The Exchanges

Q What exactly is going on there, and how far behind are we?

A So China has had a lot of success in building one type of reactor over and over and, you know, over again, which is really what we need to do. In order to do something well, you have to do it often, right? And if you have the same design that you're doing over and over, you learn a lot, you get better at it over time, you, you know, shore up the supply chain, you have a labor base that understands that, that system. And they've, they've been really successful at that. So they are ahead of us in terms of scale. Now they're also ahead of us, unfortunately, in terms of Uh, new test reactors. They're, they're also turning on innovative types of reactors that were invented in the United States in the seventies and eighties. And they, they sort of sat on the shelf and we never really tackled them. Um, we ran into a bunch of regulatory roadblocks and then we never turned them on. And China is now sort of taking those ideas and actually turning them on on the mainland. So they're being both very experimental and actually figuring out how to scale out. And those are all the thing, you know, those are the ingredients you need to have a mature industry. So it is a, it's a scary place. You know, I think that China understands that Energy dominance over the next hundred years is going to be even more important than energy dominance was in the last hundred years. And it was already pretty im…

AI assessment note: “So they are ahead of us in terms of scale. Now they're also ahead”

Q Right. This might be a wild question because this is so physical and it's an energy and industrials, right? So we're seeing AGI come into the scene. Maybe you call it super intelligence now, whatever you want to call it, but how does AGI and AI advancements, how do those affect getting nuclear online sooner and all of these reactors up and running?

A So this is probably a controversial statement, and, you know, investors probably really, really want to believe that AI is going to, like, help us build a ton of reactors. I actually think that if you're looking for an investment that is protected from AGI, it's something like this, where it's extremely physical. You're bending steel, you're stacking blocks, you're building plants, you're laying concrete. It's going to be a long time before AI touches that level of the physical world. There are a lot of things that are going to be automated first before that. The other thing I'll say is that our customers are going to be data centers, right? They're going to be people who want to actually power the AI. So interestingly, nuclear is sort of like the derivative to AI. If you believe that AI is going to go vertical, you should probably invest in nuclear, you know, because A, you're going to power it. B, you're going to be very insulated from, you know, the deflationary effects of that. Um, and yeah, of course we use LLMs and these sorts of things to help us do work management and help us control information and these sorts of things. But at the end of the day, this is a matter of, we need to bend a lot of steel, we need to weld it together, we need to stack the blocks.

AI assessment note: “It's going to be a long time before AI touches that level of the physical”

Q A lot of people want AI right now. How are you thinking about involving AI with Carta?

A We're all in. I actually just came from two internal AI meetings. AI is interesting. Um, what does it mean to be an AI company? Uh, and some people think of it as, oh, you're building AI products. So you're building a product that does AI or uses AI or some, some AI enablement product. Uh, another is you could be an AI company building traditional non AI products, but your internal stacks all AI. You are, you are building a new type of company based on, on AI. Uh, we're both. So today I spend half my time on product and the other half on AI projects at Carta and all the AI projects come in two forms. One is how do we use AI to build differentiated user experiences for our customers? So it's all in the product side. And then how do we use AI internally to be a better, faster, smarter organization? Uh, and I, I split time on both. Um, and so much of the AI stuff I'm learning, uh, It's, it's like programming. Like, it's like being, being an engineer, but instead of programming using Python, you're programming using the English language, but most people aren't trained programmers. And so much of, uh, changing the culture of Carta to be an AI first company, uh, is, uh, teaching people to think differently, just to think about, like, I was just in a meeting where a team was manipulated. We, Do a lot of spreadsheet work. We have a lot of spreadsheet lists at Carta, a bunch of Excel pe…

AI assessment note: “today I spend half my time on product and the other half on AI projects”

Q How do you use it in your workflow?

A I use it for a lot of kind of the basic stuff of like reading stuff, transcribing stuff for me, especially I can, I can dictate really quickly and turn it into, to written stuff. So there's like all these like little kind of productivity, uh, uh, hacks that are helpful. Where I find it most useful though, um, uh, is actually for other teams, uh, because I don't do a lot of things that are repetitive. I don't do a lot of things at scale. You know, I, most of my day is eight hours of listening and then I make one or two decisions at the end of the day. Whereas other people that are actually producing things, you know, they're producing reports, they're writing code, they're, Creating spreadsheets like the people that are actually doing production at scale. That's where I think it's been super helpful. Uh, and so I spend most of my time actually helping them learn to use AI to do their jobs better. Uh, and that's, that's been incredibly high leverage for us.

