Hemant Taneja & Packy McCormick on Rethinking Venture in an Era of Concentration

Thanks for doing this.

It's wild out there. I walked into this here and we had this conversation with the team where we said we just have to fundamentally recalibrate what scale means, what are we really striving for. The data points are when a company like Nvidia can add a trillion dollars of market cap in 100 days, or Anthropic, which we now have a very large investment in, is adding the kind of revenue it's adding every month, billions of dollars, 10 plus billion dollars often. You sort of say, wait, the scale that we're operating at and the concentration of value that's going to happen in a handful of companies is pretty stunning. And so the correlate to that is what are the founders of this generation supposed to be doing so that their work can actually matter. And I tell you, when I first got in the business in 2000, the whole goal... our first institutional fund was 200 million bucks and the goal was invest in some company where you own 20%, sell for a billion and return the fund. 10 years later, the word unicorn was like a measure of scale. Last decade, it was all about can you get to decacorns. But today when you have multiple, one, two, three, four trillion dollar companies that are going to be five, six, seven, eight, nine, 10 trillion dollar companies in a few years, building a billion dollar business or a $10 billion business, I think it'll be financially interesting but not compelling for founders to be their life's work because great founders want to make a dent in the universe. And so just rethinking what is a scale that our companies have to get to, how do we get them there in a world where there's so much momentum towards these large companies vertically integrating into every dimension, every profit pool. That to me is what the focus of the last 6 months for me has been.

Was it controversial?

I think there's a couple things in that. So I'll tell you, when the correction happened after COVID, the very first thing that I did, and Paul Fielding was our CFO, he and I sort of decided we're going to do something draconian, is we marked our portfolio down by 40%. And our LPs told us we were the first firm to actually do that. And the reason was we didn't want our investors thinking there's value in these companies at the valuation that were getting done in the exuberance of sort of post-COVID that everything is going digital. That wasn't there. And the second is as investors, you don't want your next generation of investors coming and thinking, look how good an investor I am because I've got this markup. And so we took it down, became rational. We did it again earlier this year when Claude happened and the whole terminal value question was being asked of software companies, which is a very reasonable question. So the reason I wrote that was actually for the next generation of investors to understand just trying to get marks in the heat of the market isn't really the best way to create value or be helpful to founders. And if I go back to my comment about scale and the fact that these companies need to get to be a lot bigger, you're not going to get there by budging them with equity. So the whole point there was rather than just thinking about, because every time founders raise capital, you're taking something away from them. You're taking their stock. And so how do we get them to the promised land without having to continue to do that when things are way more capital intense? You look at infrastructure stuff, you look at resilience stuff, you need a lot more capital. The answer can't be just keep marking up stock to finance these and we need an innovation in the industry. So my point was more get everybody in our industry to think like entrepreneurs ourselves. I think we get complacent that we sit on a pedestal and we get to play God and say yes no all day long and give you money. No, we need to be entrepreneurial about how do we serve these founders in a much more compelling way given the needs are changing and given the scale they have to get to is fundamentally changing.

So that's actually... that's not what it was meant to say. So for us, we look at our venture business, we look at ourselves as a seed firm just like you run a seed business. There's a reason why we have three senior partners that built their own seed firms at the center of it, which is, hey, we really want to be seed first. So that relationship and building that trust and ownership in the beginning and navigating ambiguity is what venture is really about. The point though is from there, the journey that these founders now have and the capital they need and the sophistication of capital they need isn't just about growth equity. And we have innovated some. We created customer value fund, we just raised for a global resilience public vehicle to be able to take some companies out. So we're innovating, but I'm telling you, I sort of sit there and say, we're not innovating nearly enough. And so the whole point was to actually just say, hey, everybody start thinking about innovating. So we have more capital innovation. And this is not about financial engineering. It's about financial innovation.

Who's going to do that if it's not our industry? We are the ones that are closest to what the needs of these founders are, and I just don't see enough of that. I think it's because we're complacent with the markup game because we think that's good enough, and I think we're not serving these founders right if we're not pushing ourselves harder.

