22:27
Hi Lior, welcome to Strictly VC. I didn't realize you dropped out of high school. I know. That's so interesting. Right now we are Talk about vibe coding. Yeah, there are a lot of founders dropping out of college and high school now. So I don't know if you, you know, before I really jump in if you want to comment on this topic. it to college. High school was fine enough for me. Yeah. So you were here last time in 2021. And remember it was really cold. Connie and I were sitting outside and she was freezing over there. Uh but it was a fun discussion. And your fund Eclipse invests primarily
in physical world products. And for some time it wasn't a very popular area to invest in because everything was SaaS and fintech. there was some excitement in 2021, but a lot of that was driven by kind of the free money, the ZIRP era. And then we had a little bit of a lull. So can you tell me what you think changed between that time and where we are right now for the companies that you invest in? I mean I think when we started Eclipse uh 11 years ago, um our view is coming from an operating background into the investing world. Um we build and invest at Eclipse was a very basic fundamental view that the world GDP, 85% of it is actually not in the world of software and IT, it's actually in manufacturing and semiconductor and data center and
defense and mining, etc. And building the companies that digitizing the physical world, the world that we live in. Uh no, it's not a vibe coding. That's cannot actually help us or pay the bills uh or grow bananas for that matter. Uh you need actually atoms. And atoms and bits behave differently when you build. Uh naturally the richest person on Earth, the two of them. Um so it's actually very profitable, but it's hard to do it. They've been built different. Um and when we started in 2015, it was very controversial. Uh I think I was the all era of software developments and enterprise softwares and SaaS and uh it felt fairly lonely in the first uh couple of years. Things changed drastically.
Um and we're seeing it in uh the public market very soon. We're seeing it with companies growing to be uh very large companies. And um uh and I think we are seeing it in the founders. The best founders in the world right now wants to solve really tough problems. And they usually live in the physical world. Yeah. So are you seeing more capital going to this area? I mean because we're definitely seeing the excitement in robotics is like something we've never seen before, right? We have several companies that are raising money at very high valuations. So are we at the cusp of something really new and happening in robotics? Can a robot do something for us that this actually hasn't done before? Like can it go into my house and
do my laundry, wash my dishes? When can we expect something like that to happen? Yeah, I think um I was in event of uh with the Jamie Dimon from JP Morgan couple of weeks ago in DC. And I was on stage with uh Sanjay, the Micron CEO. Um and you know, majority of Silicon Valley have no idea who's Micron. Uh not too long ago. Actually company is manufacturing memory chips or HBM. Um you see right now in the public market the highest all time of market caps of companies like TSMC and Micron and SanDisk and of course the Nvidias and AMDs and the Broadcom and Qualcomm and ARM of the world is just doing
extraordinary runs in the public market. And it's because I think we are seeing the movement of people understand that real moat in software gone. Right? You can vibe code pretty much whatever you want. So there is no moat there. Um what you cannot do with vibe code is manufacturing wafers. Cuz you need machines and you need them to have silicones and they need clean rooms and they need bunch of other things. So I think we are seeing the shift from a very easy economy, very soft economy, software focusing on low cost or high gross margin, low capex, very fast, very easy uh direct to uh enterprises, low ACVs.
This entire market shift back to the physical market. Robotics is one of them. And give you a one interesting stat that I'm tracking. Last year our portfolio So we have roughly 90 companies we built, roughly 95 and invested in roughly 65. Uh the portfolio last year raised 14 or 15 billion dollar in equity. The portfolio in the first quarter of this year raised four and a half billion in equity. In the first eight years of Eclipse, the portfolio combined raised less than four billion. That's incredible. So what do you think changed over that period of time? Like what is it about this moment that makes capital allocators, you know, of different kinds say, "Let's put money into physical goods technologies." I
I mean in the end of the investors are following on one thing and it's returns. So there is I think a strong belief that returns are coming right now. Naturally you have SpaceX going to have the largest IPO in the history. By no mean not a software company. Probably the most hardware asset heavy that the world ever produced going to call it closer to two trillion. Um so I think investors are seeing returns. I think um you're seeing from geopolitics point of view we grow up into a globalization 50 or so years uh that we and China uh and other countries kind of play ball with each other and we outsource a lot of those businesses into other countries more low cost. Deglobalization is kicking in.
