FULL INTERVIEW: Thomas Laffont’s Journey From Hollywood Assistant to Legendary Tech Investor

24:20Thomas Laffont:And then also, we had companies going public pretty early on. So you could do a lot of differentiated work in companies that the market cap was a couple billion. Yeah. Right. I think what really changed for us in the early 2010s was Meta, then Facebook, and Alibaba staying private for longer. We just never seen companies of that scale who were that important to our research and our market not go public. It was Meta or Facebook went out at like 60 billion, I believe, something around there. Yeah, I think around there, right, and right around 2012, 2014. So if you're used to buying a company potentially at$2 or$3 or$6 billion, that's a big miss.

25:01It's like a 10x difference.

25:02Thomas Laffont:But not only that, we actually were an investor in Google from pretty shortly after the IPO. And Google had this stretch post-financial crisis where it kind of traded sideways for a long time. And the reason was, here comes Meta or Facebook, and they're going to replicate a private internet that Google won't be able to search. And so Google is going to be under pressure. And we weren't able to talk to Meta, so we didn't know what they were thinking. Sure. As soon as the company went public, ironically, Google stocks started working because they didn't come out and tell you, we want to kill Google or we're replicating the internet.

25:38Thomas Laffont:We're doing kind of something different. Both stocks ended up working. Yep. So that was kind of a big eye-opener to us. It felt both offensively minded and defensively minded. How could you tell at that moment? Obviously, it was correct that companies would stay private longer, but I'm sure you were debating the question, is Meta the outlier or the exception that will eventually prove the rule versus there's a structural shift in private equity venture capital that will propel many more companies to stay private well into the tens of billions of dollars in market cap? We would not have foreseen what ended up happening.

26:13Thomas Laffont:We had an instinct that it might happen. Yeah. And remember that Spotify is kind of getting built at the same time. Right. And it's redefining music. And we really, you know, as I mentioned, Apple was a core thesis for us. And now we'd here comes subscription music, which goes directly against Apple's model of selling you an album. And so what's going on? They just did around a three billion felt like a lot. And then Uber. Yeah. Right. So it was it was just kind of you felt an Airbnb. Right. So I would say like those two companies, the mobile industry, coming out these companies getting big in in private markets it felt like undeniable momentum right and so and China by the way right the same so we felt like we had to participate so what was the first private investment you made Evernote I think was one of the first and we said look we're gonna do later stage deals a hundred million plus in revenue so we did deals like Evernote and Box and ironically our most successful deal is one that broke all of the rules that I just laid out which was Snapchat.

27:22Okay.

27:23Thomas Laffont:Where I think we led the series C in that an evaluation of about a billion and a half. And what was the reaction from other more traditional venture investors when you guys started leading around? I think that look venture felt very different back then. There was fewer firms there was more atrophied thinking right Andreessen is just kind of starting we actually shared a building with them so we were kind of starting at the same time about as they were and it felt like wow we're gonna bring a bit of a different competitive energy to this market it felt very clubby sure you hadn't seen like these new firms like founders and obviously you know and a bunch of others kind of really make their mark.

28:15Thomas Laffont:So it was sharp-elbowed for sure and still is in many respects, which is probably what I like the least about that market because I love talking about ideas. I love trying to be a positive sum thinker, which the public markets... Easier to do in the public markets. Correct. Did you bring the models to private markets? Were you building financial models or the team? We did. I think we brought that, we brought analytical thinking, we brought kind of deep research. Like I remember reverse pitching Aaron Levy at Box, a big deck that we had done and we had just done what we thought was kind of public market like research but we brought that to a private entrepreneur and he hadn't seen that kind of work before so that was a differentiator for a minute until other firms realized, right, we can do that or even better we can outsource it to Bain yeah right and so we had to kind of quickly kind of adapt to that yeah but in the beginning it was novel yeah and that was an industry that was really done in word and we were Excel thinkers yep yep so that was kind of a very different kind of mindset that we were bringing to the table was there any shift required in the messaging to LPs the fun structuring anything that were that you had to work through in order to actually set up the fund for success in the private markets?

