The Exchanges

Q For those who don't know your origin story for Impulse and why you decided to go after this after SpaceX, could you share more on that?

A Uh, sure. Yeah, I, I was, um, a founding employee at SpaceX. I developed, I led the development of the, the propulsion systems for Falcon and Dragon, and, and also started the origins of what became Starship. Um, my, probably my proudest development was the Merlin engine, which is, which is currently flying on Falcon nine, the most reliable rocket engine ever developed, and also the highest thrust of weight of any rocket engine ever developed. But I, I worked on Starship for the last six years that, uh, that, uh, SpaceX. So when I left, I had the plan of, you know, launch mostly being solved or is being solved. It's like the next big opportunity is to move all that, all that payload, all that cargo around in space. So I started impulse space to do in space transportation as opposed to from earth to space transportation, which was SpaceX. So we feel like The next step is in space transportation, including to, to the moon and, and landers on the moon, which we're also bidding on.

AI assessment note: “So I started impulse space to do in space transportation”

Q I'm really, yeah, I mean, I think that's like a good thread to pull on because A-sixt Z is a marketing media machine. Mark has talked about this on plenty of podcasts. We see it with Eric and all the podcasts that are coming out. I can't keep up with them. Uh, but it's phenomenal. So how does this marketing machine that help with deal flow and helping these companies?

A Yeah. So, I mean, you alluded to how, how it started. Uh, I give a lot of credit to Margit, Grace, Maya, the marketing team of the early days who built the brand, uh, before we deserved to have one. We didn't have the returns. We didn't have the portfolio companies. And then as Mark has alluded to, that gave us, um, you know, the ability to, you know, win deals, um, support our portfolio. It gave us Power in the market. Strong venture capitalists have power that they can lend to their portfolio companies, uh, to enable them in times that they otherwise couldn't. So that's with hiring executives, uh, corporate partners, customers, the like. Fast forward to today. I mean, you talk about Eric and his ninjas on the new media team. We can go to an early stage founder and almost guarantee That their launch is going to go viral. Um, that is a differentiated offering. So, um, if you look across the operating platforms, we strive for that differentiation. It helps us upfront with deal flow because people hear about it. It helps us with winning because they reference it. And then once they're in the portfolio, we think it helps deliver outsized returns.

AI assessment note: “It helps us upfront with deal flow because people hear about it.”

Q Wonderful. Ok, so, we have some really big news today that we need to address. Can you share this latest funding announcement?

A Yeah, I mean, you know, we, Zipline just closed, uh, over six hundred twenty-five million dollars, uh, which is going to accelerate our expansion both in the United States and outside the US. The big thing that happened last year is that the international part of Zipline's business, particularly the life-saving work that we do in Africa, uh, grew incredibly fast. I mean, we, uh, are now saving about 17,000 lives a year. We're gonna go from serving about 5000 hospitals and health facilities To over 20,000 hospitals and health facilities in the next 18 months. We had almost no operations in the U.S. at the beginning of last year, and we have now expanded to the point where Zipline actually does more deliveries in the U.S. than it does in the rest of the world combined. So that grew really fast, and we're expecting that business to grow, uh, by more than 10 X this year again. And so, uh, really the fundraising is designed to position Zipline so that we are ready from a capex, from a manufacturing, Uh, from an operations perspective to add a lot more metros over the coming, uh, you know, four quarters.

AI assessment note: “Zipline just closed, uh, over six hundred twenty-five million dollars”

Q That's pretty good. Ok, so what are these 11 business lines? Can you explain that a little bit more?

A Yeah, I mean, you can slice it a couple of different ways. One way to look at it is, um, How much of our business is generating revenue from transactions? So, you know, we have transaction revenue, which you can think of as trading commissions, rebates from market makers, payment for order flow. Um, we have a card product, so we earn, uh, interchange from swipes, and those are all, you know, money we make from every transaction. Then there's also net interest revenue, which You can kind of think of as money that we make from being the custodian of assets and also lending against those assets. And so we have, uh, a spread that we take on the cash that we hold on our platform. That's probably the simplest one. So we offer a great yield on cash, uh, right now, I think, 3.5% with some added incentives on top. Which still gives us 75 basis points, uh, for, for ourselves. And, you know, the cash sweep program, we call it the high yield offering has gone from basically nothing in 2022 to tens of billions of dollars under custody. So that's grown into a big business. The margin book You know, which is a tool margin is a tool that active traders use to, uh, to, to use leverage on their portfolios. We weren't really that competitive on margin to be fair up until 2024. And then we really started competing on rates, competing on user experience. And now the margin book is, I mean, it's, it…

AI assessment note: “One way to look at it is, um, How much of our business is generating revenue”

Q capital. It's over. That was the, that was the whole idea. But then some people thought, oh no, this is an opportunity. This is an opportunity to roll up companies, to apply AI, to use these efficiencies. And one of those people, or one of those funds was General Catalyst. And so you guys actually like really leaned in hard for this. What was the decision making process for it?

