Merchant Banks: Venture’s Next Act with Will Manidis

[Music] Welcome to Origins, the podcast that dives deep into the business of Venture Capital, where we learn how the people behind the capital, both GPs and LPs, make decisions. I'm Nick Turls, GP at Asylum Ventures, a newly launched early-stage venture firm celebrating the creative act of building startups. And I'm Besar Clarkson, your LP Co-host from Sapphire Partners, where we aim to invest in exceptional early-stage venture capital managers, both emerging and established franchise VCs. And I'm excited for our guest today, Will Minidas. Me too. I met Will recently on Twitter. We were in a debate about where venture is today, big firms, incentive alignment, misalignment, and actually thought it would just be fun to bring them on Origins and we can all debate these things. And one of the ideas was to try to cook up what an ideal aligned venture capital firm might look like. We can do that together. Will is not a VC, which I actually think makes his perspective even more interesting. He was the founder of Science I/O, a company that built foundational AI models to make healthcare data computable and actionable. The company was acquired in 2024 by Veridime and now Will is the SVB overseeing AI initiatives there. So I'll also chat a bunch about AI and healthcare. Will was a Teofellow, a former managing partner at Dorm Room Fund and he's also an active investor. Oh, so cool. I'm really excited for this. I love the outside-in perspective on venture because we can all learn something from everybody, right? And there's always room for innovation and hot topic is AI intersecting with healthcare. I know we've had some guests on with this. I suspect we'll keep having more. Is it it seems to be something that the venture world is really sitting down thinking about hard. So very cool. Should we bring them in? Let's do it. Welcome Will excited about this. Yeah, excited to be here. We have a ton to cover, but right before we started, we were talking about how you just came off a four-hour podcast, which is going to be a little longer than our than our pod today, but you pre-gamed, I guess, origins with a four-hour couple weeks ago, a couple weeks ago. I feel like I've slapped for a couple hours in order to get back. Good. Okay, so how did this go down? The four-hour podcast, so it tells. Yeah, so I think I have been helping with technology, brother's podcast network, right? I think the only media company that will survive the next 15 years. Right. If you think about where the company is in the future, I think like number one through 499 of the Fortune 500 will just be this podcast, but we've been helping put it together, helping do some structuring on it. I tapped in to do a full episode, and I was like, it can't be that hard to sit on a microphone for four hours and talk to guests. Let me tell you, it is impossible. Like these guys are professional, athlete levels adrenaline, just to get through the thing. It's crazy. I have so many questions. Where do we start, Nick? Well, first of all, I was saying that they invited me to be on an ab init chukin. Just for 20 minutes, I've been a bit on the how do I prep sufficiently? So oh, my goodness, hats off to you for four hours. I think you got to do the 20 minutes segment. I think it's incredible. I will. I'm such an LP. I want to wait till the half-year numbers are out. What is cell phone? Yeah, it's like it's like, it's like Yale has to go first. Like Yale goes first, you temco goes first, then you're fine. And I'm a fast follower. I'll learn it. No, mostly because I'm a total data nerd, which is why I'm so excited to talk to you about all things. But it's only interesting when you can get the half your numbers, which aren't really fully baked until really late July or August. So I'm that annoying person who wants to be specific about when they come on. So they're probably eight me and I've probably been uninvited. But that feels correct. How profitable is it? Very, very profitable. I think it's fair to say that like it is produced who more EBITDA than 99% of venture back startups ever will, which is not a high bar, but it is. Not a bar. Although with AI startups, higher bar, although the pain in how you can't read BIDDAB. Yeah, I always assume that like the AI companies are just like 90% of their balance. She is just kind of flow through to a Sam Altman control entity. Like it's not, it doesn't feel obvious to me that that's living on that P and L for that long. We'll find out. But do they, is this mostly advertising spend them? Is this the last, is this the OG advertising? Yeah, I mean, the way to think about it is it's not a news show. It's an ad show. Right, really, the sponsors are the product. So if you think about like companies like ramp that are are sponsoring end to end, including like sponsoring hit pieces and the information, you know, ad content really is the content here. Well, this is such an intriguing and I'm going to bridge to AI so we can talk about then bridging in the science. I'm going to get there, watch this. So this might be the counter to AI, which seems to be eroding the advertising because an agent to agent communication doesn't need an ad to help find something. So this might be the last place the advertisers go in which case they're going to have a ton of revenue. Yeah, I think this is like one of my big thesis is we're going to see this like rise in a elite media. I think like if you think about all the AI slop right now, it's not aspirational. It's not great media. It's not curated. Right. It's the end result of kind of unfeeling algorithms. And I think it like physically makes your stomach sick, like viewing the Facebook timeline at this point, like 900% AI, right? And I think what you'll see are things that I'm calling a elite media think like strike press the FT TPPN, which is like highly curated, opinionated things, but