The Anti-a16z? Vijay Pande’s Bet on a Tiny, AI-Native VC Firm

0:00Connie: Hi, I'm Connie Loizis. And this is Alex Gove. And this is Strictly VC Download.

0:15Connie: Hi, and welcome back to Strictly VC Download. Our guest this week is Vijay Pande. Pande started as a theoretical physicist, nabbing a PhD from MIT before landing at Stanford in 1999, where he built something called Folding at Home, turning millions of home computers into a supercomputer for disease research. In 2014, he talked his way into Andreessen Horowitz, convincing the firm to bet on healthcare and life sciences, a category it had explicitly ruled out, and built it into a nearly$4 billion practice. Then, a little more than a year ago, he walked away to build something smaller. His new firm, VZVC, co-founded with longtime investor Zach Werner, is a deliberate bet against bigness, you could say.

1:03Connie: It intends to make not dozens of investments from its new fund, but only a handful. It also doesn't feature associates or other layers. Because the two are building it from scratch, and because it's AI native, it's an interesting example of what more VC firms could look like in the future. In addition to the mechanics of the firm, we get into why he's betting concentration beats breadth right now. How a VC writing checks into buzzy biotech startups that are growing quickly thinks about how much money to put to work. And we also touch on a challenge that's central to the intersection of AI and biotech, which is that you can't scrape biology off of the internet the way you can scrape text.

1:43Connie: Every company has to generate its own data to a certain degree, but that means every model is trained on its own silo, raising the obvious question, is that fragmenting the field's collective progress? Or is there a real path toward biology's version of open source AI? Vijay's got strong opinions on both. It had been a long time since we'd caught up with Pondy, who is well-known in Silicon Valley circles through his time at Stanford and through Andreessen Horowitz. It was really great to talk with him. We hope that you will enjoy and learn from this chat too, and we will see you back here next week.

2:21Connie: Vijay, it's really great to be reconnecting with you again. I cannot believe it's been quite so long. Congratulations on your new firm. For people who've seen your name, know that you were leading Andreessen Horowitz's biotech effort for many years, but maybe you don't really know your background, can you just kind of give us the skinny? You started off, I think, as a physicist before winding up in biology.

2:48Vijay Pande: Yeah, absolutely. That's a fun question. So actually, I was in my first startup when I was 15 years old. Oh, wow. So it was me with my friends, Ed Gavin and Jason Rubin, building NoiDoc Software, a computer gaming company. It was such a great education to do that at a young age, where we were writing software that shipped on Sega Genesis and so on, but doing it basically three of us with electronic arts as our studio. And I think I fell in love with startups in that moment. but also went to college and grad school. My degrees, as you mentioned, are in theoretical physics. So I got a PhD in theoretical physics from MIT.

3:20Vijay Pande: And, you know, you might think like, well, what does that have to do with anything today? And I think one of the biggest conveniences of my career is that all the math and a lot of the, even the intuition from AI is a lot of the same math from theoretical physics, especially conness matter physics. So I've been working on neural nets even when I was in college. But at that time, biology was a really big thing. You might remember the Human Genome Project. And so there's a lot of excitement behind that. So that really pulled me from physics to biology. And combining those two loves of biology and startups, I wanted to be a professor at Stanford.

3:54Vijay Pande: And so I started there in 1999. And pretty quickly, people from Sand Hill, VCs from Sand Hill contacted me to help them with diligence and to contact the advisory boards and board directors of companies. And that was just such a joy for me. I was back in that ecosystem and so on. Most notably, I built Folding at Home, the large-scale supercomputer project, which basically allowed us to preview a lot of the issues that we now see in AI, especially the bitter lesson that compute is such a big deal.

4:23Connie: Before you go on, I just would love to index for a minute on Folding at Home, because I do think it's really interesting. And I don't know that this broader readership or listenership is going to know it.

4:33Vijay Pande: Yeah, yeah. So to set the context, and we found in 2000, it's something that's actually still going on strong. But the idea is that my group at Stanford wrote software that people download and actually shipped natively on the PS3. And by running that software, your computer became part of this large-scale computer. And we were actually some of the first to even do any sort of compute on GPUs. And so with GPUs and the PS3, we were able to build a very, very powerful machine at a very low cost.

