Lior Susan on Why Now Is the Time to Build Industrial Giants

0:00This car is driving better than me in the oil-bound. It's much harder for us as humans to let it go. It's more of a psychology and education than actually reality because the reality is self-driving cars today driving way better than 99 % of humans. I have no doubt in that.

0:16Peter High: Welcome to TechnoVenture. I'm Peter High, president of Metastrategy. My guest today is Lior Susan. Lior is the founder and managing partner of Eclipse, a venture capital firm he founded in 2015 with a focus on transforming the physical industries that underpin the global economy. including manufacturing, supply chains, transportation, energy, healthcare infrastructure, and defense. The firm partners with entrepreneurs building technology companies that combine software, AI, and advanced hardware to modernize these essential but historically underserved sectors. Its portfolio includes companies such as Cerebrus, Bright Machines, Augury, and Ursa Major.

0:50Peter High: Before founding Eclipse, Lior led Lab9, the hardware investment platform and global manufacturing company Flex. Earlier, he helped build Intucel, a pioneer in software-defined networking that was acquired by Cisco. He also served in the Israeli military. In this conversation, we will discuss why manufacturing and other physical industries represent such a significant investment opportunity, how AI and robotics are changing the economics of the physical world, and why rebuilding industrial capacity has become central to economic competitiveness and national security. Lior, welcome to TechnoVenture.

1:24Peter High: It's great to speak with you today. Thank you, Peter. Great to be here. It's a great pleasure. Well, let's begin, if you don't mind, with Eclipse, the venture firm you founded in 2015. When you founded it, now 11 years ago, much of Silicon Valley was focused on consumer technology and enterprise software. What did you understand about physical industries that the broader venture ecosystem was missing at the time? Yeah, no, absolutely. And the goal was not to start a venture film. That actually was not the goal. The goal was to go and build companies in the physical world because that's my passion.

1:57And I think the impact of the world is the greatest. When you touch the physical, we live in the physical world. We don't live in the enterprise software world. And we saw, I saw from first hand when I was at Flex, how China did it so well and how they grow the GDP from close to zero to 20 trillion by only innovating the physical world. And I thought to myself, it's just not make sense that all of my friends trying to convince me to come and build in the fintech or consumer or enterprise software when 85 % of the world are focusing on the physical element and there is literally nobody investing in that space.

2:32Peter High: Yeah, fascinating. Yeah, I also have another question I want to ask you at the outset. Eclipse is described as a team of operators building companies that transform physical industries critical to humanity is the way that it's framed. At least I've seen it framed. Talk about this whole notion of a team of operators doing so. We'll get further into the other parts of it as we get further into our conversation. But describe that if you would. Yeah, I mean, we wanted to not stop building while we transition our life from being an operator to investor. So the first part was we're selfish and we didn't want to stop building.

3:10The second part was those companies that we built, when you build a rocket and when you build a fab or when you build a robotic solution or mine, those companies being built different than just the world of enterprise software. And we felt it's required to bring more than capital to those companies. So we assembled what I will think is the best operating team in the industry. So when we are bringing capital, we're also bringing our expertise of how to build those businesses.

3:42Peter High: What are some of the common characteristics that make a physical industry attractive for technology-driven disruption in your mind? Yeah, I mean, I think, I mean, we're seeing it. I mean, only this year alone, you saw SpaceX and Cerebrus IPOs. You see a lot of market share is being created in the physical world. And I think, you know, we had this notion that we can outsource all of these physical things to other countries. And then we learn in COVID to begin with that we cannot produce ventilators. We learn in the war with Ukraine that we cannot produce drones and missiles. We learn that we cannot build enough energy sources to power our data centers.

4:24So I think, you know, we are facing choke after choke on a physical industry's transformation. And the value of solving it is so big.