people that you are aspirationally following that are so devoid of AI that they're like just fundamentally human. And I think that kind of elite super curated craft like media is kind of the future of all of this. And I think you'll see that in software too. Like I think this like anti AI craft push is more durable than we'd suspect. Okay, I'm going to say this on next behalf that is so silent. Yeah, you missed origins, but I assume that was like fourth or third. Yeah, of course. It's you and Jackson doll holding up the podcast front. It's the only other two podcasts that will survive the Chris Baker's podcast. That's how interesting. Um, I don't know why I mentioned that. Okay, tell us about well, quick, quick backstory. Give us the Rundo. I was lucky enough to join a company called Foundation Medicine. Foundation was a cancer diagnostics business that was starting to commercialize its data. They had basically limited data infrastructure when I joined. We had just signed a agreement with kind of a plucky New York startup called flat iron health. Lucky enough to see the entire life cycle of commercializing that data partnering with flat iron. That public business goes public, ultimately acquired by Roche for a couple billion dollars. They also acquired flat iron for I think 2.1 billion. Zach Weinberg will be head may forgetting that number and correct. No, I think that's something that's accurate. And then end up just being kind of disturbed by how bad data tolling is for healthcare data. So if you think about like fundamentally what the business of flat iron was, it was having medical records on one screen that you owned and spreadsheets on the other and having medical doctors sit there and translate all this messy unstructured data to structured elements inside of spreadsheets. And like I'm not smart enough to become a doctor. Like I'm not smart enough to make make it through an engineering degree. So like certainly it seems wrong that these people could study for a decade and be glorified spreadsheet jockies. So when we started science, the goal was really hey we think this large language model thing is real, right? Like we got serious about the company, the second, the attention is all you need paper came out, but we're going to need domain specific very, very narrow models to solve high trust regulated use cases such as abstracting healthcare data. So we thought high quality data, high quality models, high quality human feedback, you could imagine a much more responsive healthcare system where you import data in, get structured data out and that becomes much easier to work with. Right, and that company for about five years and then ended up selling into a company called Veridime, where publicly listed electronic medical record provider, focused on ambulatory practices. We ended up in Veridime for one very specific reason, which is that Veridime has the best data in the world. Like just the quality of medical data that the company sets on is unbelievable. And one of my grand thesis is that models don't matter data matters. And if you take the long view of that thesis, you're going to want to end up inside of places that have robust access to workflow and robust access to data. Can I dig in on the AI component? I'm curious how much maybe it was machine learning back in the day, but how much would you say was AI in the science work, the science IO work that you were doing and then thread it to if it wasn't there, what should have been doing that, right? Yeah, I mean, we were we were very early on transformer models. Like we were training, we were taking generative models very seriously before I think even internal teams type of up an AI were the thing we didn't see then was the emergence of scaling laws. It was our view that high quality data would ultimately be the rate limiter. And it wasn't obvious that if you just push sufficiently far on the on the compute curve that you could start to approximate what really high quality smaller models could do. That said, I think now that we're, you know, towards the bleeding edge of scaling laws, I think our thesis was correct that you do really want these high trust very small models that are compute efficient for specific use case. But I think the thing we didn't see come in was the kind of 20, 23 era foundation model creates, right? Where people were raising billions and billions of dollars to build even like country specific LLMs. I still think largely that was like a capital formation symptom, not a company symptom. Like I do really think the fundamental innovation of a foundation model company is being able to charge to in 20 on a billion dollar check that you write over a weekend. But like, short of that, I think we got the call right. It's a good good gig. It's great business. If you think about it, that's like a head of things every time one of these companies races around. So I think we got the call right. I think it was unclear to us how much the data would matter. That's what has been very interesting to see at Veridime. But I think we got the call right on the Transformers to models, to language models thing. What areas, when you think about maybe like zooming out a little bit in healthcare, like what areas do you think will be impacted in the short term? And what areas maybe are like a little high B that maybe take a lot longer? The big cultural shift right now is just the willingness to embrace technology. If you think about how EMRs were deployed in healthcare during the high tech era, providers did not want these things. Right? We had a past legitimate congressional incentives to pay providers to use electronic medical records. Compared to if you go to a doctor's office today, your provider is probably using GPT. Like I was in a medical appointment the other day and he had like chat GPT open on the screen writing notes for him. And patients are doing it too. Like it's