5:03Connie: So incredible. So people donating their idle processing power. How many people were involved in this? How many are involved in it now?

5:10Vijay Pande: Yeah, so I actually don't have the most recent stats, but at a day day, this was like tens of millions of people contributing. Wow. And I think the fun thing especially was that, you know, I think people could really get involved and they could see sort of what their contributions were. But maybe the most poetic way to talk about it is that it was not just a supercomputer, but maybe a little bit of a time machine that allowed us to do calculations 10 years ago or 15 years ago that now are becoming routine. And that sort of allowed us to have a lot of creativity and sort of try out things.

5:41Connie: Phenomenal. So there you are at Stanford teaching, I guess, biophysics and involved in this or no?

5:49Vijay Pande: Primarily chemistry. But, you know, like the thing about professors is we mostly run a research lab and teaching is the thing that we do to pay the bills and so on.

5:57Connie: And how did you land at Andreessen Horowitz?

6:00Vijay Pande: So in 2014, I met Mark and Ben, you know, Mark and Drewson and Ben Horowitz. And we had a mutual friend and a business partner, Balaji Shrinivasan, who had worked with me in a couple of different capacities and was actually even briefly at my Stanford lab. And so Balaji introduced us. You know, we really hit it off. I think we saw a lot of things very similarly in terms of the future of technology.

6:20Connie: It seems like that required a lot of forethought because I feel like even in 2014, it was like just not something a lot of VCs were keen to dive into. I know that you've said that biology is moving from science of discovery to something you can engineer. What does that mean in practice?

6:36Vijay Pande: Even amongst many biologists today, there's a sense that biology is so complicated, there is no way that a human being could understand it. And I think that actually may be true in many ways. But I think what's shifted is that AI and machine learning allows computers to wrap their type of understanding on something very, very complicated. It's, I think, this AI machine learning shift that has allowed something that really wasn't possible before, either to understand biology, which is to try to figure out what targets you want your drugs to hit for specific diseases, to be able to make those drugs, and now even to help in the clinical trials, which are the most expensive part of the process.

7:15Vijay Pande: So you're shifting from pure discovery to something that is very much engineering, although it's engineering in a biological context.

7:22Connie: And when you talk about the clinical trials, you mean because so much of it can be done synthetically right now, so not as many people need to be roped into the process.

7:30Vijay Pande: That's, I think, very much an aspiration. I think the big thing right now is that the biggest cost for drugs is the cost of clinical trials. The cost and time to get to clinical trials has been shrinking, and especially with AI, but it could still cost like hundreds of millions of dollars to run a trial. And the end reason why drugs are very expensive is that the probability of a drug going from the first trial to the end of the third, the probability of a success is just 20%. 8 out of 10 fail. And so you can imagine if these things cost hundreds of millions of dollars and most of them fail, the amortized cost of drugs gets really, really high.

8:03Vijay Pande: I think the lesson here is that the reason why they fail typically is not that the biologist did something wrong or did something poor. It's that all the experiments that these drugs were designed on were on animal models like mice and other animals. In the end, these animal models are just not very predictive of humans. And I think where AI really changes this is that while it would be unethical to do experiments on human beings the way you would do on mice, the ability to take human data and bring it together in a model is something AI actually is really very well suited for. This AI model is not going to be perfect, but that is going to be way better than any animal model would be.

8:41Vijay Pande: Once it crosses that bar, that's where it gets really exciting. But beyond that, you and I care not just about does the drug get approved, but let's say if you or I were sick or one of our loved ones was sick, like, would that work for us? And where this starts to transition now into is even not just in helping design drugs, but now onto the other side, being a doctor and seeing, well, actually, is this drug the right drug for me? And that's something that actually also itself can be life or death.

9:05Connie: Can you, I guess, elaborate a little bit more? You're talking about the ability to personalize medicine or?

