cascade, like they're at the beginning of the week. So a lot of it is, you know, I spent some time thinking over this weekend about this or we should be moving faster on this, or we also review goals. So I like a very simple mechanism for sharing progress on goals. It's either green, yellow or red. If it's green, we don't really have to talk about it. If it's red, I think it deserves some scrutiny. So, you know, sometimes we go through the red goals and see how we can help turn it around as quickly as possible. And that involves a little bit of ceremony. You know, I have a gavel and I hit the gavel on the...

Yeah. Yeah. And yeah, you know, it's... I think that sometimes it's really nice to lighten up some extremely serious things so that people actually enjoy talking openly about, you know, goals, things that aren't going well because the assumption is everyone in the leadership team is if not exceptionally strong at least very, very strong. So usually if things aren't going well, you know, there's a good reason. It's usually not for lack of effort and sometimes having the perspective of multiple people can really help us quickly improve things and turn them around.

Yeah, I mean usually I'm involved in the most critical product launches or projects that are going on at any given time. And a lot of times we are... so we do a lot of product events. I think this year we're going to have five of various sizes and scales, but usually the product events tend to have themes. So, actually we're doing one in a couple of weeks. Well, I should probably be more explicit about the date. We're doing one on December 16th of this year. So, towards the end of the year, and it's called Yes No. So, it's an event on prediction markets and AI. And we spend a lot of time just making sure the messaging is right, the design of the event, the look and feel, of course, the products that go into it. So, yeah, there's a big component that's just making sure that the next event that we're going to do is...

Yeah. So, I, you know, present at the events and I introduce the products, not just me, but with the teams that are working on them. So the event is like a television show.

Well, thank you. It wasn't always the case. I think we have great people that think entirely about the storytelling of what we're doing, what the purpose is. I think the events themselves are a good forcing function because if you're communicating a new product to 20 million people, it really forces you to distill it into the essence of what 20 million people can understand. So I think where I tend to sometimes make mistakes in communication is I can get too jargony, too in the details because I just assume the average person watching is in the details of the business like I am, which I think is not a great assumption. So I think like really thinking about it from first principles. If you're someone who has never heard of a prediction market for example, how do you explain what a prediction market is and why it's innovative? Why it's important for society for these things to exist? We start there and then I think once you have the foundation, you try to make sure all the products plug into that foundation. You tell a coherent story. I think it's really just about storytelling and you have to spend time and have people thinking about it and you know, if I don't think about it or if the CEO doesn't...

If you don't think about it, then I don't think the story gets told. So as a founder and CEO, you have to spend a lot of time personally getting involved in that.

Well, actually, yeah. You know, when we started doing these events, that was a huge inspiration. We were like, 'Well, we should just make them exactly like the Apple events because they've perfected it.' But then, you know, now we're on our fifth event this year. We started it last year, and it's really just we don't think about the Apple events at all. I think it's almost like learning to play a new instrument, right? When you first learn to play an instrument, you know, you just have to practice and do what the other greats have composed in the past. But then once you get to a certain level, you know enough to innovate and to like break rules and to change things. So yeah, now we really try to make it different each time. Like, how can we from first principles make this event as good as possible? And we don't feel like we have to be tethered to the old Apple model, which some people just replicate. But to answer your question, no, I don't think that was necessarily a rule that Apple had for the events. Actually, in a lot of their events, they get really into the technical details of the computers and they share the megahertz of the processors and they even had bake-offs where, you know, they would load a website with a PC and they would load it with a Mac and obviously the Mac would load 10 times faster and everyone would be like, 'Oh my god.'

Well, Jobs was, I mean, he had this element and you see it throughout history in showmanship, right? Like there was a story to it. There was a presentation. There was like you were anticipating. You were like you could feel what was going on.

Yeah. And I think, you know, he was basically an artist. He was probably the greatest of all time at that particular element of it.

Yeah. I think we haven't talked too much publicly about it, but we're doing an AI event on December 16th, which is our first big event focused on AI as well as prediction markets, but AI is going to be a big part of it as well. I think that when we talk about internal operations, what we told the team from the very beginning is, look, there's two areas where we want to start with and be absolute best-in-class in our space. And that's customer service and software engineering. Because if you think about what really moves the needle, those are the big teams that have multiplicative impact. Customer service interacts with, you know, all of our customers that are having issues or need help with the platform. And traditionally that's been a big cost center for our peers and competitors because it scales with the number of customers on the platform. And so we've done so much innovation there. By and large, if you interact with customer service on Robinhood, you're interacting with our AI agents. And I think we are best-in-class on that side. On the software engineering side, you can think of it as accelerating product velocity and development. So, we've seen tremendous increases in product velocity from our engineering teams because from a very early point, we've made it as easy as possible to use every AI tool. It started with GitHub Copilot because that was the only game in town and then of course Cursor and Windsurf until that got acquired and then of course Claude Code has been very popular, a command line interface tool recently. And I think the important thing about AI adoption is you have to track the right metrics and actually look at them very carefully. And I think the great thing about software engineering and customer support is the metrics are pretty easy, right? For customer support, you have to look at AI deflection rate. What percentage of tickets that would otherwise be going to an agent are actually being fully self-served by AI? And you want to drive that up as high as possible. And for software engineering, you're looking at percentage of code commits generated by AI. But you also want to check that with, is the total engineering velocity or are the total monthly commits per engineer continuing to increase? Because you want to make sure that it's not just AI is doing a larger and larger percentage of the software engineering contributions but that in aggregate you're making more contributions and being more productive. And so I think we started with those internally and now we have it baked into the process where we almost think of it as headcount. Okay, how much compute do you need next year? Instead of this headcount, how hard have you tried to use AI agents in the workflow for every team? And I think the next frontier where we're going to see really interesting stuff is on creative and marketing. So you can imagine to create really high quality advertising collateral would have taken a lot of deep work over a long period of time by great artists. But if you can empower those artists with the best tools, ElevenLabs, Midjourney, Runway, you know, the total amount of creative can increase by 100 or a thousand x and you can actually make them much more personalized. So I think that's the next frontier that companies aren't really talking about, but I think we'll get to the point where next year you'll see a big step change.

