From Pharma to AGI Hype, and Developing AI in Finance: Martin Shkreli’s Journey

In this episode of Gradient Dissent, Lukas Biewald talks with Martin Shkreli — the infamous "pharma bro" turned founder — about his path from hedge fund manager and pharma CEO to convicted felon and now software entrepreneur. Shkreli shares his side of the drug pricing controversy, reflects on his prison experience, and explains how he rebuilt his life and business after being "canceled." They dive deep into AI and drug discovery, where Shkreli delivers a strong critique of mainstream approaches. He also talks about his latest venture in finance software, building Godel Terminal “a Vim for traders", and why he thinks the AI hype cycle is just beginning. It's a wide-ranging and candid conversation with one of the most controversial figures in tech and biotech. Follow Martin Shkreli on Twitter https://x.com/martinshkreli?lang=en Godel Terminal: https://godelterminal.com/ Follow Weights & Biases on Twitter https://x.com/weights_biases?s=21&t=ktCZSU5uZ-QmXUJakQgoeQ https://www.linkedin.com/company/wandb Join the Weights & Biases Discord Server: https://discord.gg/CkZKRNnaf3

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:02Lukas Biewald?

  You're listening to Graded Descent, a show about making machine learning work in the real world. And I'm your host, Lukas Biewald. Today, I'm talking with a super unexpected guest. Some of my colleagues thought this was an April Fool's trick that I was even setting up this interview. It's with Martin Shkreli, who I originally became aware of, probably like a lot of you, as the pharma bro, right? This young pharmaceutical CEO who was raising prices on consumers and quickly became one of America's most hated people. I didn't really follow that for years, but then he resurfaced on Twitter and was actually talking about weights and biases and had

• 0:46Lukas Biewald?

  really thoughtful feedback on our product, had gone incredibly deep into pie torches and nerds. And I couldn't believe how technically competent he was. And then I looked at the product he was building with weights and biases. And we had a nice connection as fellow entrepreneurs. So I really wanted to have him on this podcast. And so, of course, we talk about what his experience is like getting canceled and going to jail. And I give an opportunity to talk about his version of the events. But we also go super deep on AI and drug discovery, something a lot of our guests have talked about, but he has a really interesting, different take on that. And we also go deep into his trading product

• 1:29Lukas Biewald?

  and how he built it and his entrepreneurial strategies. We also talk about what it's like to go to jail and what his experience of jail was like, and some advice that he gave Sam Beckman Friedman recently on going to jail. So this is a really wide ranging interview that I found fascinating, and I hope you enjoy it.

• 1:52Lukas Biewald?

  Martin, thanks for being on this show. I mean, you're probably the most professional live streamer that we've had on here.

• 1:59Speaker 01

  Thanks for having me.

• 2:02Lukas Biewald?

  Yeah, absolutely. You know, before you started talking about weights and biases on Twitter, you know, I really knew you as, I think, just vaguely as someone who, you know, was kind of buying pharma companies and then raising prices. And you're known as the pharma bro. I've seen you online, like, kind of correcting or like telling your side of the story. But I thought maybe I'd just offer you the chance to, I'm actually curious, honestly, to hear, you know, your take on those events before I get into the AI stuff.

• 2:31Speaker 01

  Yeah, I mean, well, obviously, you know, if you do something really noteworthy that sticks out, it sort of takes over, you know, your life and kind of how people look at you and what you do. And I think, you know, that probably happened to the hawk to a girl or, you know, any of these other people where it's like, hey, you're on front page news. This is what people will remember you forever. But before that happened, people, I think, remembered me as an investor or trader, then also as a biopharma guy. And, you know, I'd basically done everything you could do in biopharma, whether it's discover new drugs, develop drugs, license drugs, acquire drug companies. I went public at 28, my age. Not that that part's that relevant, but I had some very successful drug companies. I did

• 3:14Speaker 01

  a lot of the things that Vivek Ramswami did. In fact, he was my biggest investor. And, you know, we just kind of saw the drug industry as this very fragmented, kind of modular industry where each drug is very similar to each other. Whereas in software industry, you know, your company maybe has nothing, really nothing to do with, say, something like a rival, not a rival, but like, say, I don't know, a ServiceNow, or, you know, maybe a company like CrowdStrike or something. It's not really, they're pretty apples and oranges. Whereas pharma has like, they're very similar overlapping properties, whether it's, you know, they're all the drugs sort of work a certain way. There's a certain toxicology. There's a certain pharmacokinetics. They're very, very much members of a class, so to speak. And I think

• 3:59Speaker 01

  that, you know, we basically saw this industry as just this massive collection of class members that Was generally not well arbitrage. The individual companies would trade on the stock market, but within those individual companies, say Pfizer or Merck, they have a thousand different assets and the company's management team would have to decide, well, we'd like this one or we don't like this one. And they'd make smart bets sometimes and they make not so smart bets sometimes. And sometimes they make bets that traded off a certain utility for them that they're willing to forego for us. So I think a lot of people have since sort of gotten into that business. It's got a little trickier, but it's still this really fragmented business. So that was my main focus for a long time.

• 4:40Speaker 01

  And so I know the drug industry really well and the finance industry really well. So for me, you know, kind of doing the hedge fund world and then moving into more of a private equity pharma style treatment of the world was, was really helpful. You know, I think that a lot of traders come to Wall Street, they want to make money. But then what I think you realize over time is trading is really tedious, really difficult, really like going to tear your hair out. And you learn all these fascinating stories. You meet fascinating people like yourself and you, you start to realize that you love kind of the story, the industry, the people, the products, the, all the, all the lore, you know, is really fun. And private equity is one of these things, whether you're kind of a VC is sort of a

• 5:24Speaker 01

  form of private equity, if you ask me, but whether you're starting a company, buying the company, kind of doing a little bit of all of the above, you know, you sort of are getting your hands dirty. And I think that's for me, I realized that I was a hundred million times better at that than I was at investing or trading. So that's kind of how I got into pharma. I certainly have always looked at software as well, but we could talk about that as, as you wish.

• 5:46Lukas Biewald?

  But so wait, so you were, you, you came from originally from a more trading perspective than coming from like the deep pharma. Like you, you don't have like a biology background, right?

• 5:56Speaker 01

  No, but you know, when, when there's money on the line, you, you force yourself to learn as fast as you can. And, uh, you know, I, I was fortunate that, you know, working for Jim Cramer to start and then working for a so-called Tiger Cub, I was able to, you know, really get to talk to these management teams or the analysts or whatever on a one-to-one basis. So sometimes I feel like I learned, you know, a lot of the stuff, you know, both through hard work and my own perseverance, but also like, you know, I got to see how the drug industry worked pretty close.

• 6:25Lukas Biewald?

  And was this actually like my, my thought was like you were, you were buying existing drugs, like not like doing like science to discover new drugs. Is that, is that right?

• 6:34Speaker 01

  Yeah, no, I did a lot of science. So, you know, we, we got a lot of publications at JMedChem.

• 6:40Speaker 01

  You know, for example, it's the, the, the big sort of journal for medicinal chemistry, did a lot of crystallography, preclinical work. So I've, I've done, done it all, sort of taking a drug from an idea in my head. I remember the McDonald's where one of my business partners and I sat down and sketched out molecules for panothanate derivatives.

• 6:59Lukas Biewald?

  You would be really like sketching out a molecule on a napkin? Really?

• 7:04Speaker 01

  Yeah. Yeah. I mean, I, I, you know, I grew up as a nerd, you know, it's a, there's a sort of a self-taught nerd, but, but a very deep nerd. And panothanate is a vitamin B5. And in a certain disease, children can't make or process vitamin B5 into its active form because they're missing an enzyme called panothanate kinase. And we just said, well, if you're missing this enzyme that converts panothanate to phosphopanothanate, we could create phosphopanothanate. And we're able to sort of get the basic idea of it down. Of course, I worked with a lot of brilliant chemists. One of my chemists worked for, you know, famous Nobel Prize winner and saw the people like that. But I think, you know, probably in management, I think you maybe have a lot to say about this too. You want to know how the whole company works,

• 7:49Speaker 01

  right? I'm certainly was never the best chemist at any of the companies I started. And I certainly wasn't the best biologist. But I had you enough to ask the right questions, you know, like, you know, where, why, why won't this, for example, why is this drug, this drug looks like it's going to have very low log P? How is this drug going to be useful? And you don't have to be, you know, the perfect, you know, person at every single job, but I think you should know a little bit about this. So when, when a lot of the other pharma companies criticized me, I would often point to their CEOs and say, do you, do you even know much about medicinal chemistry or, or some of these other areas that are really knee deep in the pharma industry? And so for me, whether it's software or pharma or any other industry, I really do want to know, you know, how, how the, how the sausage is made and get my hands a little bit dirty.

• 8:33Speaker 01

  So it, it didn't happen overnight. I mean, that, that took 15 years of, of pretty hard work before it started to, some of these things started to dawn on me. And again, I'm no, I'm not the perfect person for any of these exact disciplines. I, I don't push production code at our company, for example.

• 8:47Lukas Biewald?

  Well, look, I totally respect that. But I think the criticism was right. Wasn't it like, hey, you're like buying an existing drug and then raising the price? Like, is that, I don't know, is that accurate?

• 8:57Speaker 01

  Yeah, that was the criticism. I mean, I think that it's one of these situations where, you know, don't ever let the facts get in the way of a good story. The good story was like, here's this guy, he's like really unabashed, sort of champion for capitalism. And, you know, he's doing all these things and saying all these things, and you should not like him. And what's a wonderful story, you know, for a film or something. But the reality is, you know, very, very different. You know, our drug had, for example, many other drugs that were replacement goods. You know, you could have taken just simply another medicine for very little. But, you know, it would mess up the story, you know. And then there were other things like, well, I didn't take a salary at the company, or anything like that. You know, there was no like need, you know, our company wasn't profitable, it was a very

• 9:42Speaker 01

  small drug, it was actually still relatively inexpensive. Because I guess when you do the multiplication of the price, well, that seems like an awful lot. But you really have to look at the absolute price as well as the relative price. So there's still, on an absolute basis, there were drugs that were millions of dollars. So this drug was, you know, a fraction of that, you know, it was a very small fraction of that. And today, we still have drugs that are a million dollars per year, per person. And, you know, there's very limited understanding of the orphan drug industry, which is one of the best, I mean, one of the deep regrets I have, is that I think obscured some of the great work that pharma companies do in orphan drugs, which is my specialty. The panethane medicine

• 10:24Speaker 01

  I mentioned earlier, that's a disease where maybe 500 people have this disease. But it's pediatric, it's fatal, and the biology is fairly clear and simple to sort of get over the hump and make a potential cure for this disease. So to me, I mean, when you have 500 people that are your addressable market, I mean, you're gonna have to charge a million dollars a patient. You know, there's no other way. And I think a lot of people, when you put that to them, they realize that, well, would I rather have the patient, who's a child, die, or have a insurance company who's massive and rich?