9:10Vijay Pande: Yeah, I think it's the jargon here is so-called precision medicine. If you've ever gone to a doctor with something not trivial, there's only so much they can tell. and then they give you a drug. And if sometimes that works, but sometimes it doesn't, they give you another drug, then another drug. And this happens in cancer. It happens in lots of different areas. And we would all be much better off if the first drug was the right one. This is very much a data problem. It's the data of reading your blood tests and understanding all the nuances and then understanding how you may be similar or different to other people.

9:39Vijay Pande: One of the big transitions we're seeing in healthcare right now is that when you get blood tests, typically your values are compared to population averages. but really they should be compared to like, well, is this weird for you? And what we're starting to do also on the medicine side is to be able to just understand what would be right for the individual.

9:56Connie: Just going back a little bit, you have been mired in this world for so long. I mean, really back to your Stanford days. We've been hearing about AI drug discovery and precision medicine for so many years. How would you sort of describe the process over the last 10 years? Was it slow in this sudden ramp up in the last few years? or was it just became more aware of what's happening in the last few years?

10:18Vijay Pande: Yeah, I think it's lots of different things. So for instance, precision medicine for the longest time was based on genomics. And the reality is like your genome is kind of like the blueprints for building our house from day one, but your house is fairly different now than it was from the moment it was built. And so there's other types of things that people measure in proteomics and so on that is much more relevant for understanding disease and where your body is now. And there's been now all this automation in robotic measurements that is naturally tied into AI. And those two go hand in hand really well.

10:50Vijay Pande: Over the last 10 years, there's been this steady clip. And in both AI for biology and AI for chemistry, the biology part is like, well, how can we treat this disease? And then the chemistry part being, well, how can we come up with the drug to go after that specific protein? There's been actually very significant advances over those 10 years. AI for clinical trials is the part that's much more new. And I think we're starting to see companies go after that right now.

11:15Connie: I just last week interviewed a founder. I don't know if you've ever crossed paths with him, Michael Polanski. He worked for many years with Sean Parker and was at Founders Fund for a minute back in like 2009. But he has this new startup they've been building kind of quietly in the background for the last few years. but they've tapped into a tissue supply chain, which I didn't realize exists. Of course, you know well that this happens, but they've found a way, they say, to keep tissue alive for at least a month so that they can test different chemical compounds on it and see how the skin reacts over time.

11:51Connie: That was my most recent peek into this world. I just thought that was so fascinating. What are you seeing that's really kind of like blowing your hair back in terms of things that weren't possible before that are possible now.

12:03Vijay Pande: Yeah, I mean, even just this last week had all these wild things. I think people are looking at these new cancer vaccines that are going through trials. That stuff has been in progress for several years, made its first real breakthroughs in terms of clinical trial results. One of the things that is maybe not as appreciated in the broader community is that we see all the AI advances coming up. It feels like there's another AI advance every week. I think we're starting to see analogous things on the bio side these days. Some of it's tied to AI sort of harnessing new ways of getting data. And I think you're mentioning a great example that people are bringing all this creativity to the biology side to get data that you couldn't have any other way.

12:42Connie: Yeah.

12:43Vijay Pande: And then that data is being fed into AI, which is now coming up with new ideas for even new types of therapies and so on. So like it's really bubbling up. And I think part of it, too, though, is a feeling like when people are looking for what the killer applications of AI is going to be, biology is interesting for a couple of different directions. I mean, obviously, it's mission central for many of us to help deal with this. I think it's something where intellectually the connection with data is really very natural. But then finally, I think if you just think of it from a pure tech way, it's a place where you don't have any of this data on the Internet that people can just all train the same thing.

13:17Vijay Pande: and your data can't be distilled from one model to another. It's a really interesting play from just a pure AI sense.

13:24Connie: Well, I wonder if that's what that means. Like you said, you can't really scrape this data from the internet. Everybody's finding creative ways to develop new data that they're feeding the AI and the AI is hopefully making better guesses based on the data that's being fed. But it kind of reminds me of medicine in a way and problems that arise from the way that medicine is set up, which is that it's so siloed. I feel like doctors don't really share as much information as they potentially should. And so that's why you often get like a second and third opinion. That could be very different. What happens when everybody's coming up with these very distinctive, different, smaller models or data sets?