I actually don't think anybody can do that. I think these are really hard problems to get right. Yeah, because it's not just, of course, most people don't even have metrics to track this stuff, which is, I can give them the metrics. Hopefully, they're useful to listeners. But thinking about, let's just say customer support, for example, I like to break down customer support progress in AI into three stages, or more phases than stages I should say. Phase one, phase two, and phase three. Phase one is a company puts the help center into the AI chatbot. And so the chatbot can answer any question from the help center. Most companies doing AI for customer support are in this phase, right, where it's just that's pretty easy to get going. Most people just like a better search function. Yeah. Better. Well, you know, and it's better than actually having links to the help center that you'd have to navigate, which actually I think probably still most companies aren't using AI for customer support and they're still pushing you towards just navigating a help center. We were there too, probably five years ago, 5 to 10. Phase two is all right, you're getting a little bit more sophisticated. You're not just taking the data from your help center and feeding it into the AI, but now the AI can actually go into the database. They can say, okay, I can see Shane's account. He's got, you know, $50,000 in there and he just deposited some money yesterday from Chase Bank. Okay. So, you can pull that data and you can use it to provide better contextual support, but it's still read only. Yeah. Right. So, rough numbers, let's say 90% of companies are in phase one. 90% of the ones that aren't in phase one are probably in phase two. Then you have phase three, which is okay, now I can actually do non-readonly actions. I can change stuff. Maybe I'll refund you if, you know, you're not happy with your purchase. I can refund your gold subscription fee. So that involves actually some pretty deep work, deep integration into all of the backend systems. So you can imagine the utility goes up as you get deeper, but the cost also goes up. And so actually I think very few are in phase three. That's where Robinhood's in. And yeah, it's actually not easy. And I think that most companies are vendoring this, and this is the thing that vendors are hard at because, you know, a vendor is not going to be able to save you much time in the actual process of plugging into all your systems. That's going to be work that your engineers internally are going to have to do. But I think what made it easier is from a pretty early point, we wanted to make our data easily queryable internally because we wanted to run good analytics to understand what was going on. And so once your data is queryable, it makes it so that you generally have good systems hygiene. It's, you know, in one place. It's got clean interfaces. You spend time making sure that it's correct. And I think the work that we did not really thinking about AI made it a little bit easier to plug in these models and actually run inference and build agents. And that's just the stuff we do internally in the product. We have Robinhood Cortex which is our AI model and the most visible use case of it in Robinhood right now is stock and crypto digest. So, if you go to a stock that's moved, it'll give you a real-time view of what's actually driving it. I think some of our competitors have done this as well. But I think what makes our offering especially compelling is it updates basically every minute. It updates whenever new information comes in whereas, you know, I think most companies do it once per day and it's just stale for companies in which lots of news happens.

Yes, that's right.

Sure. Yeah. And basically the company that I started a couple years ago that I'm chairman of is called Harmonic and it's building mathematical super intelligence which is AI that can solve math problems at a level exceeding humans, the best human mathematicians. And it's not quite at a level exceeding all of the best human mathematicians, but it certainly exceeded my level, which I didn't think it would do so quickly. But yeah, earlier this year, we actually achieved gold medal performance at the International Mathematical Olympiad, which is the world's most prestigious mathematics competition. These are like cracked high schoolers. Like five people got perfect scores in the International Mathematical Olympiad. And just to contextualize, I say high school, but I bet you I would struggle solving a single IMO problem. Yeah. Right now. So Aristotle, which is our model, solved five out of six. And the bet always was that this would actually generalize to other things. So if you're good at math as a human, I felt this because I was a math major. If you could figure out how to solve a math problem, you could solve pretty much any problem. So that could generalize to general problem solving, like be better at making business decisions, for instance. And so we've tested that actually over the past few weeks. They released a product for mathematicians and researchers and it's been used to solve unsolved math problems.

You saw the Erdős problem.

Yeah. So Paul Erdős...