AI assessment note: “I use it for a lot of kind of the basic stuff of like reading stuff, transcribing”

Q I'm curious how you think about this in the grand scheme of geopolitical tensions. You did mention that energy is directly linked to conflict and GDP. So what are your thoughts on that?

A Yeah, I think, I think that's really two, I would think of it as two different topics. I think You know, this is, and, and in terms of other groups that, um, we've, we've worked with or spoken with in the administration, the, uh, energy dominance council has been awesome for, for trying to push for more us energy production. And that's because they know energy production and consumption is directly linked to GDP. It's one of the most correlated things with GDP. So on the, on the GDP side, yes, energy production, specifically baseload energy production is, is completely linked. And drives things like manufacturing and AI. Um, so, on the GDP side, you know, that's how we think about that. On the conflict side, I think the point I've made there before is, A lot of people worry about conflict with other peers or near peers, and people have different thoughts on how that's going to manifest, whether it's an AI war, if it's kinetic war, or if it's just economic tensions. And, you know, I think what we've seen for the last decade or more is just economic competition, and that goes back to energy production as the foundation for all economic activity. Um, and so the way, the way I think about that is whether you're worried about None of the above. If you're just worried about growth, great energy production. If you're worried about economic battles with other countries, you need energy…

AI assessment note: “on the GDP side, yes... On the conflict side, I think the point I've made”

Q trends and investment stuff on the private market side with VC, private equity, IPOs, that kind of thing. Um, But I'm very curious from your perspective, um, including RIAs, let's say, other family offices, I work for a family office, um, wealth managers, and, um, other kinds of sophisticated investors. This topic is really timely right now because the market is so volatile, but how are you seeing portfolios shift?

A So the main way I'm seeing portfolios shift is one, Which a lot of people can point out as a rise in passive, right? So most people, you know, 20 years ago would say they're making all their money in active bets. They're going to pick stocks. They're going to find active mutual funds. They're finding a manager who think they're going to outperform. Um, and this was usually expressed in the form of a mutual fund. And now everyone's realizing the real alpha is just buying the market and holding onto it for a long time. But the, the interesting thing is that buying the market is no longer buying the market. So if you buy the S and P 500, You're not actually buying the whole stock market. You're buying, you know, there's private companies now that are carrying significant market caps that should have gone public, but they're not, they're still private. Um, there's, you know, other types of fixed income instruments. There's all kinds of things. There's private equity, um, that you are not getting in your portfolio. So the key is how do you buy a index like portfolio across the whole market and all the available asset classes? So I think the biggest trend is the merger or the blend of Private markets and public markets coming together in one place in one orchestrated system, which is what we're aspiring to build. We call it the total portfolio solution. Um, and then also trying to un…

AI assessment note: “So the main way I'm seeing portfolios shift is one, Which a lot of people can point out as a rise in passive”

Q look onto that. I've, I've heard rumors about, you know, the AI legal space, and there might be huge high-flying companies that rhyme with shmarvy, but, you know, not many actual lawyers are, are using it in-house. So, like, implementation inside is, is low, but adoption is really spread across. How deep is AI, like, actually implemented within the enterprise, and How long do you think that's going to take?

A Yeah. So we're transitioning from this phase of everyone for a couple of years have known that they must say to the world that they're using AI for everything. Every board will have said, like, we have an AI mandate. We're going to adopt or die. And so that means everyone talking a big game about how they're using AI. In practice, there have been a lot of proof of concepts, sandbox, enterprise deployments, internal deployments, kind of in low stakes scenarios. And that's enough that you can say you're using AI. But there's still a gap for this to, like, when you look at what is the number of lines of code that are being contributed to JP Morgan's code base that's generated by AI, as of very recently, that's extremely low. It's now starting to happen. We're starting to make, like, our way into, like, real use cases, both hospitals adopting this, banks adopting this, and businesses adopting this. But it feels like it's really in the last six months that you're moving from toy projects to kind of real adoption. But still that's just the frontier companies that are, uh, Like when you look at the long tail of businesses, they're still like, uh, still lots of the world runs on paper, right? Like there's still a long way to go.

AI assessment note: “moving from toy projects to kind of real adoption. But still... a long way to go.”

Q I'm not sure what the range is. It might have been, I'm not sure. We can check. But, uh, could you break down what is a kill switch in AI?