everywhere. And like if you go on TikTok right now and you search provider or insurance reimbursement, you'll find doctors writing denial or appeal letters with chat GPT. Like it's just a pull. And like I got blood work back this morning. I pay for a very nice constiearch doctor. I did not ask him to interpret it. I just put in chat GPT and I got results back. I did something very similar recently. Me too. Actually compared my doctor and chat GPT. Yeah, it makes you wonder like how good these, yeah. I made the compete and AI won. Because they gave me a better score or because they were more accurate. Because I always like the people that tell me I'm healthier. The level of analysis, the like the game plan, the way I was able to work with it to like build a game plan like just for me, also within minutes and free. So just a significantly better product. And unfortunately, then my primary constiearch doctor. Yeah. So I think that's going to become more common. And I think that's a cultural shift, not a technology shift. It didn't require copies amounts of new technology for you to become an active consumer in your health. And I think you'll see that happen across industries. Where I am skeptical is that the returns to this activity will accumulate and kind of new venture backed companies. I think people don't understand how robust the existing healthcare infrastructure is. And how easy it is for big legacy healthcare companies that have pre-existing distribution and long-term contracts to roll this technology out onto its customer basis. Like if you asked me to make a 10 year long, it's like a legacy cloud hosted healthcare software. It's not this new generation of medical adjacent kind of online health optimization things. I think fundamentally too much of the healthcare. So some of this just is classical payer-based in clinic legacy healthcare. And I just think the returns are going to accumulate there. I just think the distribution matters too much. Could I ask classification question? You're saying even though these are the old school models of going to someone and seeing something because they're the distribution to people, they'll win despite not being as innovative as some of these startups. Because they can suck out the chat GPT part and everything else is sounds fancy and fun but doesn't really move the needle. Yeah. And again, I work for one of these companies. These are my own on my employers, etc. And not investment advice. But like if you think about a company like Veridime or a company like Epic rolling out AI to the EMR, that is a much shorter gap than a company like OpenAI rolling out clinician-faced or a company even re-rapping OpenAI, rolling out clinician-facing tooling that replaces the boldness of the massive piece of software that is a modern EMR and practice management software. Like I think it's just much more likely that the legacy and convince will be able to use distribution to win here. I think we see that with Microsoft and Teams. I think it's a very similar story to what's happening in healthcare right now. And also just the fundamental data and workflow advantage these companies have to be able to give feedback to these models is just you can't replicate it. Where how about impacts on science more broadly? Zach Weinberg was actually on one of our recent pods. Yeah, he seemed pretty, actually pretty bearish on like the impact of AI specifically around getting drug to market new science. My view on this stuff is always that if you had a human being that had read as much as these models have, you would suspect that they would at least make some kind of novel connection between things. That you would at least be able to make metaphors across domains or at least make a jump. Like there's plenty of bodies of literature where there's obvious next steps that just no one has carried out. And the fact that the models haven't been able to do that I think makes me intensely skeptical that we're going to see automated scientists at scale making novel discoveries. Yeah, I also think it's fundamentally no one understands what the drug discovery process actually looks like. It is like not really a thing. We're simulation at scale really matters. It's a thing where like things like real-world evidence can improve speed to approval. But it's not like we're going to have server farms screen in mallet. We already have too many molecules. Like we have too many hits. Like it's not a thing we're going to be able to automate at scale. And I think automated science is just unlikely based on what we're seeing out of the models today. You mentioned Cotsier's medicine. We've actually I was just chatting with someone who's telling me about Atria. So that's the company? Yeah. Yeah. The Atria and Atria. Yeah. I think it's maybe a little maybe like a hundred grand a year. Super high-end they own the entire sort of like everything from imaging and testing to the doctors on staff themselves. Like yeah, I mean given A.I. And I mean the fact I've already more or less replaced my primary care with A.I. Like how do you think about that in relation to a really high-end Cotsier service? And then there's obviously like function, middle different views on function. But so I've done pretty much every Cotsier service. These are like a deep fascination for me. I think the kind of like 30 something worried well that are paying cash to get health care by choice is like one of the most fascinating demographics in the world. Like I am like healthier cross-eparriac back. And I'm spending like 20K 30K a year to like receive voluntary health care. That is like crazy and that's the consumer spending pattern we haven't seen before. The my interest in the stuff started with the executive physical programs you see out of academic medical centers like the Mayo Clinic or Cleveland. I've written extensively about these. I think these are like very fascinating indicators