• 10:57Speaker 01

  Has no problem paying a million dollars, you know, because it's only 500 people would would rather have them pay for it. And the answer is just totally obvious, you know, the the families, I'm still in touch with some of the families that that have children with this disease. So I feel like, you know, the the price increase in and of itself, we we had done this like six or seven times before. And we never had a problem with patient access. In fact, patient access tends to go up because we can now afford to sort of provide for these patients. So there was just a lot of stuff where, you know, it's this generic theme of, you know, do you are you going to use your rational mind or emotional mind? And for so many people, you know, they want to use their their emotional mind.

• 11:36Speaker 01

  And I'm really lucky that, you know, I started podcasting or webcasting and live streaming in a time when nobody did it. It was, you know, 2015 and I basically said, I I can't win, you know, I do these interviews and they they still try to make me look bad. So I'm just going to show you who I am. And I just would turn my camera on and and go at it. And a lot of people would watch that. In fact, our our most of our seed or pre seed investors, you know, you can see those on our website. It's like a lot of the guys in in tech that we all know and love. And they watched my videos and said, well, hold on, this everyone's got this guy all wrong, you know, he's he's like one of us. He sits in programs, he sits and and hacks away to spreadsheet, you know, he's he's showing us how to be entrepreneurs.

• 12:17Speaker 01

  And some of these guys ended up becoming some of the biggest entrepreneurs, you know, for example, you know, persona who co founded Rippling, Amjad, who co founded Replit, Peterson Brothers with Flexport, and these guys did really well. And in many of them, credit credit my videos for teaching them some fundamental analysis and financial statements and things like that, which you do have to sort of learn when you become a CEO. And, you know, the people who did that, they had sort of had a strong conviction that I wasn't a bad guy. Whereas the millions of people who watched the the sort of traditional media clips had a very small conviction that I was a bad guy. So the the people had a huge conviction that I was a good guy. To me, that's more important than than trying to make like every single person love you because, you know, it's sort of hard to do.

• 13:00Speaker 01

  And the people who do love you or do think you're you're a good person, they'll vouch for you and that will tend to spread, you know. It through network effects, you know, I'm really lucky that, you know, people like the one, the folks I mentioned and so many more have vouched for me and through network effects have sort of said, no, this isn't, you don't, you know, this guy isn't as bad as, as they say, as a fact, he's really great. And I'm just super lucky that I've been able to do that because oftentimes, you know, when someone's radioactive, they tend to just sort of get, you know, cast aside, their pariah, there's no way to redeem them. And I feel like, you know, we also happen to live in environments, there's a bit of luck here, where in general, people are sort of taking second looks at people and saying, well, I don't know if I should believe that.

• 13:43Speaker 01

  And, and that has good effects and bad effects because, you know, there's, there's folks that, you know, or things in general that, you know, we should be skeptical of or something like that. And, you know, say ivermectin or something like that, um, you know, ivermectin isn't really a great treatment for anything. But ultimately, because there's this skepticism authority, whether it's in the form of government or, or in the form of media or whatever, folks are almost like becoming so contrarian that they're saying, well, in fact, I take ivermectin every day. I'm gonna, I'm gonna show you, you know, and it's sort of a funny, um, sort of dynamic, but I benefit from that dynamic. And, you know, I, I'm glad that some people, you know, have, have sort of taken a second look at me and after the media stuff and said, you know, I, I rather like this guy.

• 14:27Lukas Biewald?

  I also think that like it creates this world where it's sort of like admitting like any like nuance or gray area is like taken as this like immediate sign of, um, you know, guilt. And you get this response of like, hey, like I just completely invalidate the other side. Like, you know, in your case, like I'm sitting outside of it and, you know, I'm a capitalist, like I understand, like, you know, you want to make like a profitable, um, you know, drug so that you like serve a group, especially a group where it's like a, you know, rare disease and they may not be getting treatments anywhere else. But I can also like imagine it's like being involved in a scenario where someone like can't afford a lifesaving drug just seems like so horrible. Like, did that like actually happen or it just never did?

• 15:06Speaker 01

  I'm a humanist. I would never let that happen. You know, I didn't get into the drug business only to make money. It, it, there's also this glory of, as I said earlier, with the panethyne disease. There's this glory of saying, I'm going to use this God-given ability, whether it's just a little bit of knowledge, a little bit of privilege, whatever it is, and I'm going to use it to do cool things. And writing code for me as a kid was awesome. I loved it. But the chance to make a medicine was like 100 times cooler. Because this is the same way code, you write code and it does something. You make a molecule and it does something. And it's a little bit more difficult to program. It's a little more abstract. But nevertheless, your body is kind of the compiler. You get the outcome.

• 15:50Speaker 01

  So designing a drug is an awesome joy. And it sucks that this price increase, which again, I still defend to the death, would obscure that. So I feel like it's a tough place to be when no one wants to hear your points. And I think that one of the things that I tried to sort of, this is, again, a lot of this is self-serving, to be clear. But one of the things that I really wanted to do with infectious disease in general, was, and this was in 2015, was to try to sort of make infectious disease a more sturdy system. At the time, like, and really since time immemorial,

• 16:35Speaker 01

  the general idea of infectious disease drugs was that we would develop them after the infection showed up. It was really bad. And then it's like, okay, now we got to do something. And that's a really crazy perspective. Because if you think about it, things like coronavirus were well-known at the time. And things like, again, toxoplasmosis, this is a drug, the one of the drug I raised the price of, was about 70 years old, 7-0. And nobody had ever made a better version of it. And it's like, we should constantly be sort of like, it's not a pip and stall, but like, it's close enough to like, you want to sort of have like, a better version of this stuff? And have the version number go up? Because, for example, in the drug in question,

• 17:19Speaker 01

  the enzyme is called dihydrofolate reductase. It's a very commonly studied enzyme, DHFR. But the parasite's enzyme and the human's enzyme differ slightly. And the question is, can you can you thread the needle so that you make a drug that only blocks the parasite's enzyme, but does not block the human enzyme? That that kind of idea didn't didn't exist 70 years ago. So we made one that just blocks the the parasites and you could take gobs of it and nothing will happen to you. Whereas with the antipyrsodyc, right now, if you take gobs of it, you get very sick. So, you know, these kinds of upgrades were necessary, but the generally, the medical system, the pharmaceutical system was like, look, if it's on a fire, you know, why fix it? And as a result, you know, in general, you probably at some point read a story about how bad the antibiotics business is.

• 18:04Speaker 01

  And as a result, the drug industry stops making antibiotics. Well, now if there's antibiotic resistance, we really need that. So the problem is the lead time to make a new drug is about 10 years. So you're really talking about, you know, a really tough space with with really sort of odd diseases, you know, West Nile virus. You know, there's there's one called Chagas Disease, where there's no FDA-proof drug. Chagas Disease is actually quite common relative to something like toxoplasmosis. It's a Latin American disease mostly. And it tends to find its way to America often. And even without that, it's a very difficult disease to treat. It's a very severe disease. And no American companies or labs were ever interested in Chagas. And it's like, look, people are dying of Chagas Disease. And I don't think it was a bad thing to have a company that was going to stand there and

• 18:47Speaker 01

  say, we're going to make the infectious disease drugs nobody wants to make. And that company, again, could have could have had a coronavirus, anaviral, ready to go, day one. So infectious disease as, you know, as the world becomes more local, but also, you know, theoretically, I mean, I hate to say this, but like, you know, the AI kind of doomsday nightmare scenario, which I don't believe in personally, but the Elizier, you know, Eliza, the Yudkowsky kind of theory, the Doomer theory, that, oh, these these these AIs will let us just make print viruses, like, that'll infect the world really quickly. You know, it's probably reasonable that some some company has, you know, that capability of dealing with infectious disease really rapidly and aggressively. And so we were sort of trying to be that company at the time. And it was just a

• 19:33Speaker 01

  It was an idea that was a little bit before its time and still an idea that that isn't going to be liked. People just hate the drug industry for some reason. It's it's a really terrible thing. You know, I think people in general should love the drug industry. If you, you know, for example, cystic fibrosis has sort of now been defeated by by a great drug company in Boston and many other illnesses like Hep C is now totally curable very easily. You know, these are huge human triumphs. I mean, I I'm all for celebrating our triumphs in computer science, too. But, you know, these are really big triumphs as well in biopharma, and we still kind of treat this industry with scorn. And, you know, it is really frustrating to see that and also kind of contribute to it.

• 20:11Lukas Biewald?

  It was probably a function of, you know, when you're sort of like desperate and like need look at medication and feel like, yeah, it's expensive, right? I mean, that just, I mean, I'm sure that's like a terrible feeling. It's interesting. You know, we we do a lot of work with almost all the, you know, pharma companies. This point, we've had a whole bunch of the people on the podcast, which is why I was surprised to find, you know, tweeted viewers where you're sort of saying AI and drug discovery have nothing to do with each other, which is kind of a hot take, because actually, like, these days, I think the last couple of years, most of the pharma companies have become with and by its customers and not small ones, right? Like, they're big teams that, you know, are working on, you know, drug discovery. So I'm kind of curious, like, do you think that's like a wasted effort and impossible? Or where where you come from?

• 20:56Speaker 01

  I have a super high conviction on this, you know, so I hope every VC watches it. And, you know, I'll, I'll, I'm happy to discuss this matter further. But ultimately, in drug discovery, your your major bottleneck is idea. And it's kind of, it's a little like software, right? Your major bottleneck is what do what do I point my clinical trial, medicinal chemistry, toxicology apparatus at? You know, and and in general, ideas, there's some protein target in the body somewhere that you want to change its function or stop its function or augment it or whatever. And, you know, a great example is Dupixent, which has become one of the largest drugs ever for Sanofi in France and Regeneron in here in New York.

• 21:39Speaker 01

  It's a wonderful medicine. It's very expensive medicine, but it's it's just a game changer for for a few different diseases, asthma, eczema, et cetera. And the secret of Dupixent wasn't like anything other than just picking this molecular target.

• 21:54Speaker 01

  You know, after that, making an antibody to something is very easy. We've had the technology for thirty, forty years. Making a small molecule is very easy. This is what the chemist does. You know, can AI assist those two things? Which are about ten percent of the process, you know, the actual invention of the drug, then you have to actually test it in people. Which is the other ninety percent, but you know, those all kind of break down. So I actually did this on a livestream where I broke down the entire drug discovery process. And pinpointed the part where parts where like AI could do something in. And there were very small parts of the the the sort of system that AI could help in, and probably the biggest part is the L M part.