A Yeah. So a kill switch is like, does anyone anywhere have a button that they can press if something really bad happens that shuts off AI? And this is really interesting when one of the concerns that's coming up now that AI is getting really good and getting really agentic is that there might be what's called a loss of control scenario where Basically the AI agent becomes so agentic that it starts to take action that we no longer want. It might copy itself onto servers that we can no longer control. And all of a sudden humanity stands with a big disaster on our hands. And so a close kill switch is like, can we in all deployment scenarios, retain a button that big red button we can press to make this go away. And maybe an analogy is like in any given building, you're going to run a, Uh, a fire, what's it called? A fire drill, where everyone exits the building. You cannot run a company where people can't exit the building pretty fast if something really bad happens, and that's the same thing here. Um, whether that's going to happen by any of you, I don't think so, but it's, um, it's growing an awareness, even in, in this administration, you see, uh, David Sachs articulating like these kind of catastrophic scenarios. Uh, not outside of the realm of possibility, and so you need to plan for it. You actually also hear Chinese government taking, like, using the term kill switch pretty …

AI assessment note: “a kill switch is like, does anyone anywhere have a button that they can press”

Q VCs DM me on X, and I can get all your portfolio companies set up with a free samples credit deal. Your investors are super ingrained into the space, especially not Freeman. I mean, he's now a part of the super intelligence team at meta. What kind of advisement are you getting from them? How close are you, um, And like, what is the relationship between you and your investors?

A Yeah, um, we're really close. They're just, like, Nat is just the, I would say, the preeminent AI investor. Both made a lot of great investments, but also, as you just mentioned, is very close to the metal as to how this is actually being built. Uh, both the small questions of how to build AI, but also some of the big questions. Um, Nat himself spent a lot of time on Some of the risk questions that we were dealing with when he was leading GitHub, they built GoPilot, which is like the first real AI product that worked, and he led the efforts that led to Microsoft basically ensuring all of their customers against IP risk to say, oh, they're really worried about IP. We're just going to step in, and if they get sued, we're going to cover their ass.

AI assessment note: “we're really close. They're just, like, Nat is just the, I would say”

Q And so could you explain more about how Palantir is using AI and your new product AIP within the system?

A Yeah, absolutely. So one of the things that we always pride ourselves on Palantir is we really sell Outcomes. And when you heard Sean today, that was his real message with AI. It's not just about the models, the infrastructure, all the cool stuff we talk about. Uh, our software dominates, and I appreciate you using that tagline, uh, because we focus so much on outcomes. So I work on the, our US Navy business quite a bit. And so I'll lay out the four major areas that, that we focus on. One is how do we use software and AI to have the most lethal fleet, uh, in the world? And if it was interesting about the US Navy, and one thing when I talked to admirals is they'll take the Russia Ukraine example, which is fascinating. Russia actually has a pretty formidable Navy, but they've been completely ineffective in the Black Sea because the Ukrainians are using new technology and new methods to, in essence, neutralize the Russian, Russian Navy. And so we're working with Maven Smart System to deploy across the Navy and afloat with ships to be able to have incredible battle space awareness and time sensitive targeting. So not only can I identify The target, but how do I make sure that I'm using the right effector weapon system to destroy that target? Whether that's in a cloud environment or on the edge on a ship. The other area that we're, we're looking at for software and AI is how do I ma…

AI assessment note: “we're working with Maven Smart System to deploy across the Navy and afloat”

Q request, um, based on a simple belief, I will quote, America needs better physical technology and startups are the best way to build it and scale it. You had a pretty, uh, I'd say collaborative response to that rather than seeing it as competition. You're like, no, we need more of these. So can you explain more about your position on that and how you would help engage this environment?

A Yeah, so I think that you've got to remember that there are kind of four big categories, right? There's these huge industrial companies, you know, like a Raytheon, like a GE, who need to upgrade themselves very rapidly. There are new industrial companies like Hadrian who are starting from a clean sheet. Um, there are 30,000 small manufacturing businesses in the US that are super important, even in things like making steel that have been completely unsupported by technology, including AI for decades. Um, and then there's obviously the new prime. So my view is like, you know, one, there's plenty of software products out there to go build for all these companies. Secondly, there are many micro categories of manufacturing, you know, just like open AI and AI won't do everything. They will do the vast majority, but you've, for this level of mission, you kind of need everything. I think it's great. I mean, the number one thing that we did that was important was have software engineers get excited about automating manufacturing again, versus getting stuck at Google, you know, optimizing ad clicks or trading cat pictures on the internet. So I think YC stepping into this just for the fact of getting like Stanford and Harvard kids and, you know, kids from the Midwest back involved in building software for the real world is really culturally important for the country. The second thing is, …

AI assessment note: “I think for the country, we need like orders of magnitude more than this.”

Q been down bad and that's to say the least, but I, I want to get into that a little bit more like macro wise. Like if we zoom out, of course, there's a lot of built up energy for exits and liquidity. Um, but what other kind of, like, structural dynamics are happening that's allowing this kind of new phase of M&A to take place? Competition? Regulatory? What's going on?