14:01Vijay Pande: You're onto something really big here. And I think one of the common problems, and I've seen this with people I know where, let's say they have some type of cancer and it's both an issue in oncology and endocrinology. And those two doctors really don't sync together very well. And I think medicine is the way it is because it's done by people, obviously. No human being can be an expert in everything, right? What is really intriguing about AI is that it can, in principle, be a specialist in everything. And it can start to see things that really any single human being could be equivalent to having like a team of the very best doctors all collaborating together in that moment.

14:39Vijay Pande: That's actually really unique, and it combines with a movement that's been happening in medicine for a while, the so-called functional medicine movement, where people are thinking not about medicine as it's organized in the org chart in the hospital, but how it's organized in terms of the human being, which is that all these things are interacting. And so we have to think about it from that context.

14:58Connie: In terms of having these different startups sharing information to create these bigger, better, stronger data sets, you see that happening? Oh, yeah.

15:08Vijay Pande: Yeah, yeah, yeah. I'll talk about the ones I know the best, obviously. So I've invested in it, was on the board of InCitro for many years. And so this is Daphne Kohler's company. Daphne is one of the real OGs in machine learning and AI. So they really pioneered new ways in both getting data through robots and so on, but also how to analyze that data. And part of it is just to do all that acquisition and build on infrastructure takes a while. What we're seeing now come out of InCitro is just really remarkable that there's new advance after new advance after new advance. And so it's pretty much realizing the dream that she's laid out.

15:41Vijay Pande: That's one of my favorite examples in the biology side. I think on the chemistry side is a company I know very well, Genesis Therapeutics, where I was both an investor, but that company came out of my Stanford lab. So I'm a co-inventor of the founding technology. Oh, wow. And so there, this is Evan Feinberg. Evan's taken the things that we did Stanford and gone like light years beyond that. And he's combining both machine learning, but also physical simulation and experiment, which is really unique and tackling some of the hardest to tackle proteins to drug. Those are serious things that I think if we were looking at that 10 years ago and said, you know, this is possible.

16:15Vijay Pande: Hey, I think we'd all be high-fiving ourselves. I feel like this has really come a long way.

16:19Connie: So that's great for Genesis Therapeutics and it's great for NCITRA. and I have had the chance to meet with Daphne before, and she's so impressive. But is Genesis sharing its findings more broadly, or is that its value? Is that data set that it's created?

16:33Vijay Pande: Yeah, especially in these cases, both are designing drugs, and so they're really heads down in terms of designing those drugs. In terms of sharing value, that's where things get interesting in maybe a couple different places. One are initiatives like Chan Zuckerberg or the ARC Institute, where these are private foundations that have been funded to have a tech-like mindset, but that the data is shared very broadly. It also very much speaks to the obvious role for academia. And that what often happens is that academia, these ideas get built there. And then when people have a sense that this is something that could really be translated, then they'll go start a company.

17:08Vijay Pande: I mean, that's the history of Stanford very much.

17:10Connie: So basically, because of third parties like Chan Zuckerberg and academia, there's hope that there's going to be repositories of all this information.

17:19Vijay Pande: I think one of the things that is maybe the bigger, bigger trend that I think we're starting to see a shift of is to build these atlases for information of biology, which from the technology point of view, typically are foundation models. And as they become more and more common, I think they'll be spread. And much like you could imagine, we do have these open source LLMs that do very well against corporate LLMs. I think there'll be these open source foundation models in biology that will have very, very broad impact.

17:47Connie: That's great. I hope so. I mean, of course, I understand as an investor, you want to hang on to your secret sauce. But also as a human being, I'd love for there to be a lot of information sharing. So you mentioned Genesis. That person came out of your lab. Daphne, I'm guessing you probably knew from Stanford. How do you invest in a company when there's so much going on? And how has that changed since the earlier days at Andreessen Horowitz now that you're leading your own firm?