of where health care is going to go. But I don't necessarily see them scaling. I do think what I worry about is in the kind of below the bleeding crust of something like A.I. where you're essentially going to an a-mon property and they do ex-ci-fi health care on you. That that middle concierge market is just going to be competed out by the models. Like a like $5,000 a year concierge doctor is like kind of roughly equivalent to you getting some blood work asking chat GPT for interpretation and ordering the medication you need through a service like General Medicine. That will just route you to a primary care doctor that's relevant. I think this is going to be really ushaped where you see massive returns to the quality of care that's being delivered at the low end. And massive returns to the quality of care that's being delivered at the ultra high end. I just think that middle is going to get squeezed from both sides. That's how I do think this like a Hubertman branded ultra high end care is one of the greatest businesses ever built. Especially if you can get that distribution you can get that customer base but it's going to become increasingly rare. You seem very well read on the history of venture and finance and yeah curious to hear where like your interest is peaked and then maybe we spend some time talking about like where venture is today and I thought it'd actually be fun if we could like design a new venture firm together. Like we actually start. Bees are will have to pitch herself. I'm going to pitch is the LP. I would like to bet in the next generation venture fund. Okay, sounds good. Yeah, so if you think about like the history of the world getting better, right? Like big part to the reason why the world doesn't suck to live in today is because people were able to get exot outsized returns by putting money behind risky ventures. Like basically the history of human progress is predicated on the history of like efficient risk capital formation. And I think we spend a lot of time focused on the innovation layer and not like nearly enough time focused on the capital formation layer. Like open AI is like equally a capital formation innovation in addition to a technology one. And almost I would argue a harder capital formation challenge than a technical one. Right, so if you're interested in a world where like things are better, you would be interested in a world where it's like easier to accumulate risk capital behind exciting ideas and it's really efficient to do so. And my my thought on the state of venture today is we've ended up in this like very weird local minima where the model is both simultaneously preventing very important things from getting built and also destroying itself. Like the product is just simply driving returns to zero. And I think it's very clear that we're in a very unstable equilibrium in terms of where LP dollars are going to go over time. And I do think like if you care about the future, figuring out how to underwrite risk is very, very important. And I think there is a product that comes after venture that looks a little bit like venture maybe looks a little bit like merchant banking that will absorb the vast majority of risk dollars in the ecosystem today and dramatic and be dramatically better at underwriting highly highly technical risky endeavors for the future. And I think figuring out the physics of that thing really matters. I'm gonna stop you. I love the fact that you use the term and banking, which is as an oldie, love that. Love it. Love it. But people might not be familiar with it. Do you mind defining it so people get what you mean in the difference between that and venture capital? Yeah. So I think I think Alan and Company is like the world's greatest business. So maybe it's like useful to think about Alan and Company as the canonical merchant bank here. So the way to think about Alan and Company is there's kind of two halves to the business. There's a hard dollar half and a soft dollar half. The soft dollar stuff is like transaction advisory services. Like you can hire Alan and Company to buy another company for you. You can hire Alan and Company to advise you on a capital formation event. It is a services business that is fee based where they're not putting out partner capital to do it. But they're just staying around the hoop on interesting transactions. Maybe they're running a conference business like at the Sun Valley. Maybe they're running a media business like TTPN. But there's a bunch of soft dollar transactions. And then there's a hard dollar half of the business, which is the partners putting up typically firm capital behind proprietary opportunities that they source C and run intent. What I think is incredible about the Merchabanky model is by having capital that is long duration. It is partner capital or close to partner capital, right? LPs that are in end for 100 year durations. And also not having very strict LPAs. You're able to invest in a much wider remit of opportunities. I think venture is at peak special situation right now. I think everyone's interested in complexity and structure. I think that's a lot of that will wash out. But I think what won't wash out is that you will always have outsized returns. You're able to be creative about the opportunities that come in front of you. And you're able to see them as a result of this core business, which is the soft dollar advisory work. So I think Merchabanky has kind of an inversion on venture today. Seed investing as a loss leader. Flip that to a Merchabanky model where the firm has long duration permanent capital is very closely involved with management, hiring and firing of leadership down in it, and also able to do more dynamic capital formation that's better aligned with the needs of the company. I think that is just like the world's greatest business. And the longer we spend not doing that, it's just like it's a destruction