A So I think, and a number of your recent guests on the pod have, have touched on this actually, I think that in a world of AI and as AI starts to mature, I think a lot of the benefits actually go to what I would call not the incumbents, but like the fast moving larger companies. Like, of course there will be a few companies that start from zero and explode, but you know, companies like Salesforce, like Databricks, like Brex, like, you know, everyone, we're all using AI and we're all, uh, incorporating it into the internal tools we use into the products that we sell. And so in that world, I think that there's a desire and ability to move fast, but then look over your shoulder and spend money on someone who's building quickly because you're the one with the distribution. So this is going back to what people on your pod have cited. I think distribution is more important than ever before. And the fast moving companies that, uh, can incorporate some AI, but already have distribution, I think are in the, in the best position, but they still have to look over their shoulder to make sure, Hey, like, what are these next gen companies doing? Maybe I should take them out because the product is really interesting and I can push it through my channels. You know, it's, it's honestly like history doesn't repeat, but it rhymes. It's a little bit like when you look at generational companies, lik…

AI assessment note: “spend money on someone who's building quickly because you're the one with the distribution”

Q As we wrap up, I have just two more questions, but I'm really curious. I've been reading a lot on the finance fintech space. What is the new finance operating model?

A Yeah. So I think the new finance operating model, basically what we're seeing in our customer base is a bit of a fundamental shift in that there's a lot of uncertainty and volatility in the world. And we recently published this CFO survey of 500 large, uh, company CFOs and their views on the macro and like the way they think about their teams and their tools. And essentially finance teams are really Re-architecting their whole stack and really only wanting to work with the tools that Help them better understand their business and also work well with each other. And I think the, the second point is really important because the idea is Brex wants to be one of the systems of record for a business, but it also has to work very closely with the other systems that they use. Another huge theme, as I'm sure you're aware is the rise of AI and AI agents in Finance. Finance is a space where there are a lot of numbers and there actually are a lot of ways where agentic workflows can really help, uh, and delight the back office, uh, financial, uh, manager or CFO. And basically as a part of that, we saw that our agents, specifically our expense and accounting agents saved folks, 169,000 hours over the last six months.

AI assessment note: “finance teams are really Re-architecting their whole stack and really only wanting to work”

Q What's, could you explain more on the middle, like what is this middle ground that he's talking about?

A Well, I think basically what he's saying is there's, um, there's specialists, which either means that your special, you know, your specialty is a sector or a stage or a way that you help companies, and so you're like a boutique, you know, like this is like the Gucci and the way he described it, or you're hyperscale, you are resources and your power and your, you know, access to, You know, the world, your access to the government, et cetera, because you have all this institutional strength, which I buy that too. And I think what he's talking about is that, um, in the middle, you're trading off both the ability to be a specialist and you're not big enough to be, you know, institutionally powerful like that, which is, which is a believable, a believable description of the world.

AI assessment note: “in the middle, you're trading off both the ability to be a specialist”

Q In this competitive environment, we've been seeing big numbers toss around and strategies kind of, uh, shape and in pretty absurd ways. In particular, in this conversation, you guys were talking about the Meta deal and, um, how Meta's been trying to poach some of OpenAI's talent, and then also this Metascale AI scenario. So what were your takeaways from that?

A Um, the temperature on the whole market has definitely risen, both, you know, the numbers from, you know, VCs, what people are paying, the length people are going to, to compete now. Um, you know, in scales market in particular, um, I'll give a shout out to a, you know, portfolio company of ours, David AI, which is like, you know, um, voice data, and they've, that, that market in general has seen unbelievable, Just pull like the, the, the demand from the model companies to get good, clean data that they can use is like a huge competitive advantage and it's like super critical. And so, you know, seeing that market pull for companies like David and others does, it helps me understand why Meta was willing to, you know, go so hard at this problem. Obviously, there's also just like a big talent work component to everything. But in general, I think everything we're seeing now is sort of, um, gloves off. This is the most important thing. And so their regular rules don't apply type of energy.

AI assessment note: “temperature on the whole market has definitely risen, both, you know, the numbers”

Q To go further into where nominal software fits in, could you just explain out the end-to-end hardware development process?