Meaning we needs to manufacture more things for ourself. Meaning there is more um geopolitics uh reasons to fight. So there is more war. So defense spends go up and we saw all of that in the defense companies. And what we are seeing in defense tech, uh you see energy crisis in Europe with the war uh in Russia. So you know, energy storage and nuclear. You see data centers, they need chips. So it's actually not the one thing. It's like multiple of trends that compounding on each other that just pushing the envelope now. And the last thing we talked a lot internally about the five forces at Eclipse. And five forces is actually something I saw in China when I was there in 2013 when I uh lived in China. And it's basically
capital, um talent, policy, customer demand, and technology. This is the first time I believe in America ever. I used to say from uh Henry Ford and Carnegie, I actually change it. Ever. That those five forces is a line and that's for a builders like us, this is like the best time to build those companies. Yeah. So because when we think about it, because we're so focused on AI. AI is such a revolutionary technology, sometimes it feels a lot of the demand and a lot of the innovation in the hardware space is driven by AI. But you're saying it's only one of the factors. Yeah. I think it's just one of the factors. I do think I do say, you know, go back to your previous questions around robotics. I think when you
combine physical AI or AI and robotics, you are seeing the ability to use those robots right now in different environments and manufacturing and industrial facilities and automation. I mean, you see here the probably the most successful robotics companies down the street, Waymo cars. A car driving by computer with sensors, it's a robot. Um so I mean, not too long ago people told me that self-driving cars it's 50 years away forever. I mean, this thing is doing L4 and driving significantly better than my parents. Uh and significantly more safe. Probably better than me. Yeah. Uh-huh. So um I think we are going to see um technologies that sounds to us very
crazy not too long ago. You talked about robots in your home doing useful stuff. I guarantee you're going to see it this year. This year. What is going to be the cost of a robot that's going to do things at in my home this year? Because like we had this conversation with a friend recently. She's like, "Everybody's talking about AI. What I want is a robot that's going to do some of my housework that's going to be cost as much as a washing machine."
Yeah. So I'll give you the full scenario. By the way, the first time that I um really went uh and I love robotics. I'm doing robotics for 15 years. Um the first time I actually saw where it's rich with VLAs and you know, basically SLAMs and uh reinforcement learning, I really felt uh I'm getting closer and closer to divorce because there is no reason for my wife to be married to me if that robots will be existing in the world. Uh And the point there was I this year you will be able to buy in roughly $5,000, maybe four. We will see. We are working on the supply chain. A robot that is actually not that big.
Um that's uh you will unpack the robot. It will drive in your home. And it will be able to do the following task. You will be able to say, "Hey, can you uh take the clothes here from the living room and put them in the washing machine." And it will just go and do it. And it will fail then it will try and it will succeed. Oh, by the way, clean the kitchen. And it will recognize what needs to go to the sink and what needs to go to the garbage can. And tasks like that this year you will be able to see it operating in your home. And a cost of not too far from what cost like you know a fancy washing machine. So, it's coming and it's coming fast. It's what feels like Waymo. So, this is a startup that you
invested in? Let's say I know about someone that working on something. Sounds like something in stealth. But 4,000 to 5,000 How many like does it need to be trained on my home? Like was it going to walk around? Nothing. So, it will figure it out right away. each other what the main thing it will do it will just will drive and create the map by himself. You don't need to actually do anything. And the other crazy thing like you're using ChatGPT and it start to get to know you and it's like, "Oh, Marina, but you know you we know I know." It's actually start learning you. It's learning also the behavior of the home. And how cool is that you can open the app and say
you're driving home and say, "Make sure that the house is clean when I'm coming." And you're just coming and it's like clean. Yeah. What a The best house husband ever. What a time to be alive if we have that. It's like when they introduced you know washing machines in the '50s, that was a huge revolution. So, you're you're predict that we will have somebody like a robot a machine that's going to do even more of our housework for us now. This year. Wow. Very exciting. So, let's talk about SpaceX IPO really quickly. Um I think it was a couple weeks ago was it a week ago? Time flies really fast. Um SpaceX announced that they have an
optionality to buy Cursor which is essentially web coding company. Probably leading web coding company outside of Anthropic's cloud code. And our kind of interpretation of this is that Elon wants the Wall Street and public investors to give him some value for xAI and AI. So, essentially AI companies have can garner higher valuation than a space company would. What do you think about that? Like why is AI more important now? I mean I think um uh We can probably spend 2 hours on what Elon is doing what not and do I like it or I dislike it. But I mean I think the short version is um and that's only my point of view. So, it might be completely off. Um building
building foundational model cost a lot of money. You need significant amount of infrastructure. You need chips, you need computers, you need storage systems, you need networking, you need data center, it's need power. There's massive amount of capex, there is construction. Um And I actually my two cents he struggled to find how he's going to raise enough money to xAI by himself as a standalone. But he knew that in the other side he actually create a cash cow with SpaceX and Starlinks. And he figured that if he will use the profits from that side in order to fuel and actually finance the xAI, he can actually get now the market the public markets to actually give him a premium because he have an AI play inside SpaceX that something he done in
here. So, in some way win twice. He's getting the cash flow that he needs in order to spend on the money and he get a better multiples. I think the deal with Cursor is very simple, right? He's uh they're super behind in xAI and he needs to keep up specially on you know trying to compete with Anthropic and Codex by Anthropic and OpenAI with their Codex and you know he just figure out if I either can buy it or build it and build it is too slow and the market is already moving too fast. So, he's going to trying to buy it. Yeah. As a physical tech investor, do you think public investors are wrong to value AI at such a premium compared to physical
tech? I mean essentially the SpaceX on its own would garner a lower valuation than SpaceX plus a strong AI play which he may or may not have. I mean arguably you know the highest market cap in the world is a hardware company today, right? Is Nvidia that building the picks and shovels for AI. So, um and actually if you go to the list of Fortune 100, the majority almost all of those companies right now you just saw by the earnings all of them increasing capex. The three hyperscalers and majority of the Fortune 100 companies that have any to do with AI is actually hardware companies. Not software companies.