of capital on every side. Maybe that's a silly question. But why do we think that because it's partners capital, it's necessarily longer duration and more aligned with kind of long term? Because I'm thinking there's a junior partner. He's not super rich yet. And he's like, I don't know, let's invest in that thing. It's going to be a 10X. I'm going to make a quick buck. I think about, there's risks with large number of people putting money in a pool at a very basic level. I just don't think you need large numbers of people. Yeah, I think my argument here would be that Merchabanks are somewhat definitionally boutique that the scaled asset managers of today, the enderiesons, the general catalyst, which are building incredible firms. Many of them I think have tons of equity value. These things will be well traded public names. But they're not boutique. They're running like banks, right? Like getting funding from one of these platforms is much more similar to applying for an SBA loan than it is engaging with a Merchabank. If you like work, if you think about the Pstonic, like IGSB, Allen & Company, or these Riteak Merchabanks, you have very small partnerships that are interested in building a vehicle that lasts for generations. The part of the reason you see these Merchabanks often so tightly held within families is because it solves an incentive alignment problem. I just think the entire venture industry is drunk on the two and 20 fee model. Right? Like if you were a solo-gift, yeah. Accuracy, two and 20, we love it. There are some amazing funds out there that still stick to two and 20. There's a whole lot of them that are way more expensive. Which is crazy, right? I'm clear with the fee. I mean, I think a lot of people do not actually understand that the fees have to be repaid before you get carry or and/or they don't care because the fees make so much money, the carry becomes superfluous. But two and 20 is a little bit of a, not a total unicorn situation, and venture, but it's definitely harder to come by. I know. Nick, is it fascinating? Am I the last one? Are you afraid of two and a half people? I'm not. Fred Wilson said this all the time, so I can say this publicly. Union Square is two and 20. And they say you make a ton of money, two and 20, you do not need to change that. I think if you're a primary metric as capital velocity, whereby dint of just getting dollars out the door, or you've got a private jet in 10 years, you will never have incentive to pull those dollars well. Versus a merchant banking model where you're seeing suppressor limited fees because of the partner capital, it's long duration capital. And you're able to do things that are more interesting, more dynamic, more aligned with the interest of those dollars. Then if you just sat on top of a big pull of capital. And like look, two and 20 is great for these big platforms. I just think it's not the business I would spend 80 years on. Before we get to maybe new models and exploring merchant bank, I'm gonna, I'm gonna love when we end this podcast. And we just built a merchant bank. Peezer's gonna anchor it. - In. (laughing) - Like when does the, I feel like we, I don't know, I feel like I've been having this conversation, hearing this conversation for like a long time. - Founders, I guess it founders, some care some don't. You know, if they're on the receiving end of like the capital stream, maybe they don't care. Like when does this break? - I don't, I don't think it breaks. - I think we're waiting for a break moment. I don't think venture is gonna go away, right? Like I think within three years, it'll be three to five publicly traded. Venture asks that manager, is that a multi product that are managing north of 50 or 60 billion dollars a piece? Like in the same way private credit can have its systemic flaws and Apollo is still a well-traded public name. Venture is still an incredible business to be in. I do think for the marginal LP dollar for the marginal partner that is starting a new firm, the best and highest use of their time, their calling and their capital is not the traditional venture model today. It is likely this weirder new structure that looks a little bit like something that looks like merchant banking. I think we're approximating it right now, but people raising special sets vehicles. I think that's like halfway there, right? But I do think what matters here is duration. And right now we have a massive duration problem in venture. - I don't disagree in the duration problem, but I'll answer the break question. I don't think it breaks technically until there's enough LP's that don't wanna fund the venture business that the tap is turned off. And that can happen in situations like what we're in now, where there's structural changes, right? If the tax rate, who knows where it's gonna land, if that changes things because frankly, it was the allowing of tax incentives that it originally opened up, the floodgates back then, when it was probably I don't know, $5 billion a year in business, whatever it was, tiny back in the early, it's tax changes. - And we're specifically talking about tax for the universe. - Foundations and endowments. And then your tax bill are looking at a significantly higher tax on different kinds of income. And so you have to figure out what falls in and what falls out. But also so that's one thing where their cost of capital is changing, but you also have, what I think you're getting at will is, and you and Nick and I, you've talked, you would be talked about this a lot, is if you don't get the return on your dollars, the equate to the risk that you're taking and the ill-equity that you're taking, LPs could vote with their feet and say, it's just, I'd rather do credit, I'd rather do real estate, I don't know, I'm gonna go buy art, whatever it is that works for their risk, reward and dollars, and it's all those things, but I don't think