A Yeah. So starting on the hardware development process side, Generally, our, our customers are going to fall into one of two buckets, and sometimes if they have multiple products, they're in both. So on the one hand, you have the zero to one development program, and you're starting from pure simulation data to prototyping to initial qualification testing. Um, qualification testing can take months, often takes years. Um, it's a very arduous process. And then into production testing, which is kind of, you know, things are rolling off the line and we're just making sure that everything works. Our software product overall is focused on this entire spectrum of test cases, right? So we have customers that use nominal only for simulation data and customers using it all the way on production manufacturing lines and out to their fleet. And all of this data is relevant to pull into a single consolidated database, be able to manage it, present it back to engineers, to technicians, to VPs in an easily consumable way. And then the second piece is In the hardware development process, right, you need to scale one to n, and that means bringing up a production line, but to Cameron's point about everything is a test asset, you also need to be aggregating all the data from your fleet back into the same telemetry management and analysis tooling that you're using for development to be able to close …

AI assessment note: “you're starting from pure simulation data to prototyping to initial qualification testing”

Q Well, there's a lot to break down there and so much to get into, especially the couple optimizing for IQ. Um, first I want to get into the genetic optimization software for embryos. What does this actually mean?

A What this means is, you know, parents for the first time ever can give their children the best start in life. So often, you know, the first thing parents think about when they're having healthy children is how do I not pass down like a rare disease, right? Something like maybe someone's a carrier for cystic fibrosis or something like Huntington's, for example, and they might think, okay, how do I not pass this down to the child? Then they think about maybe other diseases like common diseases, like breast cancer, like heart disease, These are conditions that, generally speaking, are just simply not tested in the clinic today. So let's say you're a couple doing IVF. The first thing that's very well accepted in the IVF context is making sure your child doesn't have any chromosomal abnormalities, like Down syndrome, for example. Ok, so that's the first thing. Then they might test for rare hereditary diseases, as mentioned, like cystic fibrosis. But then that's it. They don't test for common diseases like cancers or heart disease or schizophrenia. They don't run any more sophisticated modeling of genetic prediction on things like IQ or height, for example. And so what really we're doing is we're completely empowering the patient with the full genetic profile of their embryos so they can, they can genetically optimize their children.

AI assessment note: “empowering the patient with the full genetic profile of their embryos”

Q to take a step back and actually understand how important this moment is in the grand scheme of fertility. There are like significant shifts in the US for fertility. Um, it seems commonplace now that couples are freezing eggs, that they're using IVF. Um, and even broader than that, there's declining birth rates worldwide. Could you put some data behind all of this? What is the current state of fertility?

A Well, you make a great point. First off, birth rates are declining. Um, IVF is becoming much more commonplace. I mean, I think the statistic is that one in 50 babies today are born via IVF. That would be crazy to say a couple decades ago. I mean, at some point, IVF babies were, you know, test tube babies, right? Now it's just how people have children. And you can imagine as it becomes easier and easier, maybe you take stem cells and make an egg or sperm, as it becomes easier and easier and less invasive to make an embryo, You can imagine that IVF is going to become more and more ubiquitous. And it's interesting because, you know, I always talk about the decrease in cost of whole genome sequencing. What used to be a billion dollars is now a couple hundred dollars. That obviously DNA being the generational health tool that it is, doesn't just impact adults and longevity, but also impacts fertility. And so you're going to see the rapid decrease in cost of IVF combined with the rapid decrease in cost of sequencing A genome. These things are going to converge together to enable this just, this massive, massive new generation of parents having genetically optimized children.

AI assessment note: “I think the statistic is that one in 50 babies today are born via IVF.”

Q What are the specific categories of traits and conditions that you're analyzing? How accurate are these predictions?

A Yeah, so the different categories ranges from cancers, um, to, like, you know, heart conditions, to, like, mental health conditions, to, like, um, something like physical characteristics, like height, or eye color, or, like, hair color. And regarding the predictability of the models, there's a large range, depending on how genetic something is. For example, people might intuitively guess that something like height, right? Height, you would probably guess, you'd probably guess that's pretty genetic, right? You look at mom and dad, you're like, okay, this thing is like pretty genetic, which it is, right? And so that model tends to be a very good, strong model. Let's say something like depression. Depression's dramatically more environmentally driven, right, than height. And with that, you would expect that model to be dramatically worse. And so DNA isn't deterministic, and that's actually not a bug, that's a feature, right? Obviously DNA, if it could predict everything, that there would be no free will, right? But it can't. And so one of the things that we're really careful in doing is we actually tell patients when they're using our services which models are better, which models are worse, and the way we present results to patients always takes this into account in that if a model has lower genetic predictability, you would basically see the deviation or the prediction of the mo…

AI assessment note: “the different categories ranges from cancers, um, to, like, you know, heart conditions”

Q So why exactly is travel broken right now?