A Yeah, for us, it was a boat strategy. So we actually think three approaches can win. One, some folks view it as the model companies will end up being the winners. They're gonna keep iterating. They make better and better models. They'll go direct to the consumer. And we're obviously investors in Anthropic, large investors from the last round, from this current round. Um, we absolutely think companies like Anthropic and OpenAI and Google can benefit and the model layer can benefit. So we agree with that. We also agree with folks who say, well, it's been overestimated. SAS is not dead. If you look at public market comps, they've continued to grow, do well. There are all kinds of new, interesting SAS companies that are being started and growing. But we also think there is this third bucket. So we're kind of an all of the above. We think there'll be winners in each category, and we obviously want to be part of them. But that third bucket is for really fragmented industries where it's very hard to sell into AI native services and products. Can we actually build the AI native service platform and software and then go buy our distribution? So a few examples. One that was very early on was Crescendo, where we led several rounds of funding to build AI native software for call centers. And we teamed up Andy Lee, who ran Alorica for over 30 years, a call center chain with two amazing CTOs…

AI assessment note: “we actually think three approaches can win... that third bucket is for really fragmented industries”

Q That's so fascinating. Can you talk through the funding mechanism part of it? Like how do you structure it?

A Yeah. So we normally do one or two rounds to build the software piece. Like with the Titan MSP, for example, we gave them kind of a traditional seed series A to build out AI automation for IT services for MSPs. And they went out and they got six pilot clients and over three or four months, they sort of showed us, Hey, look at all the tickets that happened in MSP. We can now automate 38% of them. So when that happens, then we start talking to them about a second round of funding. We say, well, let's actually go look for targets because if we can buy one of these companies and automate 38%, we're certainly going to increase the margin and be able to reinvest a lot of that free cash flow into growth and create a really successful company. And so we look at a lot, especially the pilot clients that they're working with and who could, who really wants this AI technology. And unlike private equity, we screen really hard for does this company want to change? Do they want to implement AI? Um, and then of course we want to hold them for seven to 10 years and go public and not necessarily add debt and cut costs. That's a really different model than private equity. Um, but in their case, then we gave them a second round of funding when they found their target RFA. And now that they've found that target and, you know, are doing really well on the AI transformation, then they say, okay, now …

AI assessment note: “investing between a hundred and a hundred fifty million in each of these projects”

Q to tailored projects, Turing helps top companies realize AI that's more capable, more adaptable, and more effective. With Turing, discover how AI can accelerate your business growth. To learn more, Visit Turing.com slash sorcery. Spelt S-O-U-R-C-E-R-Y. That's Turing.com slash sorcery. So given the fact that you've invested into so many companies through different waves, how do you think about investing in competitive companies? I'm sure there's been some crossover across.

A I try to be really cautious about that because, you know, I've started, um, two companies myself and, um, you know, I want to be very respectful of, um, founders and what their goals are and their wishes are and everything else. And so I, I often ask both sides. I say, Hey, this thing has come up. Do you view it as a conflict? Is okay if I proceed, et cetera. And so I usually wait for that. Okay. Um, I think that, uh, there's a number of circumstances where people think something's, I'd say, 90% of the time founders think that they're going to compete and they don't. They tend to grow in very different directions, especially if it's two early stage companies. Um, and then maybe a few percent of the time things actually converge. And sometimes they actually converge on the companies that you don't expect. And so I've seen more conflicts and things that I didn't think would converge than things would. And, um, I'll give you an example. I remember I was an investor in Square and the Stripe founders pinged me and asked if I wanted to invest. And so I pinged one of the key executives at Square and said, Hey, do you think this is a conflict? And he said, absolutely. This is a conflict. You shouldn't do it. These things are going to collide. And so then I texted Jack and I'm like, Hey, is this a conflict? He's like, no, go ahead. And so I went ahead and invested, so I'm glad I did tha…

AI assessment note: “I try to be really cautious about that because... I often ask both sides.”