Look, there was a version of this that happened before as well because most venture firms, including us, were early stage venture firms and you had these public hedge funds. They came down starting crossover investing and there was this whole idea of is the market sort of reinstitutionalizing into platforms. Well, who's going to capture most of that value? Folks like us that fundamentally understand technology and where it's leading in the world and have to go get access to more scale of capital, or people that are trying to commoditize providing capital to founders because they have scaled capital. And I think if you look at the dynamic around that in the last 10 years, I would actually argue that there's a handful of firms that have really innovated and are starting to do things at the larger scale. And the reality is, frankly, if I think about alpha that needs to be created in this world of AI with existing businesses, that's where this really applies. So much more of the alpha is about taking AI operationally, innovating, and then driving abundance from a product perspective to serve your customers. Way more of it comes from that than can you actually cost cut and financial engineering. So I would say the advantages to firms like ours that kind of grew up in the era of truly driving, building the world's most important companies, not optimizing profit. So I actually would tell you, I've had conversations with many of the large PE firms, they themselves don't see a future in traditional PE anymore.

Yeah. So look, when a great founder starts a company, their ambition generally is to go change some industry. And our view is that when you want to transform an industry, no single company can do it. And I said this when we were deep in the healthcare work in the COVID era, and I was like, the Amazon of healthcare is not going to be a trillion dollar company. It's going to be a trillion dollar ecosystem. So our theory, for example, if I go to healthcare for a second, let's take Summa as a living lab, which is the Akron-based system that we bought. Our belief is that if we want a proactive, affordable, accessible healthcare system, we have to turn these local institutions, these hospitals, which are poorly run, they're all teetering on bankruptcy, into AI-native hospitals that are profitable, that can invest in their communities and can drive that proactive care. Now, when I think about, okay, we have to go do that. Let's go buy a hospital. I sit down and say, what are we going to do with it? There's no company that can go in there and go accomplish that. So what's happening inside Summa right now is we have our own AI team, Percept, and we have I think eight of our AI healthcare founders that are targeting different parts of the stack. They're all working together. And when you look at the corporate metrics that transform a hospital like that into the future version of what I think it needs to look like, there are work being done by different companies that toggle each metric. And so you all of a sudden get this operational view of this AI-native hospital where all these companies are contributing, and they learn a lot from doing that at enterprise-wide scale because there's not the inertia of, oh, we're going to do pilots and do incremental stuff and years of BA and certifications. No, we're going in there, and as long as these solutions are proven, because we don't want to be irresponsible with people's lives, which is the only kinds of companies we bring in here, we're going all the way, but we're going to go all the way as an ecosystem of companies. So I think this collaboration between founders to say we're going to together transform an industry, these living labs are necessary environments to be able to go do that. Janus is another example. We're trying to do the same thing there.

Well, the three most interesting facts about... first of all, LeBron James and Steph Curry were born there. Then, Alcoholics Anonymous was actually started at Summa. And so I think it's actually, you know, we forget these institutions that are around for 100 years. Those towns were really vibrant and they were very innovative, and just because of the tax of the way the healthcare industry developed, these companies that are still the largest employers have just become economically diminished as businesses, but there was a lot of great legacy in these places. And so a lot of our focus is on, hey, can we take an institution like that, bring in our set of founders, and take that institution to the future in the way they should be delivering care, to take care to the community.

If I was to think about just healthcare, I assume that's what you're asking. So I think there's a few things. One is because of the way the system is set up and there's no trust between insurance companies and providers, hospitals think insurance companies don't pay enough, insurance companies think hospitals bill too much. So there's these armies stood up on both sides, there's this enormous work tax in an industry that already is bankrupting us. First thing is can you just make that go away and create sort of an AI-based system that just eliminates how the payments really flow, as an example. The second thing is the burnout of the workforces, which was already pretty bad given the EHRs and the way we digitized the industry, then got extremely bad because of COVID. So that workforce is burnt out. It's actually in fact the reason why healthcare is the fastest adopter of AI, because the workforce is so burnt out. We need to get them so they can get their bearings and can actually take care of the people and we're not triaging resources. We triage a nurse capacity to put it on the sickest people as opposed to we should have abundance around it. So that's like a really interesting and important leverage. And the third is the patient experience. How do you actually make it so that they can navigate the system and they can understand and take care of themselves proactively? Because the best hospital system should be one where the economics create the dynamics that nobody comes to it. People want health. People don't want sick care. And so these levers that move these dimensions are... and again, no single company can do that. But this ecosystem of founders that we're working with together can actually go solve that problem.