the system stops until LPs stop putting money in, but it's a very slow, and then it just gets smaller. I'm in your camp, I don't think it's going away. I don't think it breaks and then never comes back. I think it just gets smaller, and then the question is, what does a smaller venture world look like, but we're already so heavily capitalized? I think you could, I'm making this up habit, and still probably be great. - Yeah, I think you could cut half of LP dollars out of venture today, and it would kill firms that you and I have never heard of. And I think if you think that like AI is gonna smush the return curve, where like there'll be much fewer zeros, but a lot longer ones along that long tail distribution, and the biggest names will get bigger and bigger, that new return curve is gonna require fundamentally different capital across the life cycle, and you'll see LP dollars flow with it. - I've always believed there's probably some structure where like you have a fun, you take management fees, if you lose the money, like you owe the management fees back. Okay, but the kicker is you get like a way higher carry. - I mean, that's like the ohera trade, right? Like you buy a bunch of options, you buy a one-way ticket out of the country, and if the options hit, you don't board the flight. - Right, I think like some of that is like gonna happen. Maybe we should put more venture managers in prison, I know now. - Yeah. - I wanna see some, I guess I wanna see some paid. - You know, I wanna see some paid. - There's no paid. - I think like, there's no paid. - Merchant banking kind of, you look, with the Merchant banking model, you put your name on the door, right? Like Charlie Allen's name is on the door, and you're like, I'm burning all the bits. Every dollar that I have is in this vehicle. I'm not hiring Epic, not hiring fatality, not buying a fancy house. I'm living in like, scar sale. And I'm putting every dollar I have into this vehicle. And if you wanna come on that ride with me, I'm gonna manage this money over the next 60 years. That's pain. And that's pain that like numeric incentives, and like fundamentally I think culture is way stronger than contracts. And what's powerful about the Merchant banking model is like you're putting your name on the door, you're putting every dollar you have into the thing, and it's a partnership that's gonna stay tight together, ideally for generations. That's fundamentally different. like locking up the marginal GP. Going back to Merchant Bank, you know, I, you know what, it made me think of, do you know, do you know Aaron Harris? Yeah. Like who's Aaron Harris? For me and the rest of everybody else? Aaron Harris is, I'm sure he won't mind. I'm sure he'll like this. He'll edit it out if he does. Yeah, no, no, I think he'll love this. So he's an XYC partner, was a founder of a New York startup a long time ago. And is I forget what it, I think it's, I forget what the company's called, but he's essentially advising like Series A and B startups on fundraising and how to raise capital. And he's run like Series A, I don't know, he's sort of operating like a little boutique bank effectively. I think he would say that actually. Where do you think that fits into the ecosystem? And why is the Merchant Bank critical than just literally managing your own capital? Because I think those are two very different pieces, right? Because if you think about, I talked to a manager yesterday, I had drinks with him that describe any managers like, you know, $20 billion of really high end endowment dollars. And he described his job at seed stage of buying a basket of lottery tickets, indexing the market, broadly enough, hoping to get into a name that you can play all of a billion dollars into at scale. And just like fundamentally, if that is the business you're in, I just don't know how you wake up every day, like that is just like not. You're both new key dollars on the LP side and you're just like, it's just not, there's no taste there. And I think what's so unique about the Merchant Banking model is that the core of your activity, the big broad funnel stuff of advisory plus transaction, how plus capital services, plus pro services. Like I think another interesting mold on this is kind of what Finn Barnes is building into the GP, where though like deploy engineers for you, where you can just stay around the hoop on a lot of things in a way that's profitable to you and doesn't require you to invest to stay close to management. And that enables you to monetize those relationships through multiple ways, right? M&A advisory services, engineering deployments, you can even imagine doing something like what Jeremy Kaplan is doing with Octave or your buying companies that you're doing advisory work with that are just upside down and the cap tables need a little help. Right. But it allows you to stay around the hoop on a big variety of transactions in a way we're just saying like, you know, we're an evergreen fund that's mostly partner capital or close to new dollars. You're still fundamentally engaging in the business adventure versus this profitable core services business plus the hard capital investment and buy out business. I think it's central and you see big venture funds moving to this. Right. Like the thrive is becoming a Merchant Bank. Can I ask you when I'm stealing from a little later on in this script? Actually, Nick, I think I'm stealing from something you were going to ask, but I can't help myself. Could you in this next generation model call it Merchant Banking next gen? How does AI play in your mind? Right? Because there's some things in the services, which I love the idea that you're saying it's so it's basically so broad. It has to be human and there's some parts of it. They'll be automated. And I'm thinking of your concierge model, but I'm betting you've mapped that to venture. Yeah. So I think broadly, the only