Q Taking a step back, can you explain the evolution of this industry and models? You gave a really good explanation on This Week in Startups with Alex, and I think this was like phenomenal, like it was clipped and everything. Could you explain the evolution of that?

A This industry had a huge shift after the reasoning models came out late last year, O-one being the first. The shift is, before the reasoning models, this industry needed simple data. Gobs and gobs of simple data. You needed a data factory. Or, ah, this industry needed somebody to find people really fast. After the reasoning models came out, the game completely changed. Uh, this is like after O-one and deep seek. Now what the labs need is not a data factory. It's not a talent marketplace. The labs need a strategic research partner, somebody who can collaborate directly with their researchers to understand where the models are weakened today and strategically custom engineer data to improve the model's performance. Uh, they need a research accelerator. In-house researchers collaborate with teams in coding, multimodality, STEM, RL gyms, et cetera, to generate data that'll improve these models. And the data that the models need now needs to be hard. That is model breaking. It has to literally break the model. So you need humans who are smarter than the models. Uh, it has to be, uh, realistic. That reflects how real humans use these models to do real work. Only then AI will actually move the GDP of the world. It has to mirror real world use, not esoteric academic use cases that test whether you've hit the singularity or not. And third, the data needs to be diverse to covers every si…

AI assessment note: “This industry had a huge shift after the reasoning models came out late last year”

Q We need to start off with your background. You have one of the most intense backgrounds I've come across. Can you just break this all down? Don't leave anything behind.

A Yeah. Um, I spent a summer in high school, um, studying Al-Qaeda and writing, uh, pretty extensively on it. And, uh, it just so happened that summer was the summer before nine, 11. So like literally on, on nine, 11, I was like, oh, there's an Al Qaeda attack. I had this huge paper that had been written, um, critiquing us foreign policy with Al Qaeda up until that date, um, saying it wasn't gonna work. And, uh, that sort of kicked off this crazy sequence of events that, you know, Led me to get recruited into like a strange section of the Navy out of college, two tours in Iraq, doing fun stuff there, and then working at the intersection of emerging technology and defense, both in uniform and then out. Left active duty, went to the private sector, joined a startup, and then, yeah, had an old boss, Mike Flynn, call me up and say, hey, come to the White House, and then spent four years And owned all cyber, telecom, supply chain, and crypto policy at the White House on the National Security Council for four years, and then started Galvanic. It's been, ah, it's been a ride so far.

AI assessment note: “spent four years And owned all cyber, telecom, supply chain, and crypto policy”

Q We've been involved in like this really rapid increase in drone warfare. I'm really curious to get your perspective on the evolution of this and like how you even decided to go after this so severely and like, what's the current state?

A Well, basically what we saw is, um, I sold my last company to DoorDash and then right when the deal closed, um, Putin invaded Ukraine, like pretty much like the next week. And we saw very early on that, um, the countermeasures we were investing in as a country, like electronic warfare, Um, or like microwave technology. I, I don't think those are going to be like long-term sustainable for some of the new drones of the future. So we saw that early and started to work on kinetic defeat solutions because you can't really armor these drones enough against a bullet. So this will be like a much more reliable system, especially as the drones start to add AI and then you can't jam them anymore.

AI assessment note: “we saw that early and started to work on kinetic defeat solutions”

Q Okay. That's a lot. That's big. So how does this enable the warfighter? Like, can you give a more tangible like use case or like day operation?

A For sure. So, you know, high level, we can sort of boil it down to just the performance. So a surface vessel today is going to go 20 knots, you know, mid twenties miles an hour. That's when you start skipping off the waves and stuff like that. So a sea glider can fly at a 160 knots or a 180 miles an hour. So we're eight times faster. So if you translate that into Moving around island chains in the Pacific, you know, Guam is about as close as the U S gets to China. And then the first island chain is, you know, Taiwan and, uh, Japan and sort of those islands there. Um, so that's like a four or five day sale. If you're a surface vessel today and a sea glider could fly there in 10 hours. So it just gives you a sense of how quickly we can reposition things around island chains in the Pacific, which is the critical theater. Um, but then also you have the, the capabilities of the vehicle compared to aircraft. Cause you could say like, oh, well I could fly a plane there, but planes need fixed land infrastructure and sea gliders can take off and land in five foot seas. Uh, that's like the key differentiation of a sea glider verse sea planes or, or flying boats of the past. So we're basically, we're a surface vessel and that we can, you know, do open water operations, but rate times faster, or we're an airplane, but we can land and take off in these, in these high sea states.