29:36Thomas Laffont:I think our LPs, first of all, LPs are not talked a lot about in venture, which is kind of interesting. Like we talk a lot about the founder. We talk a lot about the companies. But I would say for us, we are pretty clear that our customer at the end of the day is our LPs. And so the trust that they give us means a lot. I have virtually no outside investments, right? I do some as favors and things like that, but almost everything I have and own is in our funds. So we kind of act as entrepreneurs and as owners ourselves. And we ask for trust from our LPs. And I think at the end of the day, they were willing to give us a chance in our first fund.

30:20Thomas Laffont:I don't think they held us specifically to exactly what we said we were going to do, but they're like, these guys are pretty disciplined and they're pretty smart and they're entrepreneurial and aggressive and let's see what they can do, right? And so I think over the years what's helped us most in our business and I think why we're still in business 25 almost years or 27 plus years later when a lot of our peers have disappeared over time is we never lose sight of our investors and hopefully we've made some good decisions but we've also made some bad ones but but I think our investors learn more about us on how we deal with our bad decisions.

31:02Thomas Laffont:And so I think we've earned hopefully some trust from them over the years. So I remember distinctly an investor when I called about the Snap deal and saying, look, I know this is a bit off brand. This was one of the largest investors in the fund, but I just have a lot of conviction in this deal. And he said, then do it. That's what ultimately why we're investing in you. And so I think the trust that you build, the relationships that you build with your own investors over those periods of time are really important. So relationships with entrepreneurs, building models for Apple, ARPUs, projecting units, that feels all very micro.

31:43How have you processed macro statistics and factors like interest rates? Everyone talks about when interest rates rise, all the DCFs change. There's a pullback in the private markets. We lived through this with the end of ZERP. But how much are you tracking the labor market, the GDP numbers, the interest rates, and how much of a factor is that on the strategy day to day, month to month, year to year, or even like broader terms? Yeah.

32:13Thomas Laffont:So data science has really become a much larger part of our business than it was back then. Sure. So we now have, I don't know, maybe 20 or 20ish, 20 to 30 people, something like that in data science that are just processing different types of data and alternative data. So I think it's not just macro data, it's App Store data, it's clickstream data, it's credit card data. So we use a lot of that for our investment research. And some of that we even make accessible to our portfolio companies. Right. So that made us smart about a trend. We were early customers of data Briggs and Snowflake as an example that let us invest in those companies.

32:53Thomas Laffont:So that is way more of a presence in today's world than it was, you know, 20 years ago. So we're constantly looking at data as an example. You know, I think OpenAI is probably the most important company in the world today in the sense that it's the driver of AI, both consumption and spending. So I look almost every day at the chart of ChatGPT users, download share, how it's weathering the storm versus other competitors. You know, that's really something that wasn't available 10, 15 years ago. Sure. We had this amazing data set called Onavo, which actually gave you engagement data from users on phone, and then Zuck bought it and turned it off.

33:37And I remember thinking, like, damn, that guy. That's a great move.

33:42Thomas Laffont:It was the only data set that really gave you engagement data So we're always looking for kind of new data sets, right? And then obviously, so that felt like a major shift, right? Going from data science enabled research, and now obviously we're getting to AI and agentic research. How are you thinking about AI as a category from an investor perspective versus the data bricks and snowflakes? which to me feel it's easier for me to maybe understand the financials, the model that I would build, how I think about value accrual and competition in Databricks and Snowflake. Fantastic businesses, but feels like easier to pattern match against previous eras of software and tech innovation versus AI where you have infrastructure and capex and training costs and inference budgets and all sorts of different, your entire product's getting copied by open source every three months.

34:46And it just feels like a different puzzle to solve when you're thinking about underwriting those businesses. How have you grappled with that? Do you see it as an extension of the tech investing or is it an entirely new motion?

34:59Thomas Laffont:Well, for me, it was almost coming back home to what I knew because the infrastructure layer was really semiconductor-driven. Sure. Right. So I think our knowledge of semis and our team's knowledge of semis was a great head start because a lot of people just hadn't done semis. So we're kind of new to semis, and I had a lot of relationships in the industry, and that led us to leave the series being Cerebris that I think will be kind of a generational kind of company. Yeah. I love that he's been on multiple times. He's been great. So that felt very natural to us. And pretty quickly when we saw what Jensen was building and the momentum that NVIDIA was building in the data center.