• 22:33Speaker 01

  You know, so so one of the things I did with with my girlfriend, who's, you know, for full disclosure, she's she works at over over at OpenAI. We we ran through, we asked G P T to come up with a good drug target. And it gave us a list that was okay, but then we said, well, let's put this in for loop and ask for every human protein. Could you turn this into a medicine? You know, could could an inhibitor or whatever it might be? Could could interact with this protein be useful in a disease context? If you actually did run this for loop and we had a stronger system, you know, G P T is great. But, you know, next generation or next generation, you actually could sort of like change the drug industry.

• 23:16Speaker 01

  But the actual like chemistry part of it, um, you know, I think it's all LARP. Like if you if you show a chemist the protein, you know, or a crystal structure. I mean, you can make a basic inhibitor and like, it's almost like, you know, that those kids that like do their arithmetic test very quickly and stuff like that, like. To a chemist, it's really not that crazy. Now, now you still have like atomic. Roadblocks where it's like, okay, I have to actually make this thing now. And then I have to go see if it works in real life. The, the, the, the systems we have to test those through, through in silicoes, we call them, they're okay. But like, you still want to know it works in real life. And then you've got to go give it to some animals, which, you know, AI is not going to help because you have the lowest form of intelligence

• 24:01Speaker 01

  that we want to give these, these drugs to, to see if, if there are side effects and things like that. So in general, like the idea that we could, you know, chemistry efforts would be helped by AI or, or, um, you know, other drug discovery components, like, like say antibody creation. You know, there have been for 20 or 30 years, different like Bayesian approaches to like get slightly better outputs. So instead of screening like 100 million molecules, you can screen 10 million molecules or something like that. But we're not gated by that, you know, to screen 100 billion molecules is very doable with today's computers and doing it serially and like brute force. And you don't get better results by screening like 100 trillion or 100 quadrillion molecules. It doesn't really help you very much.

• 24:42Speaker 01

  So the gating AI is, you know, we're here because, you know, we, we love this technology and we're, we're really interested in it. And you built one of the best companies in the space about it. But, you know, where can you actually leverage it in, in a, in a way that's, that's real and, and really going to game change? I think that's the secret, you know, there's disease, I'll just pick a random disease, but say, say Shro, Shrogin's disease. It's weird autoimmune disease. It's Swedish, I always pronounce it wrong, but it's S G O G R E N. So it's like Shrogan's or Shrog, Shrogan's. And, uh, you know, there's, there's various other diseases like this, which is no great treatment. I was developing a drug for a rare kidney disease called focal segmental glomerulosclerosis, another disease that's really tough to treat.

• 25:23Speaker 01

  And between Shorans and, and FSGS, you know, the big win here is how, why, where? How do we actually, you know, treat this disease? Now AI could help a little bit, like say in the lab. So what some people are doing is they're taking like RNA readouts and having AI sort of look through those transcripts, which was like voluminous data and to kind of difficult to like parse. Personally, I think the harder thing is to just read the medical literature. This is actually very few people knows, but this is what got me into AI. So Vivek and I had these like,

• 25:57Speaker 01

  Rival companies, but also friendly. And our theory was, was what I espoused earlier, which is like, you know, we need to go find the world's hidden gems of medicines. They could be anywhere. They could be a big drug company, small drug companies. They could be at universities, but there's one common link. Everyone publishes in a database called PubMed. And there are 38,000 papers, 38 million papers in PubMed. And so what Vivek and I did was we scaled our teams up as big as we could for humans to read these papers. And this is a time when we, we also had IBM and Watson come in and try to teach us the latest in NLP. And I started a little NLP company because I basically looked at my team and I said, I love these guys. They're a group of PhDs and other really bright people. But there was about 18 people.

• 26:40Speaker 01

  And I said, you have to look at the entire world's scholastic output on medicine. And Vivek had a team of like 80 people. So he was like scaling harder and faster. And our basic job was to go read every paper about every new medicine and try to see if we should go license it or buy it or fund it. And I said to myself, the gaining thing here is intelligence. It's human intelligence. I'll give you a really good example. There was an enzyme called BTK, Bruton's tyrosine kinase. In 1980s, there were studies around Bruton's tyrosine kinase that, or BTK, that showed that it was going to be a key drug for autoimmune disease. A key, excuse me, the protein was a key player in autoimmune disease. Only 15 or 20 years later, did somebody say, let's make a drug and let's block this enzyme.

• 27:24Speaker 01

  It turned out to be one of the greatest medicines ever. And so all that needed to happen is somebody had to read that paper and translated clinically. And I feel like today you could do that with GPT. GPT could read every paper and say, I think this, this could become a drug. And slowly but surely, if you read all 36 million of those PubMed papers and you say, okay, GPT of the top thousand, you know, one of the very, very best ones. And Pfizer and Merck could say, oh, maybe we should, maybe we should make this drug. And I feel like that is kind of the real area where, where we can win. The, the human intelligence part is, is really the gating item of all these things. And remember, clinical trials are kind of like 50 to 70% of the cost of these things. And this is just like People to sit there and wait and see what happens with their medicine.

• 28:08Speaker 01

  So like, there's just a finite amount of help that AI can have to begin with. And then there's regulatory time, which is probably not going to be helped either. So you've got, you know, there's animal time where you have to do the same exact stuff that you do with people and animals. So there's like very little narrow parts of drug discovery that could even be, AI could even be applicable. And even in those parts, it feels like, you know, there, there isn't a lot to do. So to me, when I think of AI, I, I mix in, in drug discovery, I don't think about the transformer L L M side, which, which I think is the, the side that I think about where it could be useful. I think what most people think about is like, can you make a foundation model for chemistry or something like that? And I, I feel like that's not, you know, sort of going to be too fruitful.

• 28:51Lukas Biewald?

  It's interesting though. I mean, you know, I think everyone would acknowledge that most of the expense is in the clinical trials. But like, kind of setting that aside, it does seem like, you know, there are these lots and lots of smart people, you know, at big companies and also at like Isomorphic and Citro and, you know, and have, have built these big organizations. They've kind of convinced investors this, this is important. Like, I mean, can you like steel man their argument a little bit for me? I'm, I'm trying to like channel them because that's been more of the, the, I mean, what, what's like happening here? Like, like, yeah, surely like, where's the difference in their thinking and, and, and yours?

• 29:26Speaker 01

  Yeah. I mean, I think they think that drug discovery is harder than I do. I feel like they think that target ID is something that they can either do better at. Cause I've seen some of these sort of stacks, so to speak of, of some of these companies. And some of them are correctly kind of like pointing to, you know, um, we're trying to look at, say, patient data, combine that with biomedical data and then, you know, just sort of try to take all this data, throw it a wall and see, and see what happens. But the best medicines that have ever been made tend to be like this laser focus of this is a big medical problem. Say Alzheimer's and where and how do we figure out how to treat Alzheimer's?

• 30:09Speaker 01

  Nobody's figured it out yet. Well, there are some companies that say, let's take, um, Blood samples or gene samples from, from, you know, a hundred thousand people with Alzheimer's. Well, that's a really good idea. It doesn't require AI. You know, it's plain old machine learning to like, I have a K means clustering or something to like, look at the data and say, you know, this is, this is where the bad stuff's happening. And sometimes you don't even need, you know, even advanced math for that. I mean, you just can see that, oh, APOE is a protective allele for, for Alzheimer's. Maybe we should make an APOE mimetic or something like that. And, you know, I think that the most of the ideas in drug discovery don't require these hidden features, these representation learning that like, you know, we've been taught like AI can see in, in ways that humans can't.

• 30:54Speaker 01

  If anything, like you almost want to be able to, you kind of need to be able to see it for it to work. And so much of medicine is fairly transparent, like Huntington's disease is a good example. We know exactly what causes Huntington's disease. It's a, it's a malformation. It's a so-called PolyQ repeat of this Huntington and protein. And, you know, it's a terrible, terrible phenotype. I mean, it's, it's a lethal illness. The problem isn't in, in, uh, the problem here is like, how do you actually make the drug? You know, the, the target is well known. It's one of these like very frustrating diseases because you know exactly what you need to do, but it's very tricky to, to do. So example, CRISPR in general is sort of like a curative idea for all genetic diseases. The problem with CRISPR is you can't deliver it to target cells. So CRISPR won't get into the brain cells, for example.

• 31:38Speaker 01

  It'll get into liver cells quite well, but not to brain cells. So the technology required there is, is, is you have to sort of make a better screwdriver. And could AI help there? I mean, again, I think like to the extent we get super intelligence, you know, it might be able to say, okay, what you're all missing is that this is the way to, to sneak CRISPR inside of brain cells. But, you know, so much of this is serendipity and so much of this is stuff that, um, you know, is so hard to plan. Uh, you know, innovation in general is kind of hard to force. And I feel like You know, these companies, some of them that you mentioned, they're all good people. They're all really smart people, but but we've seen this movie before in biotech. And it's often it's actually very funny. So when the human genome was decoded, there were there were there's a company called Human Genome Sciences.

• 32:23Speaker 01

  And they said, well, we're going to be the winner here because genomics is a really big revolution.

• 32:28Speaker 01

  They weren't the winner.

• 32:30Speaker 01

  When push came to shove, they had to make medicines. And when that part, as you know, from like the VC game, once things sort of get real or it's like, okay, you have revenue now. Now we have to like grow the revenue and grow the earnings. That's when like all the imaginary stuff kind of gets thrown away. And in the drug business, it's the same thing. It's like, once you put a drug into people, you now have to like, the drug has to work and it has to change people's lives for the better. Or your your company is worthless. And so I think so many of those companies are going to cross into that chasm. Where it's like, okay, all the talk's over. You know, what medicines do you have and what do they do? And that's just a biotech company. That's just a drug company. And I feel like there's almost never been like a strategic advantage in technology and medicine.

• 33:11Speaker 01

  So the strategic advantage is almost always like, what's your idea? So antibodies, when antibodies came around the scene, within a few years, everyone had the technology. And to this day, you know, antibodies are a modality for making a drug. But it's not an end in and of itself. And AI is the same way. I could use, you know, some sort of AI techniques to filter drug candidates for some disease. But if that disease isn't severe enough to merit a drug at all, there are some diseases that are like so borderline in severity that you know, they're probably not worth making a drug. Or if that this drug is not blocking that enzyme, isn't really going to do much for that disease. The AI is being misplaced and not used in the right way.