AI assessment note: “gives you a sense of how quickly we can reposition things around island chains”

Q Wow. Yeah. And so within the conditions of those factories, what is that like?

A Oh, I mean, just in terms of them not being great jobs? Yeah, I mean, a former senior vice president of Apple told me that he would never be able to do the job of a normal laborer in a Chinese factory. Um, yeah, Foxone's not a great place to work. Um, let's see, uh, you're not allowed to speak, you're not allowed to smile, you often have nine to 11 seconds to perform your task, and it's just grueling work with a couple of 15 minute breaks for 12 hours a day, often six days a week. Um, I do push back against the idea that this is slavery, however. Um, you know, iSlavery became something that, that, that became a term of art, um, after, um, when there were a number of, um, suicides at Foxconn. Um, and the company's response to this was awful. I don't know if you know, they, they, they built, like, nets around the factories, and basically asked, this is Foxconn, not Apple, asked workers to sign pledges that they wouldn't commit suicide. Um, I mean, conditions have to be pretty dire, and your response has to be pretty Um, awful for, for, for that to be the actual policy for a company with literally at the more than one million workers at the time. Um, but you have to remember that the people working in those jobs are coming from the rural hinterlands where if they don't take those jobs, they're working 14 hours a day rather than 12 under the hot sun toiling in the fields. So you ca…

AI assessment note: “you're not allowed to speak, you're not allowed to smile, you often have nine to 11 seconds”

Q Augustus, I'm sure everyone knows you because you are a social media figure. You are now a news, like, segment every week. But Alex, we need to talk about you. So could you share more about Atmo and what you're building?

A Totally. Uh, Atmo is the first and leading company in AI for weather. So we started in And we invented the first deep learning neural networks that could predict the weather. So the same way that LLMs can predict the next word or phrase or paragraph of text, and that's turned into this entire world, uh, of, of AI companies. We did the same thing for the atmosphere where we can make, you know, these high dimensional models that predict the next minute, hour, day, week, month of what the atmosphere is going to do. And that's let us make the most accurate and precise forecasts in the world. And we run these every day now for the most important organizations. So we build next generation forecasting systems for the US Air Force, the Navy, and entire sovereign countries. So most recently, we just completed an upgrade to the entire Philippines National Weather Service that increased their level of accuracy by as much as 50% and made it far more detailed than what they had before, which was quite important because they just had, uh, uh, 22 typhoons in one year.

AI assessment note: “Atmo is the first and leading company in AI for weather.”

Q Let's break this down. So like, what was the process like and why are these the perfect investors for it?

A So the process went faster than we were expecting, uh, when you have, you know, a track record of being able to execute and use words like on time, on budget, uh, building for America. And here's the proof. We were able to go from initial conversations to sign term sheets, uh, in about four weeks. Actually, maybe a funny story. I just had my first kid about a month ago. I signed the term sheet, and my wife went into labor four hours later. So the timing was phenomenal. We got really fortunate on that, and the reason why these are the right investors, when you think about, you know, Eric and Brad at Altimeter, they understand hardware. They understand how to scale Businesses outside of pure SaaS. Uh, when you look at their portfolios, when you look at their personal experiences, They understand what it takes in terms of workforce development, thinking about who are the right customers to have? How do we go and build this into a generational business? And when you think about funds like One Investment Management, they do infrastructure, they do mining, they understand project finance and all of the different pieces that we'll need later stage as we go into hyperscale.

AI assessment note: “We were able to go from initial conversations to sign term sheets in about four weeks.”

Q Are there like, do you have to go after different approaches for those industries? I'm sure because you mentioned, and I know this too, but like, because venture is such a wonky asset class, internally it, they act like startups. Private equity is more institutional. So do you adapt your approach for servicing these companies throughout that, or is the product pretty linear?

A It, the product is not that different. Uh, what's really different is, uh, the go to market motion. We started with a sales team that sold, you know, 6000 dollar cap tables to 24 year old founders in hoodies, uh, you know, in Soma. And, and now we're selling million dollar accounting solutions to middle aged, uh, CFOs in midtown. Like it's just a very different, uh, sales motion. So that's been the biggest probably challenge for us to, to adapt to, to a new industry. Uh, another way of thinking about our business is we're very much a network business. Uh, and, uh, we try, we try to build networks, uh, directly into the product. And so obviously our, our first network was cap table to investor. Our second network is, uh, investor to LP or fund LP. What makes venture unique as an asset class for us is that it has what I'll call a global topology. You know, you bring a company on to Carta because all the investors are minority holders. They bring a bunch of investors. They bring 20 investors. And then those 20 investors hopefully recommend some startups, which bring more investors. And you get this very complex web of a global topology. In private equity, it's different. A KKR portfolio company that buys Carta cap table software doesn't help us with a Carlyle. Uh, portfolio company. They're just completely separate. A KKR portfolio company that we sell helps us get another KKR por…

AI assessment note: “the product is not that different. Uh, what's really different is, uh, the go to market motion.”