4:37Peter High: And thus, you have argued that this may be the best time in decades to build a manufacturing company in the U.S. So describe why. Absolutely. Absolutely. I mean, we call it the five forces, Peter. I unfortunately didn't invent that. China invented these five forces. and the five forces is policy, technology, capital, talent, and customer demand. And when you are aligning all of those forces in a certain way, you basically get that flywheel that we are all so familiar with, Shenzhen and other places of China. And my vision was to create the same in the U.S. And I think this is the first time in maybe post-World War II, arguably Carnegie, Henry Ford, and some people saying ever that those five forces being aligned in this country.

5:24And we are saying that we see how fast we can build those companies now.

5:27Peter High: Really fascinating. You've also said that it can cost dramatically less to launch a manufacturing company today than it did decades ago. What technological or business model changes have reduced those barriers? Yeah, no doubt. We are drinking all of the innovation that Silicon Valley is being built in the last couple of decades. Of course, we start with cloud and software. We moved now to AI, agentic solutions, computer vision. And last but not least, we reached a point, Peter, that with robotics, we can really get into scale. So when you think about it, what China did so well, they had a scale because they had a lot of people.

6:03And then they built skill all the time because they just did it again and again and again. So they had scale and scale. We can get scale today with robotic solution and automation. We can get scale with AI. And this is kind of the first time we can really head-to-head compete with them. And manufacturing that was historically sounds outrageous to even imagine you would be able to produce domestic.

6:27Peter High: Reshoring is often discussed as though relocating factories is sufficient. But what capabilities, supplier ecosystems, specialized knowledge have to be built alongside the physical production itself? Yeah, I will slightly correct the phrase. It's not even a reshore because those factors never being here before. But, you know, that kind of was how we thought about it, right? We told ourselves, oh, we give up our manufacturing. Now we are trying to reshore them back or onshore them back. When the reality, a lot of those manufacturing never being exist here. They started offshore because we give them the capital.

7:08that we give them the knowledge of how to do it. And we told ourselves, we cannot do it in a cost competitive. We cannot manufacturing electronics in a cost competitive. We cannot manufacturing industrial parts in cost competitive. And we personally at Eclipse challenge that assumption every day. And yes, are we missing suppliers' networks? Yes, but that's also an opportunity. If I solve it, naturally I can build very big business. So it's not perfect, but we're definitely seeing the five forces kicking in.

7:39Peter High: That's interesting. You mentioned that that originates the five forces in China and that there's much about it that we can replicate. What can we replicate? As you think about the models between the two, where would we be wise to choose a different pathway, perhaps relative to how China has done so? I mean, there's the obvious political, of course, and other aspects that are simply quite different between us. But beyond that, in terms of business practices? Yeah, I mean, I'm an American and not a Chinese for region. I want to live in a place of freedom, in a place of democracy. So naturally, we have the advantage of living in the lifestyle that we are living.

8:16That can be a disadvantage in the business side. Naturally, if China is making a decision to build a factory in a place, they will run over this farmland. The permits will be very easy and they don't really, really care. or no, we'll not care as we will care about pollution and about some other environmental areas. But it's fine. I actually think it's good. We compete with people that know how to do it really well. They have$20 trillion, and I want to take that$20 trillion.

8:49Peter High: Where do you see the largest opportunities in physical infrastructure supporting AI? I mean, we definitely call it internally the picks and shovels for AI. when you're thinking about it is we build so many companies that going into the data center world. And it's from things like the compute. So new chip architecture, new compute, new storage systems, new networking. We're building a lot of energy systems, batteries, solar and batteries, nuclear, different type of gas turbines, modern gas turbines. We do a lot of things around how you build those data centers, modular, prefab. There is so much to do there that I don't feel we even scratched the surface.

9:35Actually, as we speak, I'm building a new security company for data centers. There is just so much to do.

9:42Peter High: How far along are you? I know it's in stealth mode now, so there's not a lot that you can share. But again... Yeah, we assembled the team. We raised multi-hundreds of millions of dollars to go after this problem. We have some of the top AI labs and hyperscaler as a design partners. We have a lot to build, but no doubt that securing our data centers from the IT and the OT and the physical and making sure that the weights are secure is going to be a gigantic problem for us to solve. Talk a bit about how AI changes the economics of industries like manufacturing or construction or even logistics and energy, where the objective is not merely to generate information, but to control machines and improve physical output, Lior.

