Designing the Impossible: AI, Quantum Mechanics, and the Future of Materials | The Pair Program Ep95

Welcome to The Pair Program from Hatchpad, the podcast that gives you a front row seat to candid conversations with tech leaders from the startup world. I'm your host, Tim Winkler, the creator of Hatchpad. Join us each episode as we bring together two guests to dissect topics at the intersection of technology, startups, and career growth. Welcome to the Para Program. Uh, your host Tim Winkler, uh, accompanied by my co-host Sean Leahy. Uh, Sean, I got, I got one for you, man. Um, a buddy of mine told me he walked into a Lockheed Martin building last week and, uh, got scanned by robot dogs before he could, uh, before he could get through the doors. Robot dogs? Yeah, like, you know, like those Boston Dynamics, um, robot dogs you oversee at like a like a defense event or something? Well, man, whoever has the special purpose officer contract to do all the security at front desks is probably shaking in their boots now if they're going to lose that to a bunch of robot dogs. I mean, I'm not messing with a robot dog, dude. Um, yeah, we got robot dogs here. Do you? Yes, we do. Do you have to show like ID and they scan it or something, or— walks around, sniff your, sniff your crotch? He's not doing full on LiDAR scans when you walk in the door yet, but, uh, that's not a bad idea. Might actually push into that. Yeah. Well, it got me kind of thinking, right? We're entering this next phase where humans and robots are just kind of coexisting. And every day is like a new article. Like I saw like a barista bots out in Portland. I went and got like a slice of pizza the other day and like a little bot came and dropped it off, dropped the pizza off my table. But if you had to pick one, just kind of curious, What's one service industry role that you'd be, you know, totally fine seeing replaced by a bot? I've been, I've been waiting for this for, for decades at this point. I want robot umpires in professional baseball. Okay. All right. I want— and, and we're getting close. So if it's where we just started, we're recording— I think they just started— we're recording after opening day and, um, they just started allowing, uh, balls and strikes to be What was that, Joseph? Must be a Nats fan or something if he wants that. Oh yeah, absolutely. Yeah, that's right. I want, I want perfect balls and strikes calls. It'll make the game better. You know, guys will, we just need an excuse, right? We need an excuse to find out. I, I, I, uh, I think it's inevitable, dude. I, I, you know, we saw it in tennis, dude. Tennis started that, right? And then I think they, I don't know if they fully replaced like the umpires in tennis or not, but they definitely went to like the— you know, like the, uh, the box, the reviews. Yeah, yeah, we'll see. I mean, I think this is, this is that first trial year for it in baseball. I'm going with like a, like a night shift hotel, uh, front desk rep, uh, because like I've been at the tail end of like a 10-hour road trip and the last thing, you know, I'm grumpy, they're grumpy, they're just kind of like tired staring at their screen, low energy and I just cool like a bot, just coming up and scan my ID, dispense my key at that point. Yeah, I mean, again, we're getting close. You can— on— there's a Hilton app. Uh, I'm Hilton Honors member, so pretty cool. Um, but on the app you can check in remotely, get a key code, uh, to the doors, uh, onto your— into your actual like hotel room. Never have to have any interaction at all. Yeah. Um, so So we're getting closer and closer. Yeah. Yeah, man. Well, I'll use that as a little segue into today's episode. So, you know, it's wild. We're seeing robots showing up not just in service jobs anymore. They're starting to show up in places that you wouldn't really expect, like running some experiments, operating in labs, helping discover entirely new materials. And it's a big shift. You know, for decades, innovation has always been driven by software., and now you're starting to see AI, you know, moving beyond code and into the physical world. It's exciting, it's scary, but you know, we're here for it. We're here to talk about, you know, some of those changes. And we're today, we're going to be talking about what it looks like to rebuild the engine of scientific discovery using AI and automation to dramatically kind of speed up how new materials are created. Um, and a lot of the big challenges that we talk about on this podcast specifically, you know, those, the challenges that we face in energy, defense, you know, semiconductors, a lot of those aren't, they're not software problems, they're also materials problems, right? And so, uh, excited to, to jump into this one. Uh, today we got two guests that kind of approach this from a few different perspectives. Both, both are complimentary to one another. Uh, first off, we've got Joseph Kraus, um, co-founder, CEO of Radical AI. A New York City-based autonomous material science discovery firm, uh, building a fully integrated AI-driven system for discovering and creating new materials, including a, a self-driving lab running experiments at a pace traditional R&D can't match. So excited to, to get deeper into that. Joseph, uh, thanks for joining us. And then alongside— yes, sir. And alongside Joseph, we got Nathan Docter, uh, founder and managing partner at Infinite Capital, uh, an investment firm that backs some of the most, uh, ambitious deep tech companies across AI, aerospace, nuclear, and decentralized systems. And I'll note that Infinite is an investor in Radical AI. So Nathan, thank you for joining us as well, man. Great to be here. Excited for this. Yes, sir. All right, well, before we dive in, we do a little tradition here on the podcast called Pair Me Up. Go around the room, rattle off two things that just kind of go well together. Could be anything. Sean, why don't you lead us off? That's a little syrupy, but you know, I got to stick with the theme. My pair today is peanuts and Cracker Jacks. Baseball's back. I couldn't be happier. Yes, sir. Yeah, I like that. You going to any games or what's— we'll