Q People know about nuclear broadly, but they might not know about the fuel aspect of this. Could you just explain a little bit further nuclear fuel, or I guess like the nuclear process from, you know, maybe enrichment to actual generation?

A Yeah. So there's really, there's really two, two big parts. One is mine the uranium, get the uranium, and then everything else is processing. So it's like chemical processing, refining. So there's four steps there. One, one is it's typically turned into a gas. It's then enriched. It's then turned back into a solid. And then once in solid form, you can make whatever shape of fuel pellet you want. So those four steps are conversion, enrichment, deconversion, and fuel fabrication. The step we do is enrichment. Which is, you know, the, the first step of conversion is a chemical process going from solid to gas. So that's chemical reactions. The enrichment process is really a separative refining process. No chemical reactions, um, no nuclear reactions. And then you go back into solid and you, then you do mechanical shaping. And so, um, that's it. It's, it's mostly manipulating matter. Into a final form. Uh, no real nuclear processes at that point. It doesn't really go critical until, until you're in a reactor in a specific configuration of fuel. That's when you get radioactivity and energy production.

AI assessment note: “So those four steps are conversion, enrichment, deconversion, and fuel fabrication.”

Q I guess, how did we get here as a country? Like nuclear is definitely stagnated over almost 50 years now. So how did we get to the point where there's, there's actually no enrichment really here?

A At this century, it's been pretty flat. Um, yeah, not much new nuclear production, even since the nineties. And so, you know, you think about that context of 35 years, 3035 years of no real growth in nuclear energy. It's natural to think maybe we don't need a lot of field production and maybe we can trade for what we need. If you rewind to the fifties, sixties, seventies, US was most of the world's enrichment. Uh, we were 65% roughly, and we did that through a first generation technology called gaseous diffusion. So we had these big, big facilities, very energy intensive, um, gen one process, gaseous diffusion. We, that's how we led the world in, in enrichment. Fast forward through the cold war to the fall of the Berlin wall to not much new nuclear build in the U S and, and we said, okay, Cold War's over. Um, we have allies in Europe. We're friends with Russia. We should just freely trade for, for, for uranium and for enrichment services. Then we started the process of decommissioning our gaseous diffusion plants in the U S and said, we will solve this with free trade. Our allies and Russia do this much cheaper than we do it at large scale and free markets are going to be fine. Um, you know, in addition to that, the DOE was working on new technologies. Thinking that that would be what scaled up in the U S and that we would have self-sufficiency and, and we just haven't really s…

AI assessment note: “we started the process of decommissioning our gaseous diffusion plants in the U S”

Q explain it all, but you've been at SpaceX, you're at Founders Fund, you've been helping out some really critical, uh, national interest companies within Founders Fund, and you were talking about the playbook that these companies use to build these generational companies, these, like, really Goliaths. So I'm really curious from your end, like, could you just share more on your background and how that's informing building out the company?

A So I started my career as an aerospace engineer, worked at Boeing during college, then moved over to SpaceX also during college, and then immediately after graduating, that was from 2003 to 2007. And back then we were working on the original engine systems. We were working on the Falcon one vehicle and the NASA commercial orbital transportation services contract, which was structured. It was a milestone based contract to bring back us launch capacity. And cargo capability to the space station. Back when we had, we didn't have that capability once the space shuttle was grounded, and we were completely dependent on Russia for all of our crewed and cargo space launch. And so, um, you know, that experience just showed me how that, how that company was built. It was really a combination of two things. One was people from the industry, aerospace industry, who knew All the tricks, knew the science, knew the engineering, um, but had a way of doing things that was more the incumbent way. And that was combined probably in a one third to two thirds ratio with Silicon Valley talent, people from hardware startups, software startups, just saying, let's treat this problem as if it were a startup problem, almost like it's a software company. And so, you know, there were, there were naming conventions around the rockets that were like more like software releases. Um, that was something from a v…

AI assessment note: “And so that's the same thing we're doing. So people from the nuclear industry”

Q But could you just give a high level overview of Anduril and maybe their evolution over the years?