35:41Thomas Laffont:So that was kind of our first telltale that, wow, something big is going on here. So we had seen semis before in the mobile era. So we felt very equipped when AI first came around to look at it from a semiconductor, GPUs, memory, TSMC. I've personally been to Taiwan many times, visited TSMC. Great anecdote, I'm driving back to Taipei City with my host from TSMC and we're on the highway and there's a golf course and it's nighttime so there's lights and people playing and I just very innocently turned to him and said, wow, you guys play golf at night here? He very innocently looked back at me and said, well, when do you play?

36:22Thomas Laffont:and that's when I realized we're in a different level of work here so we saw it at the infrastructure layer first where it just became obvious you didn't have to necessarily worry about who was going to win the whole infrastructure layer will win so I think that was kind of layer one I think layer two then came kind of the models and obviously we're investors in a number of them I'd say the most complex element of AI today is you can almost talk yourself into a bulk case and a bear case for almost any name in tech and I think software is kind of seeing that right now right so a software gonna win because of AI or get displaced right so you've got that in infrastructure you've got the well when is the peak and what multiple peak should things kind of trade out right so it's both an exhilarating, dynamic, but also very complicated environment in the sense that, for example, when Databricks came around, no one thought, well, gee, Databricks is going to put Salesforce out of business.

37:30Thomas Laffont:It just felt like a new architecture. There's something about AI that feels a lot more disruptive. And what if your model, not today, but in two years, can just build you a workday right off the bat? What does that mean for Workday? Does that mean their data is more valuable on the left hand side? Or no, does that mean they get fully wiped out? So I think that battle, and it's being played out kind of in the public market today, right? You kind of see it in these names. How can a public company CEO actually communicate a vision for that case, the bull and bear case around AI, what effect AI will have on their company?

38:17Thomas Laffont:I mean, you're seeing it. I read something that we're kind of moving into a selection market, right? So now, like, some companies are going to do well, some companies are not. I mean, look at Square, right? Jack came out and just said, no, I'm pivoting the entire infrastructure of this company for an AI era. We think you're going to need to be remote, right? Because you're going to move faster if you're remote. And so he's kind of laid out a whole vision about how he wants to run the company. That's because you're sort of generating the necessary context because you're not getting the in-person interaction.

38:55And small teams. Yeah, small teams. You'll more easily be able to be AI native if you're basically explaining process.

39:03Thomas Laffont:Very small teams moving really quickly without a central organization or you know kind of like the Borg right no central organization organizing force the model drives everything and then you have other companies that are saying no product well the labs themselves right are all in person and they believe right that you product development needs to be done kind of in person so you're kind of seeing a lot of these different ways of you have different models some people are going to charge for tokens some people are going to charge for data access or ingress and egress so I think what we're seeing right now play out is a true Darwin-like survival of the fittest where software companies are like the mailroom exactly right you saw Anil come back to Workday right he probably thought that I think Carl is an amazing executive and is a good friend of mine but maybe he thought in order to make the changes I need to be have that kind of founder mindset right founder mode as Brian calls it.

40:06Thomas Laffont:So we're seeing a lot of these different approaches kind of compete with each other. I think it's too early to tell who's going to win. But eventually, I think we should see kind of separation between winners and losers, which should be good for our business. But I think right now it's still so early that it's not clear who will win. Is it enough to look at re-accelerating top lines? I imagine we've talked to a lot of founder CEOs, maybe their unicorn status decade in, completely reinvigorated by AI. They come on, they see that the growth has returned. It feels like a new startup, even though they're maybe coming back from a sabbatical, maybe they're coming back in after hiring an outside CEO.