see what the schedule looks like. I've learned my lesson enough that going to the really early season games, March and early April, you can get hit with a cold snap that makes sitting down for 9 innings potentially a little bit difficult if you're not wearing the right attire. So I might be looking more around the May timeframe to get down to Nats Park or maybe out to Baltimore, see an O's game. Yeah, cool. Yeah, I've never been to Camden. I've heard it's like, it's a sense. Amazing ballpark. It's amazing ballpark. To like go and not been and be so close. Yeah, peanuts and Cracker Jacks. It's a classic. I'm going to kind of stick with a similar kind of sports theme. You know, March Madness is upon us, so my pairing is going to be, you know, perfect brackets and then just getting completely wrecked by day 2. And you know who you are out there. You're the ones, you know, just like me that had Wisco over High Point. You know, I didn't even know what High Point was. I had to look it up afterwards. I come from a school, George Mason, which nobody knew about until we made a pretty epic Final Four run way back when. But yeah, you know, I think everybody had a good bracket, get a little bit destroyed this go around. Yeah. NC over VCU, Gonzaga, or no, VCU over NC, Gonzaga over Texas. So yeah, that's kind of the theme for me. It's that the love and the pain of March Madness is what's so beautiful. So perfect brackets and then getting completely wrecked by day 2 is going to be my pairing. Let's pass it around. Joseph, quick intro and your pairing. Yeah, again, thanks for having me, guys. I'm Joseph Krause, one of the co-founders and the CEO of Radical AI. Pairing for today, I'm just getting back from a trip to DC and the Hill. And so my pairing is material science. And defeating our adversaries like China in innovation. You cannot do one without the other. And I think it's starting to become understood, discussed, and really a priority in DC to really focus on advanced scientific discovery, materials being one of those fields that are important among others like bio, where there's serious investment and focus coming into so that we will win the future conflicts that we move into and be the most technologically innovative country in the world. So those two are going together well. Nice, man. Yeah, it's super timely. It's, it's why you guys, it's why we have, have you guys on this, on the podcast. It's a, it's an area that we focus in quite a bit. I'm curious, have you heard of a, a little community down there called DC Station? Or I'm sorry, Station, Station DC? I have not. No. I have to loop you in because it's basically, you know, just what you kind of discussed. We, we, come together. I mean, they've got a whole kind of like, you know, co-working clubhouse, if you will, right down in DC. But it's all about bringing investors, you know, founders and policymakers together in one room to solve those hard problems. So it's super relevant for you. I'll pass you the info. Yeah, that'd be great. I think I just got back from the Hill and Valley conference and shout out to the team over there and Delian and his crew for throwing that off. But there's been a lot of focus on this intersection of technology and politics. I think obviously technology is the most important driver in the world right now. And so naturally you're going to see a push into how it's defining policy and then, and then where like focus is forming an investment perspective from a GDP, excuse me, perspective. And even like a workforce development and industry perspective. What is the government helping to fund or step in or put support programs out for where that's all the way at the R&D side. So kind of like fundamental research and where they're pushing grants to academia or their national labs all the way through to securing supply chains, critical minerals like we talked about, I'm sure we'll talk about today. Government investing heavily in onshoring critical minerals, manufacturing capability and production capability all here in the US or close to onshoring in our allies and partners. And so that's very important, I think, in the theme today, and that's been a big intersection people are focused on. Nice, man. Yeah, we're catching you hot off of that conference. That's perfect timing because you're fired up to talk more about it. And Nathan, quick intro and your pairing. Yeah, my name is Nathan Docter, founder, managing partner at Infinite Capital. We focus on investing in generational innovation, and that's really been across the hard sciences, frontier tech. Started out in the entrepreneurial world, started a couple tech companies focused around software developers. I got two pairings, one on the food theme. I have to say sushi and a Sapporo. I do a lot of spearfishing myself. I was just in Japan. Strong. A few months ago. Nothing like fresh fish and an ice cold beer. So that's strong. Oh yeah. And then on the other side, the pairing I've been really excited about the last year and going into the year forward is atoms and AI. I think the physical AI sector is so interesting. It's an area where you can create real structural moats in the AI world. And you start to really pull in data from the physical realm that you don't have access to elsewhere. So I think that's about to transform a whole list of sectors from manufacturing to materials to robotics. Exactly. I like that, man. Adam's and AI, solid. Almost as solid as a Sapporo and sushi though, man. I, I'll tell you what, like I can't go to a sushi spot and not order a Sapporo. It's like I, I, I'm on, I'm programmed to, to, to drink a Sapporo with when I'm eating sushi. My judge for how good of a sushi spot it is is how cold the Sapporo is. You know, ice cold temperature. That's a great sign. Yeah, good stuff. All right. Well, yeah, man, I appreciate you guys both joining us and good pairings all around. Let's, let's get into it. So just a quick reset, you know, for, for the listeners here where we're going to go with this conversation. Well, let's start with what it actually means to kind of almost rebuild the scientific method in a sense. And then we'll talk