• 33:53Speaker 01

  So you still need that human judgment to sort of think about what target am I going after? And that's where I really think you could have like a GPT pharmaceuticals that that actually kind of like made decisions on. Let's make a drug that targets this disease, say Huntington's or asthma or something like that. And here are the next steps, dear humans. And I think that would probably work way better than trying to take like machine learning or deep learning. And apply them to actual like wet lab stuff or like, you know, drug discovery stuff, which is like still this like sort of relatively easy and mundane mundane process that doesn't need that extra intelligence. I think the extra intelligence, it's sorry, I guess the theme of the point here is that the tail's wagging the dog. The intelligence goes in the front end and everything that comes downstream from there is actually fairly

• 34:38Speaker 01

  easy to do and deterministic almost. But the front end decision of we're making an antibody to, you know, this enzyme, W A N D B is it's this terrible enzyme. And if we don't inhibit it, you know, it's going to, you know, cause cancer. And, you know, that's the kind of strategy that has to be delineated first. And then, you know, hopefully the drug works and does great things.

• 35:01Lukas Biewald?

  So is that why? I mean, you're not really working in this industry anymore, right?

• 35:04Speaker 01

  Not really, no.

• 35:05Lukas Biewald?

  What I guess, why, why not? Like, it seems like you're super passionate about it. And is it like wanting to try something different? Or like what, or feels like a dead end?

• 35:15Speaker 01

  I have one regulatory barrier to it, which, which I do think could be sort of over common if necessary. But I also think there's a few other things. I mean, one, software is just awesome. You know, it's just nothing like it, you know, it's, it's so great to, to write software and, and watch people use your product and watch it grow. The next thing I'd say is like, the drug industry is very capital intensive. It's very, like, for example, like, I, I, I was sort of when I, when I started my first drug company, I really liked it. An outstanding job of, of holding on to the equity of the company. And I got to hold like 10% of the equity at the end. For most software people, they, they think that's crazy. You know, that's

• 35:56Speaker 01

  a really low amount for, for a software company. But most biotech people are really jealous of me. They're like, how'd you get to own that much? You know, that's insane. Meanwhile, you have companies self-funding and owning 100%. So to me, like, it's a financially kind of

• 36:13Speaker 01

  Strange business, you know, you get a lot of scale, you know, packaged in a drug company, but you can get a lot of scale in software too, you know, and you don't have to take the dilution. So as an investor, you know, SaaS has just been a miracle, you know, this didn't used to be this way, by the way, like, you know, in the enterprise software days in the 90s, you know, you couldn't just sort of open up an enterprise software company. It was really difficult and it was a huge effort. Now, you know, everything's changed and it's changing by the day. So I feel like the reward is pretty big there. Again, pharma is one of these places where you can make the next, like, Vertex literally made the cure for cystic fibrosis and Bernie Sanders said, "Why is it so expensive?" It's like, wow, we just did a miracle, you know, we're saving 30,000 people's lives that would normally die. I had a friend die of cystic fibrosis.

• 36:58Speaker 01

  And, you know, the first question is, you know, how dare we reward these people for doing this amazing technical achievement? And it's funny is that, you know, you can be a great AI researcher and you know this way better than I do, and you can plop yourself in San Francisco, make a million dollars a year relatively easily. Maybe stretching that a little bit, but, you know, whereas in pharma, you know, couldn't require just the same amount of brain power and you make far less. And, you know, your entrepreneurial sort of goals are limited just to do a mouse experiment. In pharma, it can cost a million dollars. And, you know, meanwhile, there are guys that like reached PMF and profitability with that. So I feel like, you know, it's just a better business, it's more fun, you know, et cetera.

• 37:42Speaker 01

  I wouldn't rule out kind of return to pharma. But, you know, there's a lot of hoops that I have to sort of jump through legally, one. And then two, you know, I'd sort of have to get my head around, like, is this worth doing? I've thought about doing it certainly in Europe where, like, for example, that panethionate disease, I still have like these residual ideas for drugs that I think could, could work really well. And doing them in Europe and then sort of backing into the US,

• 38:08Speaker 01

  where I'm literally banned from the pharmaceutical industry, like one of the first people ever banned from an industry. It's, it's, I feel like lawfare incarnate.

• 38:18Speaker 01

  But, you know, I still could probably persist, whether it's with the Trump administration or through Europe or something like that. So I would never say never, but it's, it's, it's kind of one of these industries that's also been, been doing really poorly, especially relative to some of my software. Because that target ID problem we were talking about for the last hour. That is such a difficult problem that when there is a good target, every company has one. So when the PD1 miracle happened with oncology, Bristol-Myers and Merck both had this PD1 antibody that changed the game for cancer, KEYTRUDA and UPDIVA. And within years, virtually every single drug company had a PD1 inhibitor. Because there just weren't that many great targets to go seek. And even the ones that didn't work, there were a bunch of them like TIGID and some of these

• 39:02Speaker 01

  other LAG3, all these other oncology targets that kind of got hot because of the mouse study or something. Before you know it, like eight companies already in clinical trials of LAG3 inhibitor. So it's this vicious competition to find a good target and prosecute and develop it to the point where you have medicine. And the capacity problem, it's kind of like the steel industry or something like there's too much capacity in biopharma and there's, there's too much money in it and not enough like actual novel kind of targets. And again, maybe I'll be proven wrong and some great AI tool can, can actually find new targets, but you know, there's only 40 drugs FDA approved every year. It's like very small amount and a lot of those are like Tylenol XR, like something that's like just very like not innovative.

• 39:43Speaker 01

  And so there's really like 20 ish kind of drugs that are truly interesting advances. Some of them are not that great. And, and so you're looking at a very small number of drugs that are actually moving the needle for people. And the amount of people that can benefit from those financially, well, Big Pharma wants eight or nine of those or, you know, sort of 15 or 16 of those 20. And then like really big venture slash PE wants, and then public biotech wants the rest. And so for somebody to just say, I'm going to do this, I'm going to make it work. You know, it's a 10 year, 10 year odyssey to, to get one of these drugs approved and hundreds of millions of dollars. So it's, it's, it's, it's sort of a business that I fell into, you know, as a hedge fund guy, studying these stocks. And then I said, you know, I could kind of do this myself too.

• 40:24Speaker 01

  And, you know, it's certainly not a bad business if you can, if you have access to like huge amounts of capital and lots of really creative ideas. But even then, like, you know, it's just not an easy industry either. I mean, you're dealing with a lot of stochastic noise, you know, you do a clinical trial, like ketamine was, was a really good example. We were one of the first companies chasing ketamine, us and Johnson and Johnson. And the data looked so good coming out of Yale University for depression, for really serious depression. And we thought that the phase threes would be like, piece of cake. It turns out they weren't a piece of cake. Ketamine has a great effect. It's better than SSRIs. But it wasn't the miracle that, you know, Yale sort of pointed it out to be and got us all excited. It was, it was a little bit south of that, a little bit north of SSRIs.

• 41:07Speaker 01

  And it turns out that that development would not have been the easiest thing in the world. We ended up selling that medicine, but it was, it was one of these things where, you know, you don't always get what you want in pharma. And unlike tech where like, if you can make a change to the code base, you can't change a molecule. You know, you start from scratch. So there's a lot of differences where, you know, it, it, it ends up being, you know, really tricky and really hard to, to always get things right.

• 41:30Lukas Biewald?

  So actually I really want to ask you about how you got up to speed in AI. Because when I first talked to you, I was kind of astonished by your like technical, you know, depth in, in that field as well. But I also want to make sure, I, I mean, I, I touched on, um, so I'm sure it's like on people's minds or this is, which is, so what actually caused you to, to go to jail? And like, what, what was that experience like? I know we've talked about that a lot, but I just feel like it's, I gotta ask that.

• 41:52Speaker 01

  Yeah, it was wild. I mean, it was really wild, you know, it's probably what's like Sam Bankman's thinking right now. The guy went to math camp. He's probably the first guy ever to go to math camp and then go to federal prison.

• 42:02Lukas Biewald?

  You probably went to math camp too, right?

• 42:03Speaker 01

  I mean, I, I, I mean, I did in my entire, my entire summers were just like staying at home programming. But, uh, you know, you're fish out of water. I mean, that's for sure.

• 42:14Lukas Biewald?

  Well, first of all, what, what was the actual like crime here?

• 42:17Speaker 01

  Fraud. Yeah, so fraud. So fraud's like this very vague word. And like the more and more we, we look at Silicon Valley, I think you could probably throw that word up a lot more if you wanted to. And I feel like, you know, for me, being the bad guy, being like a world renowned bad guy, sort of, I think, helped government say, especially, you know, one of the ways I was a bad guy was not only telling the media that I didn't think that their opinion mattered and that I could tell my own story. It was also telling government the same thing. And I think eventually sort of push comes to shove, I was getting into a lot of fights with say Hillary Clinton and some other folks. And, you know, I didn't care about the party or anything like that. What I cared about was that.

• 43:00Speaker 01

  Ever since I studied economics, it is a company's right to charge whatever price it wants for its products and services. You can raise the price of your product or service tomorrow, your users might go to somewhere else. You know, and that's just something you you have to deal with. You know, you have to decide balancing, you know, the customers' needs with your needs and your employees' needs and your investors' needs and all that stuff. And you decide, you know, and you may change the price higher, you may change it lower depending on the case. And it's kind of one of the first things you learn, like from Warren Buffett, for example, who, you know, reading all of his work over the last, you know, a couple of, you know, 20 or so years, I've been studying this stuff. You know, pricing power is very important. And so for me, the idea that a company chooses

• 43:42Speaker 01

  its price and the company's owner chooses the price is really like holy to me. And, you know, I sort of got the sense that some of these politicians wanted to remove that power. And I was really like, look, you can't change that. That's what makes capitalism. That's what makes economics in America great. You know, the day that you say that you're going to look at my portfolio and you're going to decide what what what the prices are going to be. I mean, you might as well, you know, move to another country. It's over. And that idea was creeping in. And I think the politicians used me as a little bit of a foil for like, look how bad this guy is. We need to be able to have that power too. And, you know, I was frustrated by that. And I would sort of get into it with Hillary or whomever and say, look, you don't have that power. You'll never have that power.

• 44:25Speaker 01

  And, you know, I sort of like, you know, kind of like poking my nose at them, saying, you know, you can say what you want, but you're not going to change the price of my drug. You're not going to goad me into changing it either. And eventually, I think the powers that beat tend to coalesce around, and I'm going to sort of sound like a conspiracy theorist here, and I really did in 2015, but this idea of lawfare, you know, which is that here's a bad person, we're going to find a way to sort of like, mess with them. So I had a hedge fund that I started that did actually very well. It returned about four to five X capital. You know, in hedge funds, that's a lot more like venture capital. It's almost like normal. Hedge funds, that's

• 45:02Lukas Biewald?