40:51Maybe they're just coming back in with a new vigor but what are you seeing more at the earlier stage or mid-market stage around companies that are starting to show signs of being winners in the age I

41:06Thomas Laffont:think look the good news is no one is head in the sand about this so I do think in prior cycles you had more of a head-in-the-sand mentality right so for example if I remember when cloud got started there was do you remember virtual cloud, right? That was going to be the big thing, right? I can't just have my stuff in Amazon. I'm going to have this virtual cloud and, you know, obviously that hybrid cloud, right? That's another one. No, it's going to be half all the big data, right? All those things basically just got torched in by the wayside, right? So there was a lot of head in the sand. Similarly with the iPhone.

41:42Thomas Laffont:You need a keyboard. It doesn't have 3G. It doesn't support Adobe. It doesn't have copy and paste. Exactly. The battery life, right? Ben Gate, you may remember that one. That was a whole weekend wasted on Ben Gate. So there was a lot more to me head in the sand in prior investments cycles, in prior tech themes where people were just pushing back against the idea that this was going to work. I think AI is one where the consensus view is it is going to work. it's not it's an extension extension level event and so the sense of urgency is high so that does feel a little bit different to me than maybe prior cycles where I think that took time for people right the cares are like well I'm not gonna allow Apple to have an app store and I don't want to be a dump pipe and right so they all these things that they fought yeah and over time tech one yeah the difference to me with this specific cycle is everybody agrees it's gonna happen and it's happening quicker and the stakes are higher than any other.

42:50Thomas Laffont:Yeah. So I think every board is ultra motivated. Every founder is focused on this. Now they're bringing different approaches and, you know, we'll see which ones went out, but I would say they're tracking token consumption, right? So how much of my revenue is token based, right? How much of my COGS is token based? How much of my GNA is token based? how much of my spend per developer on cursor, OpenAI, and Anthropic is happening, right? So that in, you know, I do sit on a bunch of boards as an observer most of the time, so I kind of see a lot of that kind of happening. So the awareness is absolutely there.

43:31Thomas Laffont:People are doing different approaches. There's some that are, no, I'm in a white box, a totally new product, right, where I think I'm uniquely positioned to build it. And then there's others, well, like, no, my data set is so valuable, I'm not going to allow my customers to build apps directly using my data. So you're seeing a lot of kind of different approaches. Yeah. Right. But everyone's awareness is at a 12 out of 10. So there's no convincing needed. Right. Everybody's aligned. Every board member is aligned. Every investor. And now it's about, OK, what does that sense of urgency mean for this company?

44:07Thomas Laffont:what are the things that we need to track, and the things that we really need to go and execute on. What does it take to make it as a new hire at KOTU? So we do, and I'm assuming you mean on the investment staff, right? We do case studies. Very important part of the process for me. We usually pick a public name, right, because we want to test your thinking. And my favorite types of names are names where there's a good bull case and a bear case. and whichever one the prospective analyst argues, I will vehemently argue the opposite. Yeah. Right, just to test their thinking. Exactly. Understand their thinking.

44:48Thomas Laffont:So that's really, really important. Are you trying to pick obscure names or household names, everything? No. I mean, for a long time, we used Netflix. Okay, yeah. Right, as an example. That was a really controversial stock, both in the DVD era, then when they moved into streaming, then when they moved into proprietary content. and it was a heavily shorted stock over that period of time. So there was a lot of interesting ways to look at that name, right? And you could ask interesting questions like, well, if they increase, you know, price by a dollar, what happens to EPS? And what you realize is it was almost all profit.

45:23So EPS went up a lot.

45:26Thomas Laffont:So we're just really trying to test thinking. Have you ever had to revisit a candidate who made a really great bull case or bear case that the firm maybe didn't agree with and then they came back with an I told you so five years later? You know, I've never had someone kind of email me that, which is surprising because we've done a lot of cases. You've done a lot of case studies. I can imagine there's a couple I told you so's where I was like, oh, I called Domino's or whatever. And honestly, to me, it's not about whether they say the right thing. Sometimes you build it. It has to be, if I say bull or bear, it doesn't really matter.