18:11Vijay Pande: Yeah, I mean, I think the process is very similar. And I think all investors have some things that they do uniquely. But I think in general, we follow very similar patterns of starting to really understand the founder and alignment with the founder and then the market they're going after and what their unique technologies are. So the combination of all those things. And that's really, I think, unchanged. What we always worry about as investors is like, are we sort of backing the right horse? And is this one going to be the winner? And that's always obviously going to be a judgment call. where it gets, I think, even harder today is just how fast everything's moving.

18:42Connie: What have you gotten right and what have you gotten wrong?

18:45Vijay Pande: That's the key question. So I'll do the good news first. So the right part is that I think when I started talking about AI and machine learning and technology and medicine and bio 10 plus years ago, there was a lot of resistance. And a lot of people saying, oh, yeah, that's never going to happen. It's never going to be useful and so on. And I think that resistance, I think, now is largely gone. and that you see biotechs and pharma and everyone is realizing that AI is going to be a key part of this. That's very fulfilling to see this arc move, you know, not just for my investments, but also for these are going to be real foundational technologies that sort of bend the cost curve for all this stuff.

19:20Vijay Pande: So I think that part is very exciting. I think if I were to pick one thing that I've really tried to think about that I wish I knew in the beginning and I think it took me some time to really appreciate is that as seductive as the coolest technologies are, it really always comes back down to the go-to-market. And what I tell my founders is like, especially the ones that are coming from the science or the product side, for them to take all their brilliance and creativity and really apply it to the go-to-market side. And that the go-to-market part is as least as hard or harder than the technology side.

19:50Vijay Pande: And so a lot of my investments in the last few years are ones where the go-to-market actually is really clever. So I think about function health as being a great example where this direct-to-consumer, which is, by the way, a lot of people would say direct-to-consumer health would never work. They are very much proving that that's not the case. I think Jonathan Squirtland there, the CEO and co-founder, he really thinks about go-to-market first and foremost, and then picks the technologies to do what he needs to do to get that done. So I think in our world, especially for what the possibilities are now, I think that's actually one thing I wish I thought about from the beginning.

20:21Connie: That's interesting. I don't know anything about Function Health. So can you talk a little bit about the product and what is so special about its go-to-market?

20:27Vijay Pande: Yeah, so I think the goal for Function Health is to give you agency over your health. You join as a member and twice a year, every six months, you get your blood tested. And then it's a huge panel of tests that if you try to get it from your doctor, it costs probably about$5 ,000 the way the medical system is set up. And originally function was$500 and now it's actually like$365, so like a dollar a day. The idea is from these blood tests, you actually can see where you are today and what are the specific things that you'd want to do differently. They've layered in since Series A multiple acquisitions, including Ezra.

21:00Vijay Pande: But Ezra is basically the function analog for imaging. So MRI and CT. Imaging is like you might get a blood test and say, well, OK, now we want to know more. So the next step would be imaging. They've done acquisitions into supplements and into phlebotomy. And in a sense, they're building up a consumer-focused preventive health care system. And I think one of the biggest shifts in health care that was just starting is the shift from what is in the industry called sick care, where companies make money off of you when you're sick to prevention where you make money when they keep you healthy. From the point of view of a patient, you don't want to be sick.

21:35Vijay Pande: And there's real power in living a capitalist world. But in a capitalist world where people have made money from you being sick, there'll be a lot of incentives for you to be sick. If instead things were aligned such that companies made money by keeping you healthy, that's something where now finally these financial incentives are aligned with what you and I want. And so a function of a good example about devoted health is another really good example of where these senses have been aligned. Devoted is like Medicare Advantage company, so it's for seniors' function as direct-to-consumer. But in either case, there's this alignment between who's paying for it and wanting to keep you healthy.

22:07Connie: How focused are you on D2C versus otherwise enterprise?

22:11Vijay Pande: So I think the thing about the healthcare system right now, typically the customer is the insurance company, not the patient. In direct-to-consumer, that's something where the customer and the patient are the same. And I think there's been a lot of tailwinds post-COVID, where COVID obviously was a big ordeal for all of us. And people started thinking about, well, I want to understand my health myself. I want to be not just a product of the system, but I want to be someone actually involved with real agency in it. And I think people are realizing that that's something that now actually is doable with tools like Function.