A I think that you have incredible fragmentation. It's sort of that classic Keith Raboy quote of take a highly fragmented, low NPS industry and vertically integrate it, and that's how you create enterprise value. You have legacy softwares, different types of operators, different qualities of inventory, conflicting incentives, and candidly speaking, no one's looking after the underlying customer. Um, all of these companies are, and, and individuals are just trying to convince you to book a trip with them, or spend your money with them, and then once that's done, they kind of just like, are like, okay, great. Um, versus at Wander, literally, like we have a button in, in the concierge chat, like a thumbs up or a thumbs down, and when you hit that thumbs down button, it doesn't go into the ether, like it actually pings me on Slack.

AI assessment note: “incredible fragmentation. It's sort of that classic Keith Raboy quote”

Q So I want to get into the app, because we have many engineers in the audience, many founders, and tech curious, so the app itself, can you just break down the magic behind it all? Because I know this is actually very robust, and you've built something that's...

A Yeah, so I mean, like, the, the tech stack that we use is actually very simple. We, we made the decision early on to, um, have the company be basically entirely TypeScript, um, Which, of course, you have, like, that classic debate of, like, do you want to do native iOS development or not? Like, what do you want that to look like? And ultimately, we made the decision that we just didn't have enough money to have two different development teams, and we couldn't support, you know, different versions. So we, we wanted to have And use frameworks that were, um, you know, compatible against, across all platforms. And we also made the decision that, you know, building in, in TypeScript to have a little bit of continuity between front end and back end teams, which to be clear for, for all of you who are, you know, gonna like roast me in the, in the comments or, or whatever else, like, um, like, yeah, like there is definitely a, a more performant way to, to build, but you know, in the early days, you know, it's like you have like one or two engineers. And so, You have to make those kinds of decisions. You know, I think that probably the most interesting thing about Wander from a software perspective is that you have the booking engine and the booking platform, and that's the only thing that we surface to guess, but underneath you have what we call Wander OS, which runs 100% of the proper…

AI assessment note: “the tech stack that we use is actually very simple. We made the decision early on”

Q I have heard that you do run by some really strong life philosophies. Could you share more on that?

A When, when I was, when I was younger, um, I, I became self-aware at, like, a pretty young age. Like, um, I, I remember being, like, eight and laying on the grass and doing my homework and, like, stopping for a second and saying, like, life is never gonna get better than this moment. Like, you are eight years old, buddy, and soon you're gonna be 30 and 40 and 50 and, like, this is what's up. Like, enjoy, enjoy it. And so when I was young, I made this, this set of rules, um, to help me live, live a good life, to help, to help me decide, uh, in, in what I thought were challenging circumstances and ideas. Um, and so the, the, the four, the four rules are, and the first one is rather morbid for a kid to come up with, um, is that the worst thing that happens is you die. Um, And for me, what that rule or that mentality enables me to do is it enables me to walk into situations with a level of almost comfort that, like, hey, like, you can do this, you got this, like, you know, the worst thing that happens is you die, right? So it's like, uh, which of course, like, accepting death is a really, like, you know, crazy idea. Um, you know, the, the second one was about urgency. Tomorrow is a lazy man's today. Uh, which was if you want to live a fulfilling life, like, you can't just delay things. Like, you should try and get it done now. You should, you should take that shot. You should take t…

AI assessment note: “I made this, this set of rules, um, to help me live, live a good life”

Q I want to talk about the technical sides of this. You are the CTO. I'm sure there's like incredible engineering challenges, but you're also scaling incredibly fast and maybe breaking a lot of things. So like, what are, what's the process for you and how have you started to scale up the product suite?

A Yeah. So that's a great question. So building these products is no joke. Like we have many, many startups within WAP almost. Like our live streaming feature is a feature, yet you have platforms like Twitch and Kik that are built off their Instagram Live. We have chat, which is a feature, but you see platforms like Discord and Telegram and Facebook Messenger and WhatsApp, whatever. Um, we have other things like our notifications, like We send out over forty million notifications a day sometimes. So, these are no joke systems, and we've started to really realize the difficulty And honestly, the awesomeness, like we love solving problems in scaling some of these problems. So payments, I would say is the first major challenge we did. How can we make it so people can come on, they can drop their link to a Twitter following of a 100,000 plus and the website not go down. That was a challenge. The website has gone down several, several times. Um, but we've been able to kind of progress that and now we can handle flash drops of over a million, uh, or two million even at one point. I would say over the last year, we've gone into some crazy technical challenges with how to build chat. We've built chat four times now. Um, we've built our notification system three times now. We've built our real-time feed where we serve you content three times now. We're starting a new one next week. Um, th…

AI assessment note: “We've built chat four times now. Um, we've built our notification system three times”

Q That's fantastic. As we wrap, I'll have two more sections, two more questions. One, what are you most looking forward to this year?