46:03Thomas Laffont:It's like, how did you articulate your thinking? Did you lay out a clear model? A lot of kind of where we were starting, right? And so that's my favorite part of this job is thinking through a name and the opportunity and what could happen. and it's kind of the intellectual backbone of what we do. I would say that I think the key to our platform is, number one, like seeking big themes and big ideas. That's a big one. And then kind of the risk management piece, right? That's kind of kept us in business for a long period of time. How is AI changing the role of early analysts or analysts who are earlier in their career on the investment side?

46:45Thomas Laffont:Too early to tell. Too early to tell. Obviously, look, we use AI every single day. I use it a lot to test my thinking, to clarify my thinking. I've always had a weird dichotomy, personally, where I love reading, but I'm a terrible writer. And one of the things I like about AI and ChatGPT specifically is it's helped me actually write in a way that I can be proud of. not just sometimes I write you know I'll write something an email to somebody I'm like this is just so badly written you know and it's just I don't know how to make it better yeah I know it's not good yeah I don't know how to improve it yeah and it's so frustrating because I know what great writing is from my reading but I just can't do it yeah and at least yeah I've had a had a family member sent me something that was very obvious to me, AI generated.

47:41And I think people have an aversion to AI generated text. I totally don't get that. But the thing is, I was reading through it and I was like, this is very cool because I know this person would not have been able to articulate their thoughts in this way, but they went line by line and I know they mean it. And so they were able to communicate something that they never would have been able to communicate with text. Maybe if we sat down and spent a couple hours talking through it, I would have been able to get the gist. Exactly.

48:11Thomas Laffont:Some people take out the M-dash because they write, they're like, oh, I don't want it. I'm like, what do I care? Leave it in. Yes, a lot of emails that I write are helped by Chattypt. It's like the Arnold Schwarzenegger line. I smoke my stogie everywhere. Why judge me on what I said? Yeah. My idea and whether it's well written. Yeah. Who cares whether it's was written by AI or not? That I totally don't get that. Yeah. Or polished. In fact, I hope that more people are able to communicate things that maybe they couldn't before. Right. Yeah. Because they didn't know how or they only knew granular things, you know, and now more people can write, more people can communicate.

48:57Thomas Laffont:more people can express themselves like to me that's an incredibly empowering right vision talk about how you guys have approached investing in in you know multiple companies in the same category or the same general category how has that evolved over time were you mocked early for doing that from maybe some of the more traditional funds and then how have you manage to make it work and practice and maintain the trust of entrepreneurs. Yeah, I think conflicts, which is kind of what you're, has definitely changed a lot in the Valley as companies have stayed private longer, right? And I think we have to be kind of precise by what we mean by conflict, right?

49:43Thomas Laffont:So as an example, funding two Series A companies at the same time that are pursuing the same opportunity is an obvious conflict that I think no firm, including us, would ever do. Right. So let's just kind of be very clear on that. I think it's quite different when now you're talking about these very late stage companies that are kind of competing with each other. But look, every company is kind of competing, cooperating. You know, Apple and Google are great examples. You know, they compete, but they're partners. So I think the distinction and the conflicts distinction has to be, you know, kind of changes as companies and markets kind of mature.

50:25Thomas Laffont:Right. So I think you're seeing that become much less of an issue in mid to later stages, even by firms that, you know, would typically viewed conflict as core to what they do, like traditional venture firms. Right. Have not kind of moved in that direction. And I think it's just the nature of the market. So that that's kind of would be my first point. I think the second point is the execution of it, I think, really matters. So, you know, if I view a perceived conflict between companies, even if they're later stage, I will always let the founders know directly. And I'm not asking for their permission.

51:03Thomas Laffont:So I think you also have to be clear with the founder, because what if they say no, and you still want to do it, then you're in trouble. Now you've just broken your word, and I won't do that. But I will inform them. I'll be very direct. I won't let them hear about it from somebody else or something like that, and I'll explain kind of the rationale, right? So I think communicating directly both good news and bad news, that is something that I learned as an agent. I'm not afraid to have difficult conversations because I think we can grow from them, and I ask the same of founders or employees that I work with to both come to me and say, if you have an issue, let's just kind of talk about it and look I've hired a lot of people I've fired a lot of people over the years I've asked a lot of people to go and look for different career paths so I'm comfortable having this those conversations so you know that's not something that you know I think your your reputation and trust then is kind of the the second point on the execution right of we take information security incredibly strongly that's when you know we're SCC registered and you know even And before coming here, I got like a four page memo from my lawyer about how you can talk about this, but not that.