22:44Vijay Pande: And then on top of that, AI is a natural companion for this. Obviously, one has to be careful with employing AI yourself. And I'm not suggesting people talk to themselves or anything like that. But I think just the ability, if nothing else, to get data on your own and to know what questions you want to ask your doctor and to be almost like a peer in the room rather than just the subject is a huge advance.

23:06Connie: Absolutely. You don't want to over-rely on these chatbots, but Claude has, in a couple of cases, helped me go to the doctor and say, I think this is this thing and it's turned out to be true. I've talked to other people who've used it as another opinion in concert with the actual medical professionals and it's been really helpful. It's incredible. Also, Vijay, help me understand how you are approaching the design of this new firm. So you grew the biotech arm at Andresen Horowitz. You're managing I don't know, like$2 billion-ish?

23:37Vijay Pande: No, near the end, almost$4.

23:39Connie: Oh my gosh. Okay, wow. So you're now targeting what, like$400 million?

23:44Vijay Pande: Yeah, so right now we're doing something really quite different. And actually, so Mark Andresen, when he was talking about the two types of firms that he thought would be successful, he He talked about, and I think he's maybe saying this a little tongue in cheek, but he talked about Gucci and Walmart. And like, you know, either you run a very small firm or you have a really big one. So A6 and Z is obviously a big firm. VZ right now is named after me, Vijay, and my co-founder, Zach Werner. So he's the Z. And we're intentionally really quite small. And one of the design aspects was that we would actually be very AI native.

24:16Vijay Pande: And that's actually borne out really well in that to date, we have on the investment side, really just the two of us. We were actually intending on hiring associates, but actually it turned out to be with the agents and the things that we built up, not to be something that we need to do, at least at the moment. And it's allowed us also to go really fast in a way that I think, unless you're doing what we're doing, I don't think you could do. I think for a lot of people, the junior people are using AI, but the senior people are working with the junior people. And so for us to go all in with something that really is hard to do unless you start from scratch, it's very natural to do that.

24:50Vijay Pande: And then in terms of fundraising, we're still sort of in the front race. I'm not going to talk about that, but we have been making investments.

24:56Connie: But you're intentionally going a little bit smaller because I'm guessing with your brand and the fact that everybody's very interested in this area, you could probably raise more than what I've read that you're targeting. How do you think about writing checks right now compared to maybe, you know, five and ten years ago?

25:12Vijay Pande: Yeah, so there's two areas that I've been spending most of my time on. One is AI for healthcare delivery, which I did a ton at A650 as well. And then AI for clinical trials. Those are areas that I think are either hot now or are going to become very hot. AI for care delivery, a lot of those companies can grow really fast. Function's growing really fast. So I think tech-like growth from these is not unusual. And I think Function's a great exemplar of that. The trick, obviously, is that you can't just combine AI, willy-nilly, as in anything. You have to be able to really understand the healthcare business.

25:44Vijay Pande: And so we look for founders that obviously go very deep in those areas.

25:47Connie: So AI for healthcare delivery is one piece of it. And then the other is?

25:51Vijay Pande: AI for clinical trials. So this is the area where, let's say right now, like 20 % succeed, 80 % fail. If, let's say, only 70 % failed, which is a relatively modest gain. I'm not asking AI to do something heroic. that means that 30 % succeed. That means you get 50 % more drugs. The amortized cost of drugs go down by 50%. It's a huge, huge win. And because this is so tough, you don't need to be a panacea. You need to just move the needle a bit to be able to actually have a huge impact on the cost, to be able to have huge returns, and actually to be able to change patients' lives.

26:27Connie: That's great. Can you give me another, maybe a recent example of a company that fell right into the center of that Venn diagram?