A Yeah, so they're building a modern, Anduril's building a modern defense prime, um, you know, kind of leveraging, uh, software, AI, mass manufacturing, uh, designed for mass manufacturing. And, you know, it started out with, I think initially it was like a border, uh, security company, Uh, and I, I believe their first product was like a, a century tower. So kind of like a surveillance tower. You know, they now have, you know, many product lines across air, land, sea, and space. And so, you know, they made a re recent acquisition that got announced as company Klaus, which is a ruggedized, you know, edge compute, uh, and networking sort of, you know, product and platform. You know, they've got a whole suite of products, you know, underwater with, Dive and, and Copperhead and, and, you know, sensor networks. They've got all types of, like, counter UAS. Um, uh, they've got drones. They've got Fury, which is their, um, group five, uh, autonomous, you know, drone for, for the CCA program. So they've got.

AI assessment note: “Anduril's building a modern defense prime... started out with... now have... many product lines”

Q invested into Sun Microsystems, AOL, Intuit, Amazon, Google, Zynga, Looker, Slack, X, Spotify, Square, Brex, Plaid, Robinhood, Rippling, Glean, Figma, and many, many more. Wow. I'm like winded from saying all of that, but you guys have certainly made a big splash, um, over the course of the 50 years that you've been around. So I'd love to understand how big is the fund now, and what is the thesis?

A Absolutely. So we're currently deploying out of, this is our, our 21st venture fund, KP 21, uh, it's eight hundred million dollars on the venture side, and then on the growth side, we have a 1.2 billion dollar growth fund that we call Select, Select Three. What's maybe a bit unique is, They're relatively similar fund sizes to our last set of funds. I think what's very important to understand about KP's approach is that we believe that venture doesn't just scale with money. Uh, and there's elements of the business, especially at the earliest stages that are just so unscalable, that are so relationship focused, that that's why we've kind of kept the funds a relatively similar, similar size.

AI assessment note: “eight hundred million dollars on the venture side, and then on the growth side”

Q What kind of principles other than diligence have you learned from him? Like, what, what would you say, like, seriously carries on into how you view companies or view new trends in the venture capital world?

A Yeah, I think one of the most important things I learned from him is that it's all about the people. I think even, you know, when I was angel investing before joining Kleiner Perkins, I was really enamored by ideas, and I was very thesis driven, uh, and, you know, I had this view of where I thought the, where the world was going, and John Doerr taught me Like look to the founders to tell you where the future is going and really focus on, on the people. Um, and another lesson I learned is like, you know, some company, uh, some VC firms rather, uh, you know, take a more active stance and swapping out CEOs if the company's not performing well. But, uh, something I learned is like, that's usually, it usually doesn't work out. So, you know, you're really, In with this founder for the long haul, uh, and it, which is why it's so important that you're backing the right people.

AI assessment note: “one of the most important things I learned from him is that it's all about the people.”

Q that is VC. We're the innovation class. We're supposed to help Create new technologies and integrate into people's lives, make them much more efficient and better and fun. Um, but just to like lay it all out from your perspective, um, given you've spent a considerable amount of time there, what are the current trends of VC dollars, funds, companies, and success stories that have come out of the region?

A Yeah. I'm almost like, why is everyone so obsessed with this? Like, you know, and it's interesting. Like, I'm glad you asked the question. It's like, People are trying to take down Europe, calling it a museum. I'm like, come on, guys. Um, I mean, when you look at the numbers, it, it, it paints a, a great, good picture, actually. VC funding in Europe reached about thirty billion the first half of 24. That's a 12% increase to twenty-twenty-three. Um, you know, if the pace continues, I think twenty-twenty-four will become the third most active year ever for VC funding in Europe. You know, obviously those twenty-twenty-one years were Insane across the board. Um, and regionally you're also seeing a lot of diversification. The UK has raised the most in total. I think it was 9.5 billion. Um, but France has 4.3 and it's becoming an AI hub. And you look at dock region and you're seeing tons of companies come out of there. And then of course the Nordics has produced a bizarrely disproportionate amount of unicorns. New York Times just wrote a Sorry about this the other day. Um, and you know, we're seeing big deals as well in, in Q one of 20, 24. I think Europe had 139 active unicorns for new additions. Um, so wave at the big series C, Mistral's, six hundred forty million series B round. There's, there's a lot going on. Um, and you have, you know, talent flywheels emerging as well. I, you …

AI assessment note: “VC funding in Europe reached about thirty billion the first half of 24.”