thing we should care about is the shape of the return curve. Both private and public markets broadly have a pretto distributed return curve. The vast majority of things return basically zero or the index. Very few things to go out. Venture has traditionally been a pretty steep return curve where lots of things are zeroes and a couple things are like big but not massive. Like it's new that we have like forward motor company size returns from privates. If you think about what AI does, it makes it cheaper to build commodity software and much more expensive to build great things. So it's stretching the return curve. It means way fewer companies will die because you just don't need them. Many people you don't need that much go to market. You can build. There's going to be way more companies doing $10 million a year profitably. But it also means the companies that the far end of that return curve are going to be much, much, much, much bigger. Right. Like 10, I think we'll see corporations kind of a hundred times the scale of the ones we see today. And that means you need different capital at each stage of life cycle. So if you think about the venture business, we need to find ways to drive liquidity for things in the kind of $10, 200 million run rate area. That's an unsolved product today. Right. Private equity doesn't want to buy those continuity vehicles. Maybe solve this, but you're not going to divot down cash out forever. I view that as like the jerk week of fun rock star alley. Like there are going to be a lot of funds to do special sets up by things out in that alley. And then if you think about the far edge of that return curve and also by the way, that's a great services motion in there where you can do emanate advisory work, capital formation work, buy out work. And I think you'll see a lot of services companies, accounting firms, human based things, get bought and trafficked in that kind of middle road where you're cutting human labor kind of what you're seeing out of thrive permit holdings. And then on the far out of the curve of these like mega corpse, this looks a little bit kind of like how founders fund operates today or maybe like a one three seven operates today where like your only job is to get additional dollars into these names at all costs. You don't really care what the product you used to do that is. So if you think about the life cycle of a firm, it splatters all the way across it. And the remit of investable opportunities, which is I think the most interesting part massively explodes because you can do things that are not, you know, roll up 20 software companies. If you believe that return curve is much more vast than it is today. I mean, that's a question from a from a founder perspective. You know, you mentioned manager, unnamed $20 billion fund saying he's going to index every single scene stage startup. I take personal offense to that. Maybe him. That's just me. No one else cares. Maybe founders don't care. Should founders care? Would you care raising your next, you know, if you're raising capital for your next startup that you're part of some massive index for some big firm with that maybe think you would think about different capital options. I think like I haven't bought a house. I assume you both have, but like when you buy a house, you're not really offended when a mortgage lender turns you down. Right. There's like 25 lenders you go to. You run a process and get rates. I kind of think the big commodity platforms are going to feel more like mortgage lenders over time. Maybe there's someone like the like personality filter on top of that. But like fundamentally, if you're like a company doing 10 million a year or growing 50 percent, there's like 20 firms you can go to and run a relatively standard process without a relationship and get a cluster of term sheets that largely look like each other. And like brands going to play a role on some edges. And I think some people will select. But how about how about a click? How about a click earlier? Like pre you know, before there's data like you're you're thinking about raising capital again, you know, for the next company at seed. I just think a lot of that's going to become operators. Like if you think like if I had to do this again, I don't know what a big platform could offer me at seed. But I know what my friends that have done this before could offer me at seed. And if you think about people that are kind of scaled operators, a lot of them have somewhat permanent capital behind them. I would might I think the future of seed will be a lot more of like, you know, your friends with some permanent LP capital behind them. To do interesting opportunity is like I kind of think that the seed industry just becomes a concomprter of scout funds and people investing out of their own PNL. I mean, I was talking to a very young fan of accuracy here. Well, your perspective as well. Like I was talking to a really young fan of the other day and like, you know, whatever he's 20 years old. He has not raised capital. He was not concerned about raising money. Like, and he probably shouldn't have been like like the constraint is no longer capital. I think for like the vast majority of the constraints are different. The constraints are because he knows he'll get funding or because he doesn't need it because he can bootstrap. He's really both, but he is very high certainty that he will raise funding and he's probably right. Yeah, I just can't think of the last time I've seen a company that would be great that died on the vine because I couldn't find formation stage capital. And like I feel like I've taken a couple of these on as projects and then down the road, I always find out there's like some reason why it couldn't find formation stage capital that I missed and the company is like actually unfundable. I can't think of like a great CEO even from the far stretches of God knows where. So