52:17Thomas Laffont:And I see this. And so of course, you know, I, I have to read it and you can tell the Ralph Lauren story. Yes. I hope so. I mean, by the way, I, the way it will, you know, so side note, I believe that meetings should be recorded as an example. Now my compliance will say, shit, we can't have meetings be recorded because it creates a paper trail. Yeah, and saying in discovery, they can take anything out of context. And let's just not even talk CO2 specifically, let's just talk an enterprise. Yeah, yeah. But then I say, okay, let me pause it. Enterprise CIO says, no, I can't have my meetings recorded.

52:52Thomas Laffont:I'm too afraid. I'm like, okay, let me pause it two scenarios to you. Okay. Scenario one is in each you have a bad actor that's doing bad things, right? Whatever that is, you know, like belligerent, talking out to people, whatever. So scenario one, nothing gets recorded. you know nothing and 10 years later someone comes out of the woodwork and says by the way X Y & Z 10 year pattern of deception nothing happened okay so that's scenario one right scenario two is every meeting is recorded the first time said person does something that's not right the compliance system right which is always listening sends that person an email and says hey by the way Better you didn't do this, talk to that person that way, disclose this piece of information, depending on the severity, right?

53:42Thomas Laffont:Second time person does it again, says, hey, I now have to flag this to HR, right? I would much rather live in world number two, right? Because you know what the problem is. There's a system. There's flags that have been raised. and eventually, you know, someone kind of gets involved and you either remediate or you terminate the person or whatever, right? Yep. So to me... You avoid the back and forth of he said, she said, all of that, yeah. Exactly. And who knows, maybe that person, if they had gotten that first warning, might have realized, oh, wait, yeah, you're right, I'm being abusive or whatever the case may be of whatever they were violating, right?

54:26Thomas Laffont:So I think to me that's a better world than kind of the ignorance of while getting no feedback. And then you just learn much later that kind of you had a problem. Yeah, I mean, it certainly seems like a trend. I mean, Bridgewater's written about it a lot, Ray Dalio, but then also Grimelow. And he did that with no analytics, right? So that's different. Correct. I think to me what will happen is the analytics are going to get so much better. Interesting. Right. And these systems are going to know. and they're going to be able to look at your WhatsApp and your messages and your emails and all of your calls.

54:59Thomas Laffont:And they're going to be able to just say, Hey, by the way, just don't say this. Or did you think about this? Or maybe you could have, they're going to coach you. They're going to say, Hey, you told this customer to fuck off. Be like, well, maybe you shouldn't do that. And here's like two other ways you might've mentioned, you know, like whatever this situation. Yeah, your frustration with the situation. Yeah. No, very interesting. Instead of the token maxing dashboard, you're going to have the social credit score. Yeah. It's the bottom 20%. No, no, there's a lot of environments that it can make sense.

55:33Yeah, maybe there's a divide on like the in-person versus remote work, crowd, like different Well, let me be clear.

55:38Thomas Laffont:I'm talking about work context. Yeah, yeah, yeah. I'm not talking about after work. Yeah, I just wonder if that would become something that employees select into or out of for various reasons. Just like some people are huge fans of remote work, some people can't stand it. And there's a variety of approaches. I also think there's a clear distinction between transcription and recording. They don't necessarily go hand in hand to me. So I don't necessarily need a system that transcribes every word that was said and keeps it in some database somewhere. I'm not sure that's necessary, but key takeaways from the meeting, what was said, what was agreed upon, like that's useful.

56:18Thomas Laffont:That's a good corpus of data. To me, just because someone's quote recording doesn't necessarily mean that it's transcribing. In fact, I like the option of deleting. In an ideal world, I would recommend, well, you can delete the transcript. The transcript's not that relevant because maybe you batted an idea back and forth three times and you said something that turned out not to be true, but you figured that out later. So you don't need any of that. What you do need is what was said, what was agreed upon, was anything done out of compliance or not, right? So it doesn't mean, it doesn't imply a world where everything you say is recorded.