A So I'm not going to give away our trade secrets, but I'll tell you some of the things I am most excited about. We're going way deeper into payments, making it so you can get paid in or paid out anywhere across the world, either in fiat or crypto, offering more solutions for our creators to get access to capital or even spend their capital. We want to go way deeper on our consumer features, our discover. We should drive all the distribution you ever need. You should never need to go outside our platform to get customers. We're going way deeper on that. We're going to simplify our products so much that You can understand it when you come on, um, building things like real-time feeds and efficient notifications and going deeper on live streaming. A lot of these really awesome consumer features. And then I would say the cherry on top that I am personally most excited for right now is we're going heavy on our developer platform. It's easier than ever to create software, especially with all these different AI tools. What's still hard though, is how do you get paid for it? How do you get your customers in? How do you distribute it? How do you, um, Keep retaining value for it. We've solved all these problems. Now we can easily do it for developers. Now you can easily spin up an AI agent, an AI agent and have it integrated to the WAP platform to buy and sell and receive payments. So you …

AI assessment note: “I'll tell you some of the things I am most excited about. We're going”

Q Let's talk about that. So dig into the strategy there, because you, you mentioned you're concentrated, you do few investments, but you're writing large checks. So could you break that down?

A Yeah. Diversification is the enemy of outlier returns. And so few things matter less, but better. We want to earn the right to partner with a couple of companies, the couple of teams that have the opportunities to build like truly special companies. And so we're investing out of our seventh venture fund right now. I, you know, don't quote me on this, but I think like less than 70 total investments on, on the private side across those seven funds. Right. And so. You know, Brad's really great as a partner to all of us of like, hey guys, like, wait for your pitch. Um, you know, spend the time, you know, one or two investments per partner per year. Not that that's like a hard rule, but we do really, um, You know, we're not a high volume shop. We're not making toehold investments. Like we spend a lot of time really intentionally to identify and build the relationships with the teams that we're most excited about to, you know, make an initial investment and then constant earn the right to concentrate a significant amount of capital in each, you know, round going forward. Um, and we are very comfortable writing a large check to start because of all that work that we've done and because of the conviction and belief that we have In the team, in the market opportunity, in the company of itself, and then, you know, like, We're happy to put, um, a significant percentage of, of a fund into …

AI assessment note: “Diversification is the enemy of outlier returns. And so few things matter less, but better.”

Q You have a little bit of a hot take on this, though. Um, why do you believe picking is ironically more important than winning as a lead investor?

A So, I think coming back to what I said, I think it takes to win in these very competitive processes. The, the, the limiting factor is really who do you choose to do that work with, right? You cannot do it with 20 different companies at once. And so there is this element of taste and picking that goes into that that I think you actually have to take very, very seriously. So like one part of that is like, okay, what are the few best ideas I have right now? Who are those really, really promising founders that I believe deeply in and think have tremendous potential that I'm going to go and spend that time with? It doesn't always work, right? You don't always build a great relationship. Sometimes you bounce off each other. And I think having the self-awareness to also say, hey, like, Maybe we don't vibe the way I'd hope we did, and that's okay. I'm gonna go move on to another idea is, is another part of picking. The other part of it, though, is sometimes you meet another company, and that company turns out to be extraordinary for reasons you didn't expect, and I think having the flexibility and the open-mindedness to then kind of reprioritize that and, and really pick that one to spend time with is, is a really difficult thing to do because of the sunk cost fallacy that goes into all the time you spent with others, but I think you have to kind of

AI assessment note: “The limiting factor is really who do you choose to do that work with”

Q on guard. You always have to be looking out for things, and I have a reason every week to aggregate all the top companies in one place and make sure that I'm keeping my mind sharp. Gotta keep your mind sharp. Exactly. I want to shift over into something a little bit more spicy, so we're wrapping up a bit, um, but why do you think ASI is already here?

A Look, for me, I think ASI is here in narrow pockets, right? Like when I use deep research, like I've seen the future. It's just absolutely incredible. Um, it, it, my eyes pop out every time I get back a report from that product. It's really, really special. Uh, I think what you see happening in, in code generation and some of these other more mature use cases for LLMs, like it, I feel like I'm seeing the future every day. And so intuitively to me, it feels like we have this, now this like higher power that is like super intelligent. And so therefore like ASI for me is here in certain areas and It's not to say that these LLMs don't have problems and that the AI doesn't have like holes in its ability to do things that we do well as human. Um, as Mark said, um, you know, venture's not, not under threat yet for, for good reason, right? Um, but I, but I think what, what I mean to say by, by making the bold statement that ASI appears to be here in some sense is that I think if progress were to grind to a screeching halt, we still have what I believe to be like the most incredible economic opportunity of our lifetime in front of us. And, uh, I'm very happy with what we have today. I'd be very happy if the next 10 years was about taking what we have today and applying it to the world and bringing it to consumers and businesses and every nook and cranny of their lives. But I also think …