10:27I built, when I was at Flex, we built multiple teams focusing on our automation with the notion that we wanted to onshore manufacturing. And naturally, we don't have the same manpower that we have in China and Asia here. So we wanted to relying on automation. You were not, Peter, what took us years to program those robots, you can do today in a matter of a few hours. so i can go and build agents that running on industrial plcs that can program robots and give robots flexibilities that humans have in a matter of hours yeah um and that's just uh and of course i'm simplifying it it's very hard there's a bunch to do it will fail constantly but um ai connecting to robotics will be the largest transformation for manufacturing in the future out out.

11:23Peter High: Yeah, very interesting. Physical AI has to operate in environments where mistakes can damage equipment or even endanger people, different from software, typically at least. How does that change the requirements for things like accuracy or liability, you know, relative to conventional software from your perspective, Lior? Yeah, I mean, humans damage equipment and hurting people as well, right? I mean, I was on Bloomberg TV on Friday and talked to my friend's head over there around the whole AI safety discussion. And my point to him is like, oh, you know, AI can kill people. And it will, I guess, unfortunately.

12:04And I told him, yes, but also when we invented cars, that kills people as well. We're still using cars because it also did so many other great things for us as humans. what we need to do is set the rules and set safety mechanism like the safe belt and the airbags and regulation of what you need to have. So yeah, I have no doubt that physical AI will damage equipment and will hurt people. But my assumption is going to do it much less than what humans is doing to humans.

12:36Peter High: What is the adoption of self-driving automobiles, for example, autonomous vehicles, say about the rate with which these models could be adopted? The reason I ask this is, in some ways, it's human psychology that puts places a much higher bar to giving up control, right? And therefore, the standard also is higher in terms of what we are willing to allow the technology to do or the mistakes it might make. What are your thoughts there yeah it's uh um i had my parents uh coming here to visit us in the bay area uh maybe two years ago and i put my my dad in a waymo car uh hey it's very fun to see someone in his age just getting into self-driving he's like son where is the drivers i'm like it's going to work just fine uh and he's like hey it's going too fast i'm like that don't work uh but when we finished the experience, he say, this car is driving better than me and all your mom.

13:38And I'm like, 100 % right. But it's much harder for us as humans to let it go. It's more of a psychology and education than actually reality, because the reality is self-driving cars today driving way better than 99 % of humans. I have no doubt in that. I would say on the physical AI, though, Peter, that I think it will be slightly different. Yeah, a lot of cases in the physical AI, it's a control environment. So when you do it in a factory, you make a decision, you do it in a warehouse, you make the decision, you do it in mines, you make a decision that is a control environment. You do it on the road, it's not a control environment.

14:15And I think that's much harder for society to accept.

14:19Peter High: Yeah, good points. Very good points. Humanoid robots have attracted a lot of capital and attention. You've expressed some skepticism about the category, however. Talk a bit about why. Yeah, I mean, personally, I don't have anything against humanoids, and I hope humanoids will be successful at some point. I think it's just, I think it's very arrogant for us as Silicon Valley entrepreneurs to say, oh, you know, we should build a robot that looks like a human. I always really like to say, as a technologist, which type of a problem we can solve with technology. because arguably, let's say in the world of manufacturing, legs is a constraint.

15:01You actually don't need legs, but humans have legs, so we created the legs for humanoids. I think humanoids will be useful for certain of our application. I think there is still a long way to go on form facto weight capabilities because I don't think you will want right now humanoid in the current form facto walking in your house. I will not want next to my kids. So I just wanted to temper a little bit the hype around that more than anything else. Do I believe we'll have human noise in the end? Yes. But I think we should temper the hype.

15:34Peter High: Yeah. Yeah, yeah. Very interesting, Lior. Thank you. A lot of the founders that Eclipse has as part of it, part of the operation, part of where you invest are often from companies like Tesla, SpaceX, NVIDIA, Amazon, other leading technology organizations. And I wonder, as you think across those, what are some of the common characteristics you look for in the founders and the kinds of experiences they bring to bear for the ideas that they're, and I realize it's a general question across numerous entrepreneurs, but what are some of the common characteristics you look for? Yeah, it's interesting.