Q Yeah, that's really helpful, and that's a good perspective. The three pillars, um, for conviction. Yes. Okay, so I guess, like, going back to upfront strategy, while I was there, you were operating with the barbell strategy. Are you still operating with the barbell strategy, and can you explain that? Yes, I will explain it.

A Uh, I will explain it, and yes we are. So barbell meaning, uh, if you imagine a barbell that you might lift, then it's got two weights on, uh, either side and a thin bar in between them. So we invest at seed stage, and we're typically writing a three and a half million dollar check. We can write two, we can write five, but let's say the median check is somewhere between 3.2 to 3.5 depending on the vintage. And then we don't do a lot of A and B investments. I'd say almost zero. Now that doesn't mean we don't follow on. We will follow on and do our prorata, but we're not entering at the A or B round. And we have a very specific reason why, which is The market for A and B between 2010 and 2022 went up by nine X. So nine times more capital in those markets than in any other sector of the market. And as a result, valuations went up like 300%. And it just, you got priced out because if you had a two billion dollar fund You know, let's say two funds ago, they were three hundred million dollars and they were writing five million dollar checks. Now they raised two billion and they want to write twenty five million dollar checks. And you can't write a twenty five million dollar check at a 15 free, you know, and take the majority of the company. So you end up paying 60 free. And so I think that a round is the single most overvalued round in venture capital. Now, it just so happens that if…

AI assessment note: “Uh, I will explain it, and yes we are. So barbell meaning”

Q For people that don't know, what is the biggest difference between your engines and the other engines that are on the market?

A Um, well, I think our engines are very much designed for the Neutron vehicle, and, um, you know, it is a reusable launch vehicle at its heart, so what that means is, um, our approach to engine design is slightly different than an expendable engine, because an expendable engine, you know, it needs to run for a 190 seconds on the first stage, say, and then it never needs to run again. Whereas, you know, a reusable vehicle, especially the way we're doing it, we want it to just run and run and run and run. So, you know, the qualification, um, burn time for an Archimedes engine is one hour. So the engine, you know, has to do 40 starts and run for one hour. And whereas, you know, Rutherford for Electron, because it's a single use engine, like its qualification or its, you know, acceptance test is, is like five minutes.

AI assessment note: “our approach to engine design is slightly different than an expendable engine”

Q So what kind of decisions did you make there?

A Well, I mean, I made a constraint in the very early part of the program where I think I lost a lot of credibility with the engineers, and I, and I said, look guys, we need to be able to turn this vehicle in 24 hours, which is, I'll admit, is a stupid constraint, but what that constraint did is it, it drove a whole bunch of really, really important design decisions like, you know, propellants, engine lights, um, you know, you know, structures. The Hungry Hippo fairing came out of that. Because it's like, guys, we don't have, you've got 24 hours to fish a fairing out of the ocean, recondition it, blow all the salt water out of it, and get it back on a rocket. It's like, that's not going to work. So, you know, that one, that one, you know, requirement drove the design for the Hungry Hippo fairing, so we never throw away a fairing. So sometimes you've got to be cruel to be nice, and that was, that was an example where we put this unrealistic constraint on the team, but it drove a whole lot of goodness everywhere out, you know, throughout the vehicle.

AI assessment note: “I made a constraint... we need to be able to turn this vehicle in 24 hours”

Q Ok, I have a hard question. How are you going to become a trillion dollar company?

A I mean, as an advertising business, again, I said this earlier in the pod, but you got to make money, otherwise you don't have a really good advertising platform, and we fortunately make a lot of cash, and we like to think, just in traditional finance terms, you're worth your cash flow after SBC and what you expect that to be many years out into the future and the terminal value of it. And if you think about where we are today, We probably on an EBITDA basis, I think we're over a seven billion dollar run rate, and we generate somewhere around 75% cash off the EBITDA dollar. Now, the business itself, and then that's sort of after SBC. If you took us three years ago, that number was much, much smaller. It was probably one 20th of where we've gotten to in three years. To take it from this scale, Up and be worth a trillion dollars. We've got to believe that we can get to thirty billion plus of cash flow a year. If we could get to thirty billion dollars plus cash flow a year, depends on how quickly you can do that in the future. Investors probably give you a pretty good multiple on like cash flow to get you to that. And so when we think about the things that we're working on, everything that we work on has to have a very large economic opportunity. So, um, typically in investor terms, you think about total addressable market. Well, The games category we've done really well and we bu…

AI assessment note: “We've got to believe that we can get to thirty billion plus of cash flow”

Q So what's the main difference between using something like that versus what you're using?