AI assessment note: “I think ASI is here in narrow pockets, right? Like when I use deep research”

Q So can we talk more about the brand? Because you guys have done that really well. You created a little, um, like hype storm on the internet and on socials. I'm wearing the healthiest hoodie that's ever existed. It's incredibly soft, and I'm not just saying that. Um, but could you share more about your brand strategy and how you got to that point?

A I think one of my biggest learnings Over the past decade in startup world is just the power of eponymous brands, right? Like meaning it is, you know, the name is really what it sounds like. Uh, and companies like Robinhood did this exceptionally well a decade ago. They came into the market, uh, in a time, you know, post 2008, and obviously they're a forty billion dollar company now, right? So it's something has worked. Uh, but they came into the market in a time when people were feeling super financially disenfranchised. Uh, and the brand literally stands for giving to the rich or stealing from the rich and giving to the, giving, giving to the poor. Uh, it's all about financial empowerment, right? There's a lot of like vibe to, to, to Robin Hood and the story and the ethos of, of, of the company. Uh, and they built a tool that helped people take control of their financial life and disintermediate, you know, the traditional almost financial advisor, um, or, or, or, uh, brokerage system. That brand has obviously hit mass market acclaim, and we kind of see there being very similar seeds right now in healthcare, right? Uh, United Healthcare CEO getting shot. People are very frustrated. You can't go more than like seven minutes on social media without seeing something about, you know, the Make America Healthy Again movement, uh, nanoplastics. It's just, it's reached a boiling point.…

AI assessment note: “power of eponymous brands... our whole ethos is to superpower people's health”

Q well. I will say, uh, I will say that. I, I think you'll attract some very highly qualified people and talented people, especially if they made it this far in the pod. So. The, the meals do help though, I will say. Totally. Well, um, so just to close out, I'm really curious on your take. If you had unlimited resources, what would base power look like in five years?

A I, I'll give you where I think we will be in five years, and not because we have unlimited resources, but I think we, look, we, we have a group of investors behind the company that have a lot of resources, and also see the long-term vision, and see that this is not a company that, you know, is designed to be around five years, and then get bought, and, you know, go do, like, we want to build a company that endures. We want to, we want to be part of the S&P, we want to be around for 50 years, we want to, Uh, build critical infrastructure for the country, and what that looks like is being in all 50 states, uh, going up the stack into, into utility infrastructure, going down the stack into the, into the home, uh, and building more products and services to lower costs and increase reliability for energy consumers, whether that's homeowners or business owners, um, and you'll see us do all that over the next couple years, and, you know, we're, we're honored and, and proud To partner up with firms like Addition and Adresa and Lightspeed and Valor and Thrive and Altimeter and Trust and folks who see the long-term vision have the resources to support us as we go from, um, you know, an 80 person company in one state, uh, to, you know, hundreds of, you know, eventually thousands of people in all 50 states and hopefully, um, you know, touching a, a material percentage of the country's powe…

AI assessment note: “what that looks like is being in all 50 states, uh, going up the stack”

Q So I guess the big question is, why now? Like, what is the scale of the US energy problem?

A Yeah, it's, uh, the, the why now to us is, is, is quite clear. The, if you look over the last few decades, um, if you just look at kind of the electricity sector, load growth is what's called in the industry or basically like the demand for electricity has been kind of ticking up very marginally, a percent or two a year, um, as there's been more population and as, you know, people have computers in their homes that didn't before they were using electricity. But you had this sort of offsetting efficiency gains of, you know, incandescent light bulbs changed to LED light bulbs, and now you're using actually less electricity than before. And only over the last two to three years and, uh, what we believe of over the next few decades that there's a knee in the curve, right? There's a, there's a, there's a much higher slope on electricity load growth. And the reason for that is honestly, it's maybe self-explanatory and relatively simple is that the, End use of energy is moving from fossil fuels from natural gas to heat your home to electricity via heat pump. People are getting heat pumps installed. The end use of energy in the form of gasoline or diesel is now going into electricity for EVs. Um, and then of course, uh, data centers and, and, you know, fueled by AI are also driving this, this load growth phenomenon. And so electricity demand is, is at this sort of turning point Uh, in …

AI assessment note: “the why now to us is, is, is quite clear.”