26:33Vijay Pande: Yeah, so we have not made an investment in this space yet. And we've been looking and there's many out there doing interesting things. One A6 and Z investment is Formation Bio, which Ben Lue is a CEO co-founder. But I've seen like tens of companies in this space. They're all taking different approaches. But I think one of the things that was striking is that many of them have claims that they can predict whether a drug would work or not with pretty high accuracy. That's something that would be a fantasy years ago. And it's actually, I've seen that now in multiple cases, that reality is going to change the game quite a bit.

27:06Connie: And then in terms of the checks, again, has the way that you value a company changed at all?

27:10Vijay Pande: There's two principles we have. One is that AI really makes the power law more important than ever. What I would predict is that the big companies would get really big, that there will be trillion dollar healthcare companies that are built with these types of tools. that's number one. Then number two is that with AI, though, it doesn't take too much to get to, let's say, 10 million run rate. It used to be 10 million run rate was kind of like escape velocity, like, okay, they're going to become a serious company. But with AI and agents, you can build that pretty fast, but that doesn't mean that it's something that could scale into something really big.

27:44Vijay Pande: Because of that, I think Zach and I have been taking a strategy of either coming in a little later, which would be sort of what some people might call growth, but like or maybe or late venture and looking for things where there's, let's say, 30 million run rate. You know, I did the series A function, but like, you know, they were already around for a while. So like, you know, these letters almost sometimes don't mean as much as you might think. So really, I'm looking for a business that's really a scale enough that I could really sort of see the path from here to something really big, either that or incubations.

28:14Vijay Pande: And so there's an incubation we're doing right now, which is still in stealth. But typically, that's with a founder that I've known for a long time. And so this one's someone I've known for almost 20 years. Those two areas where either we feel like it's proven and we can see sort of line of sight to something big or that we're coming in early and we're really helping shape it.

28:32Connie: That's really interesting. So is incubations a little bit more anomalous though based on your network? Or would you say that's going to be a meaningful part of your strategy?

28:40Vijay Pande: Yeah, it's the other thing that we're doing is we're doing a pretty concentrated portfolio. So this is not something we're driving 30 bets per year or something like this. This is, you know, making a few investments per year like this, but that driving still a fair amount of money through those investments, especially at the later stage or even for a lot of these AI companies that seed rounds can get quite large as well. You know, I don't think I have formally did an incubation, but I spent a lot of time with Evan talking about Genesis when he was in my lap and he really ran with that. But I've been around the hoop with founders at the beginning of companies in multiple cases.

29:11Vijay Pande: And I think all the things that I've done in the last 10 years, I think, has given me a lot of sense of what works and what doesn't work. And so those are fun conversations to have.

29:19Connie: That's great. And who are your biggest competitors for deals?

29:22Vijay Pande: Well, so the funny thing about this model is that typically we're not trying to compete for a hot round that people make room for us and they make rounds for us. So it's a very different thing than like, oh, I'm trying to get the hot series A or series B. And largely people want us as investors because of what Zach and I can do and how hands on we can be. I think there's not a lot of people that I would say are following that strategy. But when I look at people who are inspirations, I look at like Antonio Brasovas at Valor. He's well known now, I think, because of the SpaceX deal, but he's been doing what he's been doing for 20 years.

29:57Vijay Pande: His mindset for how he helps companies is an inspiration. And also like what Thrive has done with a more concentrated portfolio and so on. That's also, I think, a real inspiration. Obviously, A6 and Z is sort of in my DNA as well. But I think those other ones are, I think, a new addition to how we think about things.

30:13Connie: And when you say concentrated, how concentrated? I mean, do you think you're going to invest in like 10? What did you tell your LPs?

30:19Vijay Pande: Yeah, so we're talking about like probably five. Oh, wow. Yeah, not a lot of investments, so very concentrated.

30:25Connie: Then would the idea be to, are SPVs something in your world or would you?

30:30Vijay Pande: Yeah, so SPVs are definitely a very useful tool. And I think the benefit of SPVs that actually, when you have a fund, typically you can't do this, is that you can't match make your LPs to the particular deal. So, you know, the incubation is going to be a different type of deal for LPs than the mid to late stage deals. In those cases, we can find very strong matches. And that founder LP fit is really a beautiful thing and can really, I think, only be done by SPVs.