There's actually not a single one soft pure software companies and you know you say, "Oh, Google." I'm like, "Nah, TPU." Most of the money they spend is actually on the chips not on the models. So, I do think the public market start valuing again real asset cash flow, EBITDA and a little bit less easing this whole notion of maybe what Silicon Valley was so obsessed with in the last 20 years of oh gross margin is a magic. I don't give a about gross margin. I care about EBITDA. Like you can have 50% gross margin and have 20% EBITDA. You can have 85% gross margin and lose money. So, so I think we are seeing that shift the public market. I mean we spending a lot of time getting close to the public market by ourselves and
the demand there for real businesses that have a real mode that in most of cases have a physical element is never been so high. I also wanted to ask you about Waymo and capex that's needed to support many of the companies that you're essentially investing in. So, I don't remember where I heard this or read that Waymo would not have been able to you know become what they are right now. They wouldn't have been able to raise so much capital had it not been for essentially very low interest rate environment for such a long time.
Cuz that was just very intense expensive um investment. What do you think about other things that you're investing in right now cuz they're also capital intensive and we're not in a low interest rate environment right now. Yeah, you answer the question you answer in the way I was going to answer. We are not in a low and I just give you a stat that we never seen so much capital going into the physical world. The other thing actually with Waymo it was not the low cost of the it's actually was Alphabet that was financing Waymo from their balance sheet. So, actually it's almost like neutral for the debt market because it's Alphabet that was financing it. We have I think what Waymo paved the way and we
have a company called Wave in that sector that actually doing end-to-end and much lower cost version of collecting the data that is the majority of the spend there when you build a self-driving car. But I think what Waymo did and I give them a full credit is go a little bit back to your questions on the robot this magic robots in the house that will do these jobs. They took something that was very frontier notion. It's like, "Oh, self-driving cars. It's probably will happen in one day many years ago and it will be my grand grandkids." And actually show it that is right now and it's here. And it's can change how we are move as people and goods.
So, you know, we spend on transportation every year in the United States 1.2 trillion. So, even if took your capex, if you can actually transform that industry naturally you're going to enjoy the upside. But in order to build the next Waymo, will there be something like Alphabet or you or do you need to rely on institutional investors to back those companies? it's actually the that's the mention of the company that we have called Wave. They're in the self-driving cars. They just announced the recent round. They raised another billion dollars. So, it's not nothing by no mean. But the progression that they were able to do to achieve what Waymo was achieving early on with 1% of the capital is
incredible and how they do it. They live they leveraging their case AI. So, they actually using end-to-end model and they teach the car how to drive from simulations and synthetics and they are like doing basically transfer learning something that Waymo originally didn't had because they did a rule base. It was pre transformers. So, I think what we are doing we are taking all of this AI and innovation and technology and implying it now in the physical world and that's lower the amount of capital that is required to achieve those transformations. So, you started this conversation saying that there are a lot of forces that are happening in the physical world. But it sounds like the transformer model
is a huge force because it just lowers the cost to build a lot of the things that you invest in or it's also great tailwind because you have all this infrastructure that's supporting the AI spend. So, AI does come back to AI in many ways. Remember one of the five forces technology. All right. And AI is a huge tailwinds to transform physical industries and it's not only that, right? You know we saw before that the machine learnings and sensor fusions and compute and edge. So, we are taking all of the best in breed that is exist and kind of apply it to do the physical world and that's allowing a much, much faster of closing the loop to build companies over meaningful of SpaceX or
Waymo or Cerebras or Wave or pick your favorite company. Okay. Just one last question about kind of the investment side of things. After 2021, there was obviously COVID correction and a lot of the capital intensive companies couldn't, you know, they had a harder time raising capital. Now, how what are you telling your companies? How do they prepare if something like that happens again, which it will? Like what is the solution? So, actually 2020 20 2020 and 2021 was mainly a shock to the software companies. They actually had a much harder time because, you know, it was all-time high for SaaS and multiples. In some way, we never had it yet in
physical industries and I sent a note to our entire CEOs couple of weeks ago of like, "Guys, this crazy ride our company is raising 14, 15, 20 billion dollar every year will not be here forever. Be super smart and a discipline of how you use the capital. Are you delivering what we call core value drivers internally? Are you actually hitting those inflection points that have the commercialization in order to unlock the commercial sides because you never know when the market will shut down and you wants to take an environment like that to stack your bank account, but not necessary necessarily use it."
Wonderful. Thank you, Liar. We're out of time. This was a lot of fun. Really appreciate it. so much.