Look, Janus is another 100-year-old world institution. It's got about half a trillion dollars it manages. It's got 65 million retail users all over the world. And if you think about the world of haves and have-nots, these retail consumers don't have access to what's going on in our world. So first thing is, can you democratize opportunity for them? Can you change the experience and increase their sophistication using AI? And that's sort of the way you're going to transform the experience of the end customer. But obviously in order to do all that, you first have to make Janus itself AI-native. So again, Percept, our AI team, is working with them to operationalize sort of an AI-based operating system at Janus, and then we start to think about innovating so that their customers can have abundance of opportunity where they can invest. And so I think it's another example of a living lab where we can say, just like what is the health experience consumers should have, what is the wealth experience consumers should have in this next era.

Well, look, I think there's definitely flywheel effects for us because if we learn... first of all, if we have the economic leverage in asset management with a platform like this, we can actually go scale the platform, buy more asset managers. If you think about where's that capital going, it could actually be made available to the capital entrepreneurs in our industry. And if those capital entrepreneurs are accessing this capital, then by default they're giving these consumers access to our industry where so much opportunity is getting created. And so I think that's very top of mind for us is in this world of concentration, how do we really diffuse opportunity and create prosperity for everybody, because I don't think we're going to be allowed to do it any other way by society.

Well, I don't think we are focused on trying to concentrate. I think we're trying to create conditions for founders to become big enough, then as an ecosystem that is sort of the counter-concentration, of our architecture for innovation, so that that can happen. Like if you are trying to innovate in the space industry today, think about the kind of moat SpaceX has. Do we think the space industry should be run by one company? If not, what are we doing to create conditions for founders to succeed and democratize that opportunity. I'm obsessed with this. We have to solve these problems. I really admire the folks that have created this kind of concentration. Again, we're investing in them. But part of what makes capitalism work for society is a level playing field where new companies can actually continue to emerge. I mean, that's for you as well. And are we really making that infeasible in today's environment and how do we really challenge that? So like what do we need to have in terms of scale to create those conditions? That's the scale, that's the criteria for where we're focused.

Yeah. So I actually think the long-term view of the economic infrastructure of AI that needs to power society hasn't changed. So if you think about the continuum, you've got these frontier labs that are continuing to push the edge in what's possible, the level of intelligence, but every time there's a new model, any one of them, there's a certain amount of workloads in society that don't need more intelligence, and I think we're like probably 90% to 95% of the work that needs to be done that most people do in society, talking about knowledge work, can be done with the models today. So I think breaking that frontier to open new frontiers in research and having a company that enables that is very important. I also think for markets where strategic autonomy matters for different countries, they're going to want to have sovereign model capabilities that enable that. So I think that's really important. But when I think about an enterprise and you look at all the token spend and what's happening today, I can't imagine a world where it's going to be that we're standardizing on one of these. I think you're going to have... and this is why we invested in Cognition. I think you want to help these companies figure out how to shard their workloads into where do we need the sophisticated frontier capabilities like Anthropic, and where is open source good enough, and what's in between, and what can we build ourselves. So I think you're going to have this continuum and so that infrastructure, there's a place for open source. We invested in Mistral. So we're deep believers that ecosystem needs to exist and that this is going to exist. What I did not realize at that time was the pace with which they're going to concentrate. And then yesterday I was actually speaking at an event and it just hit me. I'm like, of course this was going to happen. I've been talking about it for a while saying this is the first technology where every CEO in every industry in every country is thinking about it at the same time. Well, if that's the dynamic, then every one of them is going to experiment with it at the same time. They're going to figure out their use cases at the same time. So you're going to see unprecedented growth and that's really what we're seeing. It just hit the zeitgeist like no other technology before has. It's wild to see.

What's the economic case for like an Anthropic as they continue to go? This is a very hard question to answer, and I can tell it's also a very hard question for Dario to answer. I'm projecting here, but the reality is if they said we're going to be just the platform and provide this intelligence but it gets commoditized by open source, then as he's been saying, unless we get to a trillion dollars of revenue, we can't have the infrastructure that can power these models. So that maybe won't get him to scale. So what's the next step? Figure out what profit pools you want to go capture up top. And when we invested at the $60 billion valuation, the bet was that cloud was going to be their app. Sort of in reverse. Usually you have a killer app and then you become a platform. They were a platform. They actually, in one of those rare cases, figured out a killer app. Then they've gone and looked at, hey, should we do stuff in legal, and given his biology background, they want to do stuff in life sciences. So the challenge I think for these companies as they migrate to have enough profit pool to sustain and feed the beast, so to speak, so they can actually survive and provide the good that they can provide, is at odds with how a platform should behave. And I think they need to figure out where is it enough that they can then also democratize innovation in other parts. And so I can tell you, the company is kind of dealing with that tension real time to figure this out.