30:57Connie: I want to let you go, but I wanted to ask something that's overly broad. Forgive me, but it is hard for me to understand from the outside looking in. What is overhyped in AI and biotech right now?

31:07Vijay Pande: So this is a longer topic, and my colleague Daphne Kohler has written really beautifully on this. The reality is that AI can find insights that we can't get from just humans alone, and I think we've already seen that. The thing that always gets tricky is when there is this call that AI is going to cure everything. And the reason for hesitance there is not about sort of any doubt about AI. It's about doubt of the data, that LLMs work because there's so much data to learn from. And when the data is just simply not there, then AI can't magically solve that problem. There's some counterarguments that are really intriguing for people looking at evolutionary data from other organisms and sequences and new ways of viewing disease.

31:48Vijay Pande: But I think a lot of people are talking about AI is going to cure all disease are not coming from that place. And I think it's coming from a confidence in the algorithms, which I think is founded, but I think of unfamiliarity with a lack of data.

32:00Connie: Right. Are there any deal breakers for you and a founder?

32:03Vijay Pande: Sure. Yeah. I mean, one of the things that I think is most important to me is that we can really trust each other. And so founders that have a high integrity that, you know, do what they say they're going to do. For me, this relationship is, you know, I'm expecting to be like five, 10 years plus and into ideally their next company and so on. I want to work with people that are thinking long-term like that. And ideally, these are people who are not just trying to win and beat other people, but really thinking about the question, how do we win together? And that's something that I'd love to see more of that spirit in Silicon Valley, probably.

32:37Connie: Absolutely. But also, I guess when you're investing at a slightly later stage, even though the nomenclature is all over the place, that can give you a little bit more confidence. I always think it's got to be really tricky.

32:47Vijay Pande: Yeah.

32:47Connie: When you're backing somebody who you're just putting a lot of trust, in.

32:52Vijay Pande: Yes, usually the people at later stage we've known for a while. And then at the incubation, I think it would be unusual for me to do an incubator on someone that I also have not spent a lot of time with in the past. So like, I think that's been something and everything I'm saying is very natural and everyone might think this is kind of obvious. But what is really unique about our approach is by having a concentrated portfolio, you can do things like this that you just couldn't do at scale.

33:17Connie: And then Vijay, you mentioned Function Health, you mentioned Genesis. So So you've got like three more bets you're going to make or?

33:24Vijay Pande: Yeah. So I was talking about especially investments for my 16 Z days that I'm still in touch with and so on. But like, you know, we're just getting started. So, you know, we have made investment in devoted health and looking to do two more investments right now. So it's still pretty early days. The key thing that limits us is that just the amount of time that Zach and I want to put into companies. The jokey version of this that I tell my LPs is that, you know, typically adding a company is like you're going to have 25 investments in a fund. And so that's something you do pretty quickly. It's like adding a Facebook friend or something like that.

33:54Vijay Pande: But like for Zach and I, it's like going to like my wife and I, why don't you have another child or something like that? You know, like a really high bar, like this long-term commitment of we're putting a lot of time into this. This is a big deal for us.

34:05Connie: That's really interesting. That definitely seems like counter-programming in this market. So yes, it'll be really interesting to watch. Well, listen, thank you so much for making time for me. And I really am just trying to sort of understand this industry, if it can be called that. It can't be much better.

34:18Vijay Pande: It was great chatting with you. And I think it's important. I think this stuff is like siloed too much. And so I think you asked great questions. I think that's the type of thing I wish there was more of.

34:28Connie: Oh, thank you so much. I really appreciate it. And I hope that we talk again in the not too distant future.

34:32Vijay Pande: That would be fantastic.

34:36Vijay Pande: The Strictly VC Download podcast is hosted by TechCrunch Editor-in-Chief Connie Loizos and me, Alex Gove of Strictly VC. Strictly VC Download is produced by Maggie Nye with editing assistance by Teresa Lancansolo and Kel. Thanks for listening. We'll see you back here next week.