A So our airframes right now are, are mostly made out of, um, carbon plates. So carbon is really light and really strong. Um, but it doesn't have some of the advantages of what they were doing, uh, with those metal bodies is that you get to use it as a heat sink. So we actually do this on our drone now where the bottom plate is a, an aluminum plate, and we, we, uh, contact the, Motherboard and the radios with thermal interface material when we close out the drone with that bottom plate. Um, and so that acts as a big heat sink, but we're still using, it's just a flat plate that's cut into a shape in two D versus having to be cast or CNC'd, which is just a lot cheaper for, for mass manufacturing.

AI assessment note: “our airframes right now are, are mostly made out of, um, carbon plates”

Q This was a question I asked Saronic, but how do you deal with completing missions when communication goes down, but they might still be active? Do they still complete the mission? How do you, how do you deal with loss of communication?

A Yeah, so, um, with this drone, if it gets jammed, then that's kind of game over, because this one isn't autonomous, but we're also announcing Archer AI, which does have autonomous capability. One of those things is terminal guidance, where basically, uh, jamming is usually the strongest on top of valuable targets, and so it just gets exponentially harder if you're an operator, if the operator is sitting 10 kilometers away, as you try to get within hundreds of meters of the target, The strength of the jamming is just so high compared to your communications link. So that's where a lot of drones fail. So with Archer AI, we are, um, introducing the terminal guidance feature where once the pilot has selected the target and the drone has a lock on it, it doesn't need input anymore. So even if radio comms get cut, it will continue in and hit the target. Um, we'll be expanding on that feature set and, and trying to build Um, more and more useful autonomy that does allow for better mission success in contested environments and, you know, when comms are totally lost, but usually it's actually a, a gradient where, you know, comms might be spotty and you need something to kind of get you through a rough patch or get you right through the end of the mission.

AI assessment note: “even if radio comms get cut, it will continue in and hit the target”

Q So, for people who don't know what Optical connectivity means. Can you explain that? We're moving from copper wires to optics and transferring data with light, but what does that mean?

A Obviously there are fibers and fibers are essentially glass. They're just conductors. Instead of having wires, copper wires, you have fibers and fibers basically contains the light, just like a wire or a copper strand contains electricity. And so what you get with Uh, fiber optics is the ability to move data faster over longer distances with less heat and less power than you get with, uh, electricity. So if you think about the way data centers are architected today, it's mostly copper wires, right? Ethernet cables are essentially copper wires. And copper wires, the physics of copper is there's a resistive property. So As data tries to move across the wires, it gets hotter, and the resistance prevents it from going all that far. And so now inside the data center, you have speeds are increasing, bandwidths are increasing, and copper can no longer carry these signals, uh, the distances they need to go. And so you have more and more optics now coming into the data center. And optics essentially is just light, right? We're just shining light along a fiber, and that, that light is modulated, meaning it turns on and off at really, really high speeds, and that's what creates the, the transmission.

AI assessment note: “fibers basically contains the light, just like a wire or a copper strand contains electricity.”

Q So you joined a CEO in 20, 25. What is it like to take the reins in the craziest market ever? And also give a little bit of your background of where you were before.

A Yeah, I mean, this is, uh, you know, an interesting thing. I think if you look at most CEOs, They either retire or somebody retires them for them, right? You don't see too many CEOs change jobs. And I've actually been lucky enough now to be CEO of three different public companies. The first public company I was a CEO of is a company called Finisar that was optical. That's now the heartbeat of one of our biggest competitors, but also one of our biggest partners in Coherent. Coherent's built around The engine that was Finisar that we ended up selling to them. Then I was a CEO of a semiconductor company. I'm mostly a semiconductor guy, and I was CEO of a company called Synaptics for five or six years. And then I joined Lumentum, as you're pointing out, at a time where The board of directors showed me a forecast and they said, hey, here's what we think the company is going to do. And I, of course, as one does, discounted it by about 50%, and it's been up probably four X over what they even showed me. So we've just hit the company with the right products at absolutely the right time. Really, our team, before I got there, the previous CEO set it up well, right products, right roadmap, And we have all of the goods that people are looking for now as they convert their data centers to more and more of the fiber optics.

AI assessment note: “I've actually been lucky enough now to be CEO of three different public companies.”