  Not normal, I'm just saying.

• 45:04Speaker 01

  But, but, you know, it was great result. Nobody lost a dime. I didn't make any money either. But, you know,

• 45:12Speaker 01

  I fought the case really hard. It's very hard to fight these. I was acquitted of a lot of the conduct. A few of the charges did stick, which were things like I misled investors about our fund and things like that. None of the investors were upset, but it was technically a crime, according to them. I still maintain full innocence. And if you go into a courtroom, it's one of these surreal things where like, everything is stacked against you. There's no way to win. It's a 99% conviction ratio. We actually have a higher conviction ratio than Russia and China.

• 45:42Lukas Biewald?

  So what was the jail thing?

• 45:45Speaker 01

  It was interesting. I mean, it was definitely, you know, I grew up in Brooklyn. It sort of hearkened to that time of like, low income, you know, sort of people from broken homes, a lot of poverty, sort of the root of some of these problems. It sort of showed me, there's a really interesting world that like, you don't exactly get exposed to where I think people feel like they want to climb in society, but there's no ostensible path for them, whether it's through school or through whatever. I think it's a lot of mental health problems, like people with severe, like a dyscalculia. I tried to do math to folks and like, there are people that just could not, you know, add and subtract small numbers because of like,

• 46:27Speaker 01

  just a profound learning disability or something like that, probably went undiagnosed. And, you know, every story I met, every person I met, you know, one common thread was that they came from either a single parent or no parent family. And I found that to be really remarkable. I also read lots of business books and so often was the case that the business leader would end up coming from like fairly well, well-to-do family or a family that had, you know, sort of two parents and a loving household. So, you know, certainly lots of exceptions to that too. But

• 46:59Speaker 01

  in prison, there were no, almost no exceptions. I met one person that had like the two and a half kids backyard, kind of like, you know, went to college sort of story out of like thousands of people. So there seemed to be that common thread. I ordered tons of books. So like, I had a copy of the so-called wizard book, the, you know, the 6P, you know, MIT programming book, and many other like, sort of classics for nerds, you know, whether Goethe L. Scherbock and stuff like that. I would just be sitting there. No computer, so you're not allowed to have a computer there. I have a couple of funny stories, I'll keep brief, but we would smuggle in cell phones, you know, and try to sort of use those.

• 47:38Speaker 01

  And then there was this computer in MDC Brooklyn that was supposed to be used for discovery. So you'd look at evidence on it and prepare for your case. And it had a web browser.

• 47:51Speaker 01

  It wasn't connected to the internet, but I was able to get to a dev console. And I would, I was able to like change HTML elements. And there's like some guy, like some MS 13 gang member or something. And I would write his name and I would write some disparaging comment. And it would be like really huge on the screen, like font 72. I'd be like, hey buddy. I'd look, find it, and everyone would laugh because, you know, and they'd be like, how'd you do that, Martin? It's crazy. And, you know, good old JavaScript. But it's, you know, it's got to be really hard for, say, Sam Bankman. I mean, the lack of a computer is almost like corporal punishment. Like, it's almost like unconstitutional in that it's in how we are so wedded to these devices and taking people away from a computer.

• 48:36Speaker 01

  For Sam, who, before he went to jail, he sent me some messages about this. And his biggest concern was internet and computer. And I think for a lot of us, that would be one of the major concerns. And it's almost unconstitutional, in my opinion, that it's sort of cruel and unusual, you know, punishment to say, you know, you're going to get disconnected from the internet. You're going to get disconnected from all the world's knowledge. And, you know, you're just going to sit there and do it the old fashioned way, reading books, using the telephone, stuff like that.

• 49:03Lukas Biewald?

  I mean, but it seemed, I mean, did you feel like there could be benefits to that also? Like, I mean, I honestly think if someone took away my computer for a while, I might, you know, learn better, like I might reprogram my brain, I don't know.

• 49:16Speaker 01

  I talked about this with one of the who's who in the VC world. And, you know, he was joking that, you make it sound great, Martin. So it's, actually, you know, it's, I read like three or 400 books in prison. I made this list, I have to upload it at some point. But since I've been home, I maybe read one book, cover to cover.

• 49:35Lukas Biewald?

  Wow.

• 49:35Speaker 01

  And so you just don't have the time if you're running a startup or something. But sitting there, I mean, I could read books about whatever. You know, I could read a book a day sometimes. And, you know, you just have nothing else to do. And it was so enriching and awesome. But it was also, you know, you still would rather have the computer, I think. But, you know, there, you know, you could kind of make a silver lining to it at least.

• 49:57Lukas Biewald?

  Didn't you feel like angry about the situation you're in? You don't seem angry, but I don't know. Like, I would think, like, given what you told me, I would be feeling so frustrated.

• 50:08Speaker 01

  I think, I think that, you know, you know, you probably know this really well, probably maybe better than me as an entrepreneur, that you just have to roll with the punches. Like, you lose a client, you lose a key employee, you know, stuff like that happen. You go to jail.

• 50:27Speaker 01

  These kinds of things are setbacks, but they're not, you know, going to define you. And, you know, every month at our startup, we try to, like, set a new revenue record, and we try to have it, you know, a certain percentage above last month. And I don't know if we're going to make that number this month. And it's really, it just started. But, you know, I'm nervous about it already. And it's one of these things where, you know, I can only control what I get angry and nervous about, you know, and ultimately getting nervous about something that was like this cascading, the fact of like, I was one of the last people who was like canceled, so to speak, before like cancellation just stopped because people just stopped believing in it. And I was one of the last people to sort of get skewered by the media before the media started to fall apart.

• 51:12Speaker 01

  And I feel like, you know, whether or not those are sort of copium and like, you know, just sort of my own excuse for it, I still have to live my life, you know? And getting frustrated and angry is just sort of not going to be very useful or healthy for me or anyone around me.

• 51:27Lukas Biewald?

  Well, let me calibrate. Okay, so I spent the last 15 years of my life worrying about like revenue numbers and other KPIs and like losing sleep over it, right? So I like totally understand that. So if that's like a five on a scale of one to 10 of anxiety, like, where is like thinking about like going to jail in your future? I would imagine that would be a lot higher.

• 51:46Speaker 01

  Yeah, it's like an eight or a nine. I mean, I, you know, my father passed away a year or so ago, and that was like a 10 to me. You know, it was actually harder than prison. Prison is, you know, there's an old saying in criminal justice, they're like, you only do two days, your first and your last. And I do think that the sort of Damocles is actually worse than the actual prison time. Now, I also got seven years, which was, you sort of do around five of that. So for Sam, you know, he's, he's got 25 years. You know, we don't know what they would have given, I think maybe they did sentence Andrew. I don't remember, but. You know, you are rolling the dice with the judge and the judge doesn't like you. They can give you 30 years and nothing you can do about it.

• 52:25Lukas Biewald?

  So did you actually give Sam advice about, did he call you?

• 52:29Speaker 01

  I did. You know, I don't, you know, I kind of feel for him, you know? A lot of people don't like him and just think, you know, he's, he's human garbage. But ultimately, you know, I think he should get punished, but I don't think, you know, 30 years was the right number.

• 52:48Speaker 01

  He sort of operated a financial institution like the same way like JP Morgan or, you know, look at that bank, Silicon Valley Bank that went under. You know, financial institutions are a little bit wild. He also, he sort of helped create a topic in a way, right? So like, you know, this, you know, I don't know that he was like this terrible mastermind criminal.

• 53:12Lukas Biewald?

  So like, what did you, what did you tell him?

• 53:14Speaker 01

  I just sort of, you know, to be explicit, I mean, I basically said that prison is a very different world from the world you grew up in. And it's a lot of minorities. There's a lot of people that just from different walks of life that you've never even thought of. And you just have to sort of respect that there are people out there that are just going to like different things, do different things, and you just have to roll with the punches. And it's a lot of cultures that you're not familiar with. You know, I have a passing familiarity with some of it from, because where I grew up, but he really grew up with silver spoon. He just basically has to, you know, get rid of that and sort of, you know, become a more malleable person. But I think that he is fairly like dedicated to the person he is and like, he's going

• 53:59Speaker 01

  to nerd out through prison through from start to finish. And I, I don't know that that's like, I don't think he should become a gang member, but I also think he has to be flexible because you want to be comfortable in prison and you have to communicate with people and you have to be friendly and charismatic. And I don't know that he's like perfect at that, you know? He was sort of like, he had a lot of anti charisma, you know, tended to, tended to put his foot in his mouth quite a bit. So I don't know, I wasn't the best, but I, but I also think like, you know,

• 54:32Speaker 01

  I, I just feel really bad for the people that get super long sentences. I think anybody on planet earth could do a year or two in jail. You know, I think that when you get 25 or 30 years, I, you know, it's, it's, it's got to be soul crushing.

• 54:47Lukas Biewald?

  All right. Well, I guess switching back to AI, do you want to talk about the company that you've that you built?

• 54:53Speaker 01

  Yes, I mean, I started sort of doing some AI work and I ended up pivoting back to sort of more trad software. We tried to make TTS stuff. We tried to do

• 55:05Speaker 01

  different stuff in LLM world and we couldn't find a great fit, you know.

• 55:12Speaker 01

  The TTS world was really awesome. I really loved it, but it was also like I was always sitting there trying to convince others, but also myself that this would be a big market. And, you know, I remember the nuance acquisition that Microsoft did. And, you know, it's a really big market, but ultimately they succeeded not by making TTS technology or speech to text ASR. They succeeded by kind of packaging with a product, especially targeted towards like doctors and stuff like that. So to me, like this is age old question of like, do you want to, you know, build the technology? Do you want to build a product? And what layer do you want to be in? And I think like, you know, it'd be really fascinating to see how companies like Eleven Labs progress.

• 55:54Speaker 01

  But one of the problems that I constantly thought about, again, trying to convince other people is that we're going to have an audio sort of web. And that, you know, maybe not every part of the web is going to be audio driven, but even a little bit more, you know, is going to be a big deal because the LLM output, people don't want to read all this stuff. I almost called the company TLDR because people just don't want to read and read and read pages and pages of text. And it was, they'd rather hear it. And it's why people watch podcasts and listen to them. There's less cognitive load than sitting there reading a transcript. And I feel like, you know, I felt that after a while, I was like just convincing myself and I'd enter, I'm not even sure I believed it, you know? I think the TTS space is like a billion ish in revenue total across every company.