16:11A, I believe it's changed. I was saying the first eight years, our founder characterization was someone in his late 30s, beginning of 40s, works in SpaceX, Tesla, NVIDIA before, maybe first-time entrepreneur, but kind of not first-time operator and kind of have a lot of points on his belt. That is a little bit crazy and wants to solve something really big. and it's just that was kind of the mold of the people that we interact with because I think the young people historically was much more going after enterprise softwares and fintech and crypto and things like that. Now it's interesting we see now the young folks swinging and wants to solve problems that are really big in our sector.

17:01So we also adjusting and kind of opening our aperture to what is the founder fit for Eclipse. So it's changing all the time.

17:10Peter High: Yeah, very interesting. You've also suggested that industrial startups are less likely to fail because the technology doesn't work, but because go-to-market expertise is lacking. Talk a bit about that. That runs contrary, perhaps, to what the common belief is there. Yeah, I think, you know, from someone from the middle of our country, when you probably will look on what we do is like, ah, probably the rocket didn't land where you wanted and it fell on the technology. I will tell you, in 11 years as an investor and an operator and 15 years before as an entrepreneur, I never felt, we never felt on that the technology didn't work.

17:48Technology, by the way, don't work in certain times and you need to iterate till it works. It's almost 100 % of the time you fell on the go-to market and as a result of that and your ability to raise enough capital to go after this very big problem.

18:02Peter High: Very interesting. Eclipse sometimes works with experienced executives before a company formally exists. We were just talking a bit about the cybersecurity company you're mentioning, for instance, and thus helping validate the market, recruit the team, secure early customers. Talk a bit about why you've chosen to operate more like a company building partner than a conventional venture investor in that way. We talked a bit about the outside. I'd love to have you dive into that a little bit further. Yeah, I'll say again, it started with like a pure being a selfish or we didn't want to stop building while we are investing.

18:39So it's kind of started with like, should we just keep building? And I'm like, yeah, absolutely. And one thing led to another. I think it's now a much bigger scale and it's just fun. And, you know, as an operator, it's keep us fresh because we are constantly building. We are constantly in the field hiring people, working on solving problems for big customers. I just feel like my network and my skills is being fresh. It's not like, okay, I was an operator 11 years ago, but then 11 years into it, I didn't build anything. So I think the second thing that we are seeing is just keeping us fresh into the cutting edge, both on the talent and the biggest problems.

19:22I think the last one is we just get a lot of amazing people that say, we are not sure what we want to build, but we know we want to build it with you. Can we come inside the house and stay as long as we need? And it's just a very, very easy answer for us to say absolutely. So in any given time, there's probably six, seven teams that working inside Eclipse on incubating something. Not all of them will end up as a company, but a lot of them will.

19:51Peter High: Are they physically together at all? Yeah, yeah, yeah. Yeah, yeah. They are coming and sitting in our office. And the other side, we have one side is where the investment team and the operations sit. And then the other side is we are hosting different teams. And we also have a loft that we host some of those groups as well. They physically spend a lot of time with us. And do you find opportunities for them to spend time together as well, to learn from each other? Or is that less important? It's just very natural. well, we all eat lunch together and we are not, it's not a program. It's not a factory.

20:26We just let it, let those people just interact with each other. And it's fun to see that one group start talking to another group and, you know, start to collaborate and say, oh yeah, I have a friend you should meet. But we are not enforcing anything. It's just naturally happened.

20:40Peter High: You've talked a bit about the growing importance of government expertise in the work that you do. Talk a bit about that importance and why it's a characteristic you look for? I think we need to. I think historically, we as entrepreneurs in the private sectors, we did not do a good job of bringing along our governments in to help us build. And when you build in the physical world, policy and government's affair can be a huge blocker, or it can be a huge tailwind. Depends how you use it. And we hired someone, Thomas Torch, that is working in DC and helping our companies navigate these very, very complex things called our US government.