56:50No, right. No, it's funny because there's so many opinions on this, but we record everything all day and live streaming on the internet. How have you processed a number of these venture funds that have become publicly listed that are taking a lot of the different names that Code 2 is in, you know, trying to find basically the most in-demand secondaries, putting them into funds. As I've watched, I think what I've seen is like, yes, it's very obvious. There's an incredible amount of demand from the public to invest in these names. But the big issue is that as soon as that demand floods in, you know, supply and demand, price shoots up, and then you have a bunch of people investing at effectively, you know, 10 times what the private valuation or the underlying asset value.

57:40But how have you processed it? Do you think there's a more elegant solution over time?

57:45Thomas Laffont:So we do have a fund called CTECH that I'm allowed to mention that addresses some of this. And people can kind of go online and research more about it. What I would say generally is people want access to these companies. And I think there's a lot of arguments for going public. One of, I think, the most powerful, in my opinion, is to democratize access to companies. Let's take an Anthropic. Let's take an OpenAI, right? And enabling the retail investor, enabling the Trump accounts, which I think is a marvelous idea. Then my friend Brad Gerstner really spearheaded, and I give him a lot of credit for this idea of, wow, Why can't we have every single new child already have be invested in the market and participate in the value creation of these companies?

58:42Thomas Laffont:So I love that democratized nature of it. So what I think it speaks to is there is incredible demand. Right. Let's say you were sitting. You're not a VC investor. You're maybe a dentist. And you're like seeing OpenAI and Anthropic. And you're like, wow, why am I not able to participate in that? You know, why is it just like an elite group of funds and accredited investors and so forth and so on? That to me will have to change. And I think it should be bipartisan, frankly. Totally. Right? And I think there'll need to be some guidelines and stuff put into place. And we don't want bad behavior and all that kind of stuff.

59:31Thomas Laffont:But I think there is incredible demand from the retail investor base to participate in the value creation. And I think what we're learning as a society is the cost of not having broad participation is incredibly high. Completely agree. Right. And we'll be, right, if you have a whole generation of young people that don't own their house and have student debt and don't feel like they're economically levered to open AI or anthropic, or even more so are directly threatened by their technologies, I don't think that's a great future for any of us. So I'm not saying that opening out or anthropic or in public is the one thing.

1:00:20Thomas Laffont:Obviously, there's going to need to be a lot of things that are done. But I do think the transparency that comes with it, the democratized nature of it will make a huge difference. Yeah, makes a ton of sense. Jordan, anything else? Last question. What are... Can I make also one point on in-person? I'm so glad I got to do this in-person. I did not want to do this remote. Because the tactile feedback you pick up in-person, as an example, for the viewers at home who have not been to this office, the open jar of creatine, just... What, I mean, what a fall. Open jar. Just for everybody to describe it, there's a goodie bag.

1:01:04Thomas Laffont:And there's just a jar of creatinine. It's just wide open. The scoop's right in there. It's so inviting. If you want to power up before you get on, it's just right there. It helps if you're sleep deprived. That's just the kind of tactile feedback you want. Yeah, you don't get that on Zoom. On Zoom, no way. I'm really glad you can be here. Well, I'll save the last question for your next appearance. Okay. Let's do it again soon. Well, thank you so much. Thank you for going way over. Yeah, we went way over. Delighted to be partners in company AI now with you guys. I'm really proud of your success, honestly.

1:01:34Thomas Laffont:I love hustle and people that break into industries. And so congratulations. Yeah, we bring a very mailroom approach to podcasting. We do. Yeah, I can think about this. There's a lot of mailroom here. A lot of suits here, too. Well, thank you for watching. Tune in tomorrow, 11 a.m. Pacific. Leave us five stars on Apple Podcasts and Spotify. Sign up for our newsletter at tbpn.com, and we will see you tomorrow. Cheers. Goodbye. See ya. We love you. Goodbye.