• 56:39Speaker 01

  And a lot of that revenue is lives at like IBM and AWS and GCP and Azure and companies that you like, you wouldn't necessarily think of as, you know, great providers. And I feel like I don't know if people are going to make money in this space. Like, I feel like it's this very like niche space. We still wanted to do it. It was a lot of fun, but I feel like, you know, It's just really super niche, even though like the Sesame AI demo I thought was really amazing. This a rebase thing and you know, I feel like, you know, there, there will be room for that. And I feel like there's, so I'm of two minds about it. Like, I feel like every content company like Disney, Disney should have, you know, a Darth Vader

• 57:21Speaker 01

  or, you know, whatever characters they have, and they should have them custom made for you and your kids and things like that. And, you know, there's no reason SpongeBob shouldn't have a dedicated episode just to you every day. And, you know, this is all doable with, with TTS and, and increasingly with other companies. But you also have the bigger companies that have better scale, like OpenAI does TTS and things like that. So I, we gave up on it just because we couldn't get real big revenue traction. And one of the funniest things, I think you'll appreciate this, is that, you know, we, we took this saying by Greg Rockman really seriously. He said that, you know, it's easier to teach good software engineers AI than it is to teach AI people good software engineering.

• 58:01Speaker 01

  We tried, but you know, it wasn't, it wasn't that easy. And, you know, we, we, we were just sitting there, you know, reading, you know, reading everything we could find, deep learning. We, we made a lot of like, you know, uh, we, we, we rewrote a bunch of Torch. We, we made a, I made a pull request, my own pull request, deep, deep speed of the Microsoft library, but like ultimately this stuff ain't easy. You know, there's a reason it's, it's a, it's, you know, it's, it's very difficult and

• 58:30Lukas Biewald?

  And it's frustrating to you, right? It's like, I feel like, you know, whenever I write software, I get the satisfaction of like, you know, shipping like every day, every week where it's like, I feel like you can usually spend like months on the good, you know, ML projects and just you're like, shit, I got nowhere and I don't know why.

• 58:46Speaker 01

  Yeah, yeah, it was, it was really challenging and we, we kind of followed this wild TTS story about this guy, James Becker, who kind of solo developed, you know, this, this soda TTS and he's over at OpenAI now, but at the time he just open sourced the soda TTS and lots of companies tried to sort of take it and, and make it a managed service. And I think, you know, the, the key there was, you know, first mover advantage was really important there, really high quality audio was, was really important. Every little mistake that a model would make could be disastrous to your reputation. And it was one of these words of mouth things where people would say,

• 59:30Speaker 01

  What's the best TTS? Oh, it's definitely this. And, you know, when you get that kind of word of mouth going, you know, you're going to do really well. But I still wonder if this market's going to be, you know, that reliable. Because one of the fears you have to have in the back of your mind is that, okay, as compute becomes cheaper and data becomes cheaper and these models become open source, what's to stop like some model in hugging face from, that's like really good and can run locally? Like, just gonna use that. Like, why do we need anything else? And speech still has like a bit of like issues with like trying to get the whole context of like the whole page or the whole book. And a lot of times people just chunk speech into sentences and pray that it has the right prosody and stuff like that.

• 1:00:12Speaker 01

  But like, it's very different if like somebody in this one sentence, this character's a bad guy. When they say the bad guy's name, it might actually inflect differently from like if he's the character's lover or something. So, you know, that kind of context is sort of missing from TTS. And but these last miles will sort of get met with. And I think like dialogue models and like really cutting edge stuff where like people are talking over each other and like really natural sounding speech, it's going to happen at it. I don't know that like there's some big customer that's going to really do well with it. My dream was like customer agents and service agents would need to be embodied with these TTS voices.

• 1:00:50Speaker 01

  We'll see how these companies do. I mean, I think it's a really cool space. So we really like that part. I also thought about doing AI for physicians, which was its own interesting thing. We see every day that, you know, knowledge workers are going to get disintermediated. We're not going to have knowledge workers anymore. Well, healthcare costs are something I know a lot about. In the news about and a big, big chunk of healthcare costs, if not the majority, are are human labor. And that labor can, can, a lot of it is like workflow charts, like flow charts and stuff like that, like hypertension. You know, if the patient is not on any medicine and they're 130 systolic, use this drug. That's not working, use this drug. That's not working, use that drug.

• 1:01:31Speaker 01

  And it's like actually fairly simple stuff, like a doctor is, you know, is important, but like you could pretty much encapsulate all medicine in an LLM. It's basically already happened. The question is like, how do you actually get that to actual real costs? And I think that that we're going to learn that like doctors have this very serious lobby that they enjoy, you know, these huge reimbursement fees. And that the idea that a machine could do it better, uh, you know, is something that's very scary to folks. We, we had a nice talk with a, a Middle Eastern country that didn't care. They said, we can cut our healthcare costs a huge amount by just forcing people to use an AI doctor. And we, there, this was a country that was willing to sort of just say, to hell with it. And I feel like eventually that's going to happen in the US.

• 1:02:16Speaker 01

  You know, two thirds of this deficit, uh, or three fourths of this budget deficit that we have, uh, I don't know, maybe it's like a half is Medicare, Medicaid. And so like this gigantic deficit that's causing all these problems that could result in like US insolvency someday is just healthcare. And, you know, all that could be done by AI theoretically. And, you know, maybe someday human robotics too. But you know. Why did, why did you do it? Why are you, what, what? I think, I think like we did okay with it. It, it's felt like, it felt like, you know, getting commercial traction would be difficult. Other people have, have sort of danced around this bush. Like some people, a lot of people are taking it, um, from the clinician payer side, which I think is smart.

• 1:02:59Speaker 01

  We were trying to do it from a consumer side, um, you know, and, and get people to sort of reduce their costs and, and one day maybe get insurers to pay for it. I feel like we didn't get the, that much traction. Like we got a lot of people trying it, but we also felt at the end of the day that like, this is actually going to be the big value prop potentially for the foundation model companies that like, you know, Elon actually tweeted that, you know, I think he probably had to be pretty careful what he said, but he said, use Grok as a second opinion. And like, that's true. You know, I, I mean, it's, it's probably a better second opinion than, than the most doctors and, you know, what an, what an awesome thing. But like, do you want to, you know, compete with that? And, you know, have your own, like these dedicated LMS for specific, like.

• 1:03:43Speaker 01

  Areas, I don't, I don't know that those are gonna work. Look at, you know, cursor and stuff like that. Like, you're just piping into, you know, the foundation model L L M. So you're not making a code specific L L M. Uh, at least with the same technology or same, same sort of, um, you know, a framework or our architecture. So I don't know that like a medical L L M makes sense. Like it's just a regular L L M and, you know, that's really done well by a bunch of companies and, and, you know, trying to do that didn't make too much sense for us. So like, it was one of these things where we got usage, but not revenue and B D C in general is like, you know, something where it's really hard to get revenue. So we, we sort of settled in on this financial product that, you know, I've always had this dream of and being the lifelong trader, lifelong investor, you know, building my own sort of financial suite was, was

• 1:04:27Speaker 01

  always my dream. And in fact, how I got to get sort of all love with JavaScript. I started BASIC and Pascal and stuff like that. And, you know, I never programmed JavaScript until, um, I literally got arrested. I mean, then I turned, I, it was the darkest of moments. I turned to JavaScript and the, uh, I, I had some free time. I was gonna fight the government, prepare. And I basically said, you know, let me go back to like the original love and, and I very quickly like, you know, realized that this was awesome and I, I missed it from my youth. So I started to make some little financial programs and it evolved into this really big suite that we have now and it, and it's, it's the most fun I've had in a really long time. But in terms of A I,

• 1:05:10Speaker 01

  you know, I think the pressure to put A I on everything is, is sort of like, you know, uh, it, it, it's being seen in finance a little bit and finance software. It still hasn't really been done right. So I, I, I hope to do something right there, but we, we were almost like completely trad like software now. Like I don't think we have any torch in anything we do right now. Like at some point I do, I do think that we can um, do some cool stuff, uh, with A I, but I feel like. You know, the contrived kind of like chat with this, you know, chat with your documents or you know, whatever it it's a little bit too forced. And I do think there's a real huge use for AI in finance. It's just not so obvious. And

• 1:05:53Lukas Biewald?

  Well, I so I, you know, I was watching your your kind of live streams and you know, you gave me a super cool demo of your your product. And I guess like for people that are maybe like listeners and not traders, it like appears to me like you it's sort of like this like, um, kind of monster BI tool where you like pull up tons of graphs and like tables and stuff. And and seems like you like to have just like zillions of like, you know, metrics spread out like all across your screen. And I guess I was like, looking at that as like a non trader and thinking like, I'm not sure that's how I would want to ingest data about my own company. So like, is there something different? Like, are you just really feeding your like ADHD and anxiety as a trader by like, putting so much information there?

• 1:06:36Lukas Biewald?

  Or is it actually like, helping you trade better?

• 1:06:40Speaker 01

  It's a really good question. I I think that if you walk around the trading desk, like, it looks like that, you know, like, the people are constantly on the phone. They're like, you know, they're ADHD whether they like it or not. And I feel like sifting through lots of securities, all these things kind of are, to me, very reminiscent of a programmer's job, in that if you're going to spend eight to 10 hours a day coding, you want the exact right command line setup, the right, like, even people like, get very serious about their their Vim setup, like, every little thing's got to be perfect. And I'm not like that. But like, you know, there there are some things that I like. And if I'm going to do it for 10 hours a day, I might as well get those little things done right. And I think that

• 1:07:25Speaker 01

  financial software is very similar. If you're going to, you know, it's very well known how long bankers spend in the office and stuff like that. But any financial person spending, you know, 10 hours in the office every day, and they they are going to look at stocks or bonds, or whatever they look at, they're going to do it constantly over and over and over again. And I feel like you have to give them a command line style tool that helps them get that information quickly. And it's going to be a little bit ugly and scattered and all over the place. But that's sort of a feature instead of a

• 1:07:49Speaker 01

  Bug because if it looked like, you know, a SaaS site, it would frustrate the investor because they're like, okay, well, I just looked at Apple, but now I want to look at something else. I want it to take like a millisecond. I don't want it to take like refreshing this page or trying to hydrate the page or just, just show me the data right away. And I feel like, you know, that, that's really important to Wall Street. Now, you know, in terms of AI, there's, there's a lot of places where this could help. And I know that a lot of the quants are starting to do work with L L M's. And so we're, we're sort of trying a race to provide those tools too. But I think like the stuff that I've seen, which is like chat with a 10Q document or a 10K document, you know, it can be a little bit simple, but you know, a lot of the times like Ctrl F works, you know, and I don't know that.