21:19But it's exciting to see, I think, especially the current administration, I will hope that it will continue a very opening to work with us on tough problems like AI, like semiconductors, like energy data centers, manufacturing. They really care and they really want to have the private sector weighed in. And we want to do a better job as a firm, for sure, but I hope also as a community of tech entrepreneurs.

21:46Peter High: As you're thinking about the role of regulations evolved across time, especially relative to artificial intelligence, given the fast pace at which it is progressing, what are your thoughts there, if you don't mind my asking? Yeah, I mean, I think there is no doubt in my mind that regulation is going to get much more scrutiny around this cutting edge. And again, you can ignore it. You can hide it. It will catch in the end because it's just really powerful. You put an AI on a robot, that robot can become really powerful. You put AI on a car, that car is now very powerful. You put AI in advanced model, now that model is really powerful.

22:27And this is where governments will step in and regulate or put some guardrails. And my preference is to help them doing that rather than being told what to do. Yeah, very interesting.

22:39Peter High: Manufacturing capacity, energy infrastructure, supply chains, like a lot of this has national security implications. And I wonder, has industrial investing now become more intertwined with geopolitics from your perspective? Yeah, I mean, I think I remember that like a one of the forces is naturally a policy and it's for and it's for a reason. We see it as an important as capital, talent or customer demand. And it's really interesting, Peter, that we are seeing now companies being built with a policy in mind. And it's something that a few years ago would be unheard. And you see those very, very early stage companies thinking about the policy, hiring GA and policy team out of the gate, interact much more often with Washington.

23:29So it's really encouraging.

23:31Peter High: Very interesting. I'd love to have you maybe share some examples to personify some of what we're talking about as well, the portfolio companies of yours that are out of stealth mode that particularly excite you. But offer a few examples if you would. Yeah, I mean, I think I work really closely at one of our companies outside of Boston called Vulcan Forms. The company is manufacturing a precision metal parts. and those metal parts goes into a variety of products a variety from consumer electronics into medical implants and artificial needs to data centers and of course military parts that will go into drones hypersonic missiles etc when we started the company in late 2017 US manufacturing close to zero out of those titanium parts It was predominantly close to 100 % in China.

24:27And our review was not only we need that, because we need to have supply chain security so we can control those very strategic parts, we also want to do it in a technology matter. And the company is designing these very high-speed lasers that allow them to manufacturing those parts with very little people involved in the process. And we're partnering with Professor John Hartz out of MIT in order to create that company. And yeah, it's exciting. We are building now three new factories as we speak outside of Boston in Devons. We are going to be the first at scale, vertically integrated company from powder to manufacturing the parts to post-processing in the US before end of next year.

25:13At scale, meaning we'll be able to do billions of dollars of revenue. So just exciting times.

25:21Peter High: Amazing. Very, very good. I also wanted to ask you before I let you go, what do you read or listen to in order to stay current as somebody who is investing in new ideas that are by definition innovative and therefore not representative of how businesses is being done today, but how it might be done in the future and therefore needs to you need to look around corners, be aware of trends as they evolve and so on. who do you read or who do you listen to that you would recommend to others? Yeah, my answer might slightly surprise you. I actually get most of the intel that I need of just being in the field building and meeting people.

26:08And you usually get the real insight not by reading it, is actually being in the field and seeing it from first end or hear it from first end. And I do have, I create in Codex, this agent that are doing certain of things for me in order to gather my information so I can every morning read that information. So yes, agentic capabilities of pulling data is very, very useful. But I will tell you the best source of information that I have is just to be in the field and build companies.

26:39Peter High: Fascinating. Well, Lior, Suzanne, thank you so much for taking time with me today. a fascinating conversation, emblematic of where you are investing, how that's evolving. Thank you so much for sharing your perspectives. It's been a great conversation. Peter, great to be here and I appreciate your time. Excellent. Thanks for listening to TechnoVenture. If you enjoyed this conversation, subscribe wherever you get your podcasts and leave a rating. Thank you and I look forward to seeing you next time.