far outside of network that's been unable to get sufficient early stage capital. The constraints are different constraints. Like I think about like New York in 2013, like if you were able to raise money and that was an if like you that was actually a legit advantage. Like the fact that you had two million dollars in the bank or three million dollars in the bank was a legit advantage. The idea that that like three million bucks in your bank today would be the thing is like comical. It also just feels like Twitter is like kind of the country club of tech where like the networks are much more public and much easier to get into like my entire network and tech is from Twitter. Like I never lived in the bay for longer than a few months. I like lived in Boston for most of my career. Like I do think online networks are way more permissive to break into than real world. Country club networks. Why do you think there's we should we should wrap soon. Thank you for spending time unless you want to go for another four hours. But I'm I don't think I can stay for 10 hours. I can have a run TV P.M. Yeah, let's just be like we that's the new goal. It's the new thing we're doing this daily. Yeah, yeah. It is funny. There is like unnamed podcasters today trying to do a daily show and it's really funny to watch them do one episode and then tweet and be like God, that was exhausting. I can't do that anymore. It's like, oh, the media is hard. Yeah, turns out doing the work is a mode. I'm a diamond. Really? It is it is such a mode. If someone came in and pitched a merchant bank attached to a venture or maybe Okay, someone come, Will comes in. He's built this really great murder bank. He has this amazing group of partners that do most of the investing themselves and they're like, these are great news. We're gonna, we're gonna take on one new outside LP and we've chosen you. - I'm excited. - What do you do with that? (laughing) - Oh, I love these thought exercises. I can tell you why a lot of LPs would block. It's because they have buckets for their dollars and it doesn't fit my venture bucket. It doesn't fit my whatever bucket. So, but there are LPs that say, I'm looking for a return number and if you hit my return number, your structure is, is and I, so I would like to run the numbers based on that concept, but at the end of the day, depending on what your mandate is, you also have to foot with your mandate. But yes, I think there's, there's definitely a lot of LPs out there that have a return hurdle and it's the question of how it works and then all the usual diligence questions about people and structure. And I'm also thinking, I know this is not what you said well, but we have seen businesses where it's more of a media component and that is not the venture backable side, but it is, it's its own business. And then there's a venture fund associated with it, which is a bit of a work version of what you were pitching Nick, but we've definitely seen those get funded in venture land. - I believe in that's like, I think kind of the future. Like the Patrick or Chansey Harry stepping thing, I think is kind of like the first step to a merchant banking model. I think it has not been fully exploited yet. Like on my elite media take, I think like these, they're not lofty enough, yeah. And they're often backwards looking. I think there's like an immediacy to TPP and that puts them in the hoop a little closer. And like if you think about where that thing can go, I wouldn't be shocked to see that become a merchant bank. I wouldn't be shocked to see that become a wealth management business. Right, like there's 16 different products you can build on top of that. But I do think your point on like, this is not easily buckledable, was correct. And I think a lot of LPs are gonna make a lot of money by kicking in things that are outside of the bucket in the next five years. It scares me a little that emerging managers, capital E, capital M have become their own asset class. And there's like a treadmill. And you can like fill out the forms and raise like 20 million dollars. It's like, oh, that means those returns are gonna drive to nothing. Like that, the second it's become that industrialized, it's gone. - Let's hope not. On behalf of everyone who's raised a 20 million dollar fund, but actually, when is well taken? - No, I think the takeaway is build a gigantic media conglomerate ASAP. - ASAP, you know, you might be too late. You need like, you need two brothers. You need two brothers. They need to set on a mic for five hours. - I'm trying to, two siblings. Let's be a little open. - In the S&P, that's the game. Wait, one last question. What's the, what's the merchant been called? - You have to put your name on the door. - This is a great minute. - In the meantime, it's a great fun take. But I think any venture fund with its name on the door, you should trust. There's a reason, injuries and horror, what says a good reputation. They're putting their name on the shingle. You should be very skeptical of scaled asset managers, where their name is not on the shingle. - Oh, but I can think of two major examples, three that don't, four, you can. - No comment on the ethics of those individuals. But I think you gotta put your name on the shingle. (laughing) - On the spicy note. - Thank you, well, we appreciate it. - I saw it. - Appreciate it. (upbeat music) - Thanks for listening to Origins, the show where we discuss the venture ecosystem through the lens of myself, a GP, and Besar, and LP. - Be sure to tune in next time for a bonus mini-sode where Nick and I unpack today's conversation with Will. - You can find that right here in this feed, and while you're there, please rate and review the podcast. It's super helpful for us and other folks looking for venture insights to find our show. I'm Nick Turls. - And I'm Besar Clarkson, and we'll be back soon. Thanks for listening. [BLANK_AUDIO]