• 1:08:32Lukas Biewald?

  But like, what is a trader doing? Cause like, you know, I look at like, it's like you're looking at a bunch of graphs and then trying to make some prediction. I mean, that sounds like you. How are you going to possibly be AI at that? Like, isn't it that like best case scenario for AI to, to do well?

• 1:08:47Speaker 01

  I think, I think AI trading is going to make quant trading even better and even harder for humans to, to, to do well in. I mean, I think that humans have a small edge when it comes to like judging markets, emotional states of like traders, other traders and things like that. Integrating lots of like disparate pieces of information. Like if you see a little clue about inflation from one company and then another little clue from another company, and you start to connect a narrative that like, okay, maybe inflation is going up. I'm going to do something with, with bonds or something like that. But that's the kind of thing that maybe AI can't see right away. But in terms of like super high frequency trading, the computers have dominated that. Trading the news is something that I'm really interested in, that I know a lot of other like L M type folks are interested in, whether they're coming from the quant side, pointing to the L M side or the L M side pointing to the quant side.

• 1:09:33Speaker 01

  You know, the, the question is, can you have a computer that just reads breaking news really quickly and reacts to it really quickly? And that's one of the products we're working on that I think is going to be really helpful because you'd be surprised at how much breaking news there is that hangs in limbo and the market's like, still trying to figure out if this information's real or not. The L M can make that call on the second and then make the trade. And I feel like that's going to be a big game changer. But of course. Like all things in finance, everyone will have it within like six months or a year. And you know, you can very quickly react to things like just the other day, just today. It was Hims announced that they were going to carry a GLP product, which was going to be good news for Hims.

• 1:10:14Speaker 01

  But it sort of took the market like a few minutes to really like digest it, whereas the LM would like be immediately like, yeah, buy that. And I feel like that is a product that certain big firms are making, but nobody's making it for the little guy. And I'd like to make those types of products for even, I mean, firms too, but also the little guy. And there's all kinds of other little nuances in finance that could really benefit from this. Traders themselves, I mean, they've mostly been wiped out by computers. I feel like there are still hedge funds out there with humans. They pale in comparison usually to the quant hedge funds. We're also considering, you know, how do we bring quant tools to people too?

• 1:10:53Lukas Biewald?

  But wait, are you trading? Like, it looks like on your live stream, you're trading, is that? So you yourself have not been wiped out by.

• 1:11:01Speaker 01

  No, I think there are still people that can do it. I think that it's a losing, it's a fool's errand. I mean, it's sort of like Kasparov versus Deep Blue. Like you could see the writing on the wall even years ago. And today, like, they're still good chess players, but you know, it's really hard to play, you know, Deep Blue. And I feel like you're going to have a time and place where like, say, private equity or venture capital. Like, these are places that are like, AI is not going to compete with for another 10 or 20, 30, 40, 50 years, maybe. You know, somebody's going to come up with a startup that competes with yours and say, who are these guys? Is it that guy I met at MIT or is this that guy I know from Anthropic? It's like, no, it's a fucking machine. It's an oh wow.

• 1:11:43Speaker 01

  And you know, that would be, you know, really, really ironic because you're sitting there like, let's unplug it or something, you know? But the reality is like, the markets have always been competitive. And you know, I feel like there are some people that believe they often trick themselves into thinking that they have an edge. And one of my goals is to help people see that. Signing up for Robin Hood, putting your weekly paycheck into it and like trying to guess that the stock is going to go up or down is not something that is probably something you should do. And, you know, is there a way to do it a little bit more intelligently that is less risky and maybe more rewarding is one of the things we're trying to think about. But I agree that like trying to make trading software or software for research analysts in 2025 where the research

• 1:12:28Speaker 01

  analyst has kind of become a dinosaur and a thing of the past is something that like certainly like for us, like we understand that like Wall Street's completely changed. It used to be cigar smoking, you know, blue blood sort of frat bro type environment in the 90s. And now it's mostly quants, IMO winners, stuff like that, you know, and I feel like you have to make tools for those guys. And so we are sort of trying to get there. And I think we will, but I also think like Wall Street hasn't collapsed yet, you know, there's still sort of people doing that stuff. And, you know, you also have a lot of other things you could do with this type of framework, like Bloomberg is really interesting company where like they're the time series data, data, data company. And

• 1:13:10Speaker 01

  they, they do it really well for stocks, but they kind of stopped doing it for other things. And I think that you could actually look at tons of unstructured or structured data and render it more easily and taking the unstructured data and Google and LLMs can do that better. But if you know what data you're looking at and you're looking at it every day, like we can graph that, you know, it doesn't have to be a stock or a bond. It can be, you know, weather or it could be, you know, some other shipment data or something. And so like, I hope that, you know, we can sort of show that stuff. It's working really well, you know, we have millions of dollars of revenue now. I'm really lucky. You know, it's been a, it was really hard sort of finding PMF, you know, we tried three, three different things and none of them sort of connected. The first one

• 1:13:53Speaker 01

  didn't connect at all with anything. The next one sort of connected a little bit with users, but not revenue. The next one, the TTS product connected a little bit with revenue and a little bit with users, but this, this product's just sort of, you know, Gone off the charts and I think-

• 1:14:06Lukas Biewald?

  Well, the thing that I, well, I don't know, the thing I admire about your, your product from, I guess, like one entrepreneur to another is I think very similar to, you know, with some biases, I think we were kind of building this really specialized thing for an audience who really understood and more than understood, I think we like love them, you know, like we kind of came from that, like really want to make something good. And I kind of see from your live stream, like you're doing the same thing, like, you know, you are like a user of your product, you kind of know the people using it. You have a very idiosyncratic UI that clearly comes from like a lot of like caring about it being a good product for, for your audience. Like you really feel that in your, in your product. So I just, I just really admire that from afar, although I'm absolutely not a trader.

• 1:14:45Speaker 01

  Yeah. I mean, this is what Toby said. I saw a video the other day where, where Toby Lutke said something like, you know, the only way you're going to ever make a great product is if you really care about it, you really love it. And I feel like, you know, we probably should have seen that all along because I love investing, I love trading and every little detail, you know, is something that I obsess about because I know what I would feel if somebody was selling me this software. I'd say, no, I don't want to use this anymore. If it's going to have this thing there or that thing's going to be wrong. And the developers don't always understand that. We, we actually try to make or find people, but, or encourage our, all of our, all of our team to invest. And so they really understand like, you know, maybe not more than 10 minutes a day sometimes, but at least a little bit of market activity.

• 1:15:28Speaker 01

  So they can sort of understand what the user's going through. So I feel like it's, it's a really, it's a really amazing area because traders have money. They love spending money on software and you know, it's a really kind of remarkable, awesome thing. So, you know, we, we, we sort of sat around the first year or two of our company sort of meandering and I don't, you know, I don't regret it, but you know, now that like our revenue is just, you know, going crazy and you know, we're hiring like crazy and all that stuff. I'm really lucky and happy that we're, we're sort of hitting that, but I also now have to start thinking about other things like, well, how do we 10x our revenue? And then how do we? Make products that people really, really like and really, you know, are valuable for them. You know, I think again, the stuff like Robin Hood and other companies, they, they, they help the user to gamble.

• 1:16:13Speaker 01

  And I want to help our users really actually make money. And, you know, it's something that's, you know, I'm pretty passionate about and we'll see if I can sort of, you know, get that to work. We still haven't actually released our product. So it's all been sort of in this kind of weird beta stage. So one day we'll tweet about it or put it in the Wall Street Journal or whatever. And it'll, it'll sort of hopefully 10x or 20x then. But for now, it's like, these are, this is actually very hard software to make. And um.

• 1:16:38Lukas Biewald?

  Do you want to, do you want to put a link in our notes or tell us?

• 1:16:42Speaker 01

  That's fine, you know, but you know, the, the, you know, trying to get, because I saw if perplexity was trying to do a finance thing. And I think that come off sort of like made a wisecrack about it. That, you know, just showing one or two charts doesn't, doesn't replace a Bloomberg and that I do think that's true. It's really hard to sort of like negotiate with the Frankfurt, you know, stock exchange and then Osaka and Tokyo and like hurt all these cats and get all these vendors and like write tests to make sure that if any of that data is late or something like that, you do, you do some emergency thing and like it's mind moggling how complex this machine is. And that's sort of a moat in a way. And the UI I think is where things really shine. I mean, just the, it's vim for finance in a way.

• 1:17:25Speaker 01

  And I think that's like to anybody who's stuck sitting down like in an office, just hacking away, you know, at finance. It's code in a different language. And, you know, this is the sort of like thing you need. So I'm really, I'm really lucky and happy. And again, you know, I had a lot of patience for investors to sort of like find our footing. And now, you know, we were, we're in the same grind that like everybody that gets the PMF is, which is what's the revenue? What's the growth and how fast can you grow? And, and, you know, every year is going to be, every month is going to be a challenge.

• 1:17:56Lukas Biewald?

  It's more fun searching for PMF, I'll tell you, it's better.

• 1:18:02Speaker 01

  Yeah, I think like looking at like the levels of the world, I actually tried to buy a company for levels. I called him, I sent him a DM and he said, oh, great, Martin, I love your stuff, I'm a fan. I said, that's great, give me your number, let me, let me give you a call. He said, no.

• 1:18:17Speaker 01

  I'm too, I'm too, uh, what are the words, I'm too, I'm grinding right now, or like, I'm too, I'm too locked in. I was like, I'm trying to buy your company, you know, I didn't even say anything, I was just like, alright, man, let's stay locked in. But you know, I think these solopreneurs are, are really something, I mean, I, I think that you're, you're going to see more and more of these and like, you know, we may get our first solopreneur IPO or something someday where, where somebody's like, yeah, I built this all by myself, I built a whole SaaS company, I own 100% of the equity and, you know, it's a multi-billion dollar company and I think we're going to see that's, you know, more and more, thanks to the companies like yours and others like that have built this amazing ecosystem. You can never do it for our company, by the way, so don't even try, but the, the complexity is just, you know.

• 1:18:58Lukas Biewald?

  You need at least 10 people or everybody.

• 1:19:01Speaker 01

  Exactly.

• 1:19:04Lukas Biewald?

  Okay, well, actually, where do you, I mean, one of the, I don't know, the spicy takes that I found of yours from like, I don't know, like six months ago or something was saying, you know, AI is going to kind of enter a bubble and the bubble is going to pop. Which like, you could argue that was extremely oppression, I'm not sure like where you feel like we are in like a hype cycle or if you're expecting like a bigger bubble. But it does sort of feel like we actually kind of had a, maybe like a, a peak, maybe two months ago, maybe we're coming down from there, what, what's your take?

• 1:19:30Speaker 01

  Yeah, it's hard to say, you know, I, I think that

• 1:19:34Speaker 01

  in the 90s, this happened with dot coms too, there was kind of like this, it wasn't just dot coms, but it was just, it was a mix of like dot coms and other tech, including like communications technology. And so you had this, this, this bubble and it felt like maybe you're getting to a sigmoid. But then it sort of did another big gap, I feel like that's probably going to happen here. I think this is like maybe one of the bubble, standalone bubbles and, um, you know, one of the funny things about, about it is that in every bubble, sometimes you actually sit there and think, maybe it's real, you know, and I think the smartest people after a while are like.

• 1:20:11Speaker 01

  I don't know, maybe I should do internet stocks and you know, I feel like that's, that's exactly what's gonna happen here. I think OpenAI will go public. I think it'll be a trillion dollar market cap. I think that, you know, the bubble will get bigger. You know, I feel like it's possible it'll all come stumbling down tomorrow. But I also feel like almost more than the internet, like AI has immediate impacts for a lot of organizations. And the other key thing is that the biggest beneficiary is gonna be companies like Verizon and Procter and Gamble and stuff like that where they're, they're gonna be able to like, save a billion dollars here and there, you know, which is a huge amount for the S and P 500, where they can save some money.

• 1:20:54Speaker 01

  You know, deploying AI tools and maybe they can spend money using products like yours to sort of say, what, what, what can we do with the data that we do have internally? Really smart companies will, will, will pass those savings on to efforts like that. But I think that the internet did that for so many different S and P 500 companies. It's what fueled global growth. And so AI has been fueling global growth and will continue to, you know, make GDP growth go up and productivity go up. And I feel like, you know, that is, you know, that's when a bubble can actually continue to inflate. I mean, I feel like, you know, when, once you have, you know, OpenAI is a lot of revenue, you know, in the dot com era, nobody had revenue. So you, you have actual progress, you have actual product, you have like amazing technological advances.

• 1:21:39Speaker 01

  You know, this is sort of the real deal. You know, the, the big question is, you know, what do investors do about it? And I feel like investors are somewhat sober. You know, you have some of these valuations that look a little bit large. But by and large, like, you know, folks aren't funding, you know, just as a company has, you know, AI in the name, you know, back then, if you had a dot com in your name, you'd go public or instantly you'd take it public. Today is, it's not that easy, you know, it's, it's still, you know, the big firms are not just funding anything. They're being fairly judicious about, you know, where they put money to work, whether it's Sequoia or Founders Fund or A16Z. And I, and I think that. When you start seeing people get really sloppy, it's kind of like when the bubble has really hit.

• 1:22:24Speaker 01

  And I feel like, once those firms started like to do like really dumb stuff, which they tend not to do. But even the smartest people sort of get wrapped up into this. You know, I, I think then we'll, we'll sort of know the bubble's really here and gonna blow up. But for now, I mean, it, it's sort of worked, you know? I, I feel like, you know, you, you're closer to the, to the vest on this. You know, you're in the thick of it, but so far, I mean, I, evaluations that I see for like brand new startups. You know, the founders want a lot, but like oftentimes the investors really do get to make the call. And, you know, I, I don't sense that it's really, really bad yet. I feel like, you know, a hundred times sales for, for a SaaS, fast growth startup is actually fairly rational.

• 1:23:09Speaker 01

  So, you know, you, you see that number kind of thrown around a bit and, you know. So there's no substance here, right? I think that one of the problems is going to be, you know, how do you actually? Take AI and, and make something real, you know, we're, we're, we're just having like the idea itself is, is probably not going to do very well. I mean, if you try to say, I'm going to do AI for X, you know, that, that's sort of a recipe for disaster. But I do think that, that, that, you know, really focused ideas and really good teams will, will be able to do it. I also feel like, you know, at the end of the day, like, you know, your customer, business, business rules haven't changed. Your customer matters, you know, your product quality matters, you know, all those things matter.

• 1:23:51Speaker 01

  So like, I feel like I do see some companies and you probably see this too, where, you know, they, they, they feel like they have a really good engineering team. They feel like they're going to do something really special on the technology side and they don't connect to all those customers and they don't do anything. So I feel like that has happened, but I wouldn't call that a bubble bursting as much as it is like, you know, this happened to dot coms too. Or it was like just a misapplication of, of the technology. So your view is that we're at the start of a bubble? That's that's good summary of your. Yeah, yeah, it's so hard to explain, right? Like, I feel like in dot com, there's just indiscriminate craziness. It was like mania. You don't get that mania today. Like, it's still very like, you know, kind of rational and sober.

• 1:24:32Speaker 01

  You know, when you see something really wild and really like sometimes something like the robotic stuff that's happening right now is a little bit like a sign of like, you know, non sobriety. And so like, if you see like something where, you know, some robotics company with no product, no revenue, is raising at a hundred billion or something wild, you know, that's, that's obviously bubble behavior. But at the same time, like, you sit there and think, it's human labor. I mean, maybe that could be a hundred billion or maybe it could be a hundred trillion dollar company. So I feel like there's, there's, there's, there's relatively sort of like, I don't know that we're, we're even in a bubble. You know, I feel like a lot of it is rational.

• 1:25:11Speaker 01

  You know, we'll have to see. But, you know, and again, you see more than I do, but Wall Street hasn't had this parade of IPOs. Uh, you know, there are bits of IPOs where

• 1:25:22Speaker 01

  companies have had to temper their expectations from the private market. So I think like to some extent, there's like a, a lot of VC money and the roles have reversed in the dot com bubble. There wasn't much VC money and companies went public to get the stupid money and to get the crazy prices. Now you, you kind of get the crazy prices privately. And when you go to the public, you have a reality check where it's like, no, Wall Street's not gonna, they're not gonna buy that. Um, and I feel like, you know, because the private markets have been disintermediated by these mega, mega sort of VC and growth funds. We, we actually can't tell, right? Because like the price discovery isn't necessarily there. So you have like a five billion dollar company announced it's closing up.

• 1:26:04Speaker 01

  That's how's that possible? And it's like, oh, the VC's quit on it. And it actually wasn't worth five billion. You know, it was just a mark. So I, I think that so many things about like the ways you can even detect a bubble have changed. So we'll, we'll probably know the answer to this like five years from now, way better than we know it today.

• 1:26:22Lukas Biewald?

  Awesome. Well, I've kept you over. I want to let you go, but I did have one last question I wanted to ask you is just like, I saw you bought the Wu-Tang album. Were you a Wu-Tang fan?

• 1:26:32Speaker 01

  Yeah, I'm a rap fan. I'm from New York, home of hip-hop. You know, I think that art, you know, is one of these cool things you could do with art is,

• 1:26:41Speaker 01

  and sort of when you make a lot of money, the patronage of art, you know, so many people will buy like Picasso or Manet or Manet or whatever and say, well, this thing is really expensive. What a trophy, you know? And I feel like, you know, if you do sell a startup or, you know, come into some money, like, one of the best things you could do with it is make your own trophies. And I feel like this is like the coolest thing you could do. You know, assuming you do want, anybody would want to do this, but like recently the $40 million dinosaur skeleton was like the most complete, you know, dinosaur like ever found was sold at auction for 40 million. And it's like, that's something like I could see like, a founder like you or like somebody who like had a big success, say like, oh yeah, I want the world's like best dinosaur in my backyard, you know?

• 1:27:26Speaker 01

  And for me, here is this music album where nobody had ever paid a lot of money for a piece of music, ever. And, you know, I thought it was a really cool idea that should be explored more because music is so important to us, whereas like, impressionist art doesn't play a role in my everyday life. But like for a lot of us, we have our headphones on, we're listening to music. It's a huge thing for us. So why doesn't Taylor Swift or Drake or somebody like that make albums for, you know, a patron? You know, and they get to really channel their voice without thinking about commercial output and just say, I'm going to make the best art I can. And whether people like it or not, you know, it's irrelevant because Elon gave me 10 million to do it.

• 1:28:09Speaker 01

  And I feel like that's like, you know, potentially a new paradigm for art. And it was kind of the old paradigm for art. So I feel like there's, that was a really interesting statement to make, but I've done some other things like that where, you know, the traditional things you do with money can often be just like really tried and it's not you. Like you buy a painting, you're like, I did this. Or like you buy Summer House or You do all the rich person stuff and then you're like, this isn't me. You know, so for me, like, one of the things I really wanted to do is collect really old computers, like the first Apple ones, like, you know, like, really, really old things like that. I collect letters.

• 1:28:43Lukas Biewald?

  There is a, there is a museum in Seattle. Did you ever go to it that had the, all them working? It was amazing.

• 1:28:48Speaker 01

  Yeah, I haven't been yet, but you know.

• 1:28:49Lukas Biewald?

  Wow. Yeah, I guess, um, I think it was one of the Microsoft, Paul Allen, I think, was also a collector of the stuff and made it available. So if you do get a good collection, let me come over and play with it.

• 1:29:00Speaker 01

  Yeah, you know, they're, they're, they occasionally will go to auction for these like Apple ones and twos and, you know. Thanks for having me, yeah.

• 1:29:08Lukas Biewald?

  Yeah, no, it's funny. I actually, I, um, I love, I mean, I remember when Wu-Tang came out when I was a kid and it like, it really spoke to me as like, yeah, definitely not the Target's demographic. And then I, I remember going to a show in like 2010 and running into like 20 other like startup CEOs that I knew in San Francisco, just being like, huh, this is, I mean, absolutely terrible show. Like, I think they had gotten old and like, were really bad, but, but, um, it's just funny to, to like, see that connection with like a lot of the people at once.

• 1:29:34Speaker 01

  Someone did do this, like, um, the numerical analysis or some other data science analysis of Wu-Tang vocabulary. And it was like many sigma to the right of like virtually any other rapper. Now it's, it's a group of guys, so it's a little bit different, but even still like the lexicon of Wu-Tang is like massively larger than almost any other group. So I thought that was really interesting and that, that the use of the metaphor was also like very, very significant in their words.

• 1:30:01Lukas Biewald?

  I think a weird mythology that I really, I don't know, I got like really into. So yeah, anyway. Yeah. Also, Finn. Anyway, thanks for taking the time, man. I really appreciate it.

• 1:30:09Speaker 01

  You so much.

• 1:30:09Lukas Biewald?

  All right. Take care, man.

• 1:30:10Speaker 01

  Bye.

• 1:30:13Lukas Biewald?

  Thanks so much for listening to this episode of Grading Descent. Please stay tuned for future episodes.