about how, you know, AI is being used to design entirely new materials. And then I want to dig into this self-driving lab. I've heard a little bit about this happening in a couple other research areas in universities, but what this all means long-term from both a builder and an investor perspective. And Joseph, why don't we kick things off with you with Radical? So doing some research on your background and what you guys are— problems you're solving, And you said you're not just building a better lab, you're rebuilding the scientific method. And so for a lot of people, right, science still kind of looks like hypothesis, experiment, analyze, repeat. What do you think is kind of broken about that model today and what is it that you're trying to do there? Yeah, that's a really good question. And that gets at the essence of why it's so important to build the technology that we're building. The way I always like to describe it is you just think about how many important problems we have in the world, right? And pick your industry that you care about, whether it's automotive and aerospace, if you're in the manufacturing or automotive, or like we in semiconductors and energy generation and electronics, the most important industries in the world all come back to a core thing, material science discovery. And as you correctly said, Tim, when you think about the scientific process, it is a really broken process. It is very long timeline because it's fragmented. It's incredibly expensive to run from lab discovery to something we actually use in all of the products that we interface with today. And the different, the parties who do that are all different. Again, it's like different people do discovery, different people do testing, and different people actually use those end materials in real manufacturing and products. And so this ability to actually think about linking all these together is fundamental to the way the world works. And in our opinion, obviously our biased opinion, one of the more important technologies humans will ever build. And so when you think about the scientific process, how do you start to change that? How do you change exactly how I just described it? Well, you, what you do is you focus on going from a serial-based process as you described. I make a hypothesis, I run some computations or simulations, then I make it in a lab, Then I look at the results of what I made and I think about it, and then I go back and I do that loop over again and I go step by step and each step is preceded and required the former step to actually, you know, kick off. And then you think about our technology and building a fully robotic self-driving lab, and it is the opposite of that. It is a parallel-based process where all of those steps are continuously going on at the same time. So we are making new hypotheses, running different simulations and computational workflows, experimenting with new materials, and analyzing the data coming out of that experimentation all simultaneously. And we are pulling the information out of all of that and learning with it. You know, that active learning loop that's actually training our AI scientists to do science in this more effective way. And, and why is that important? Well, it's quite simple. Right now you have 10 scientists working on one problem, and that moves us to a world where you have one scientist working on 10 problems at a time. That is some of the more impactful technology that I think we'll ever build, and that's why we get very excited about the mission of the company and the technology that we're going after. Yeah, it's really neat. And you know, this is kind of bringing me back into, you know, my little bit of 8th grade science class where I was just kind of Reminiscing on, you know, some of the experiments that we did, one of the things that stood out to me with this process that you're describing is how much wasted effort exists. And like, so failed experiments, right? So failed experiments don't really get shared. So you've got labs all over the world kind of repeating, you know, some of the same work. What changes when every experiment, even like the failures, become a part of a system that actually learns from it? This is a really good question, and this is the, I think why, why Nathan is spot on with his theme. Like these are the areas that the physical world or AI and atoms as he called it, are going to be deeply impactful. Because number one, there are no datasets. As you just said, there's no successful experimental datasets except what really exists in scientific literature, and that's a small cohort. But the unsuccessful experiment dataset, that would be one of the largest datasets from a science perspective in the world. And when we don't even capture it, it's, it's, it's all up here in the brain, or, or as you said, is unshared. You know, I talk to people who work on similar material systems that we do, and we've done the same things in the lab that we could have learned from and innovated on and built on top of. And so I think this idea of capturing To put it very simply, the data from real-world processes, kind of what Nathan was alluding to, there's an immense moat that will form for the people who can do that successfully. And that's very hard to capture, right? This is, you know, the one difference with software is that you can run unit tests very fast in software, right? Like seconds or minutes if it's a larger test. In materials, weeks. Months, years to do experimental tests. Very hard to get the size that can start to be truly impactful in the domain. And so from that, I think, you know, why is it so exciting is this messy data. We call it messy data. The kind of the negative results is what we typically refer to it as. It's actually where all the information is, right? You think about the first time, you know, you make your favorite recipe to use the simplest of examples. The 10th time is better. Why? Well, because you like throw the salt in at a very specific time, right? Or you're like, you flip it in the pan like right when it gets that charcoal on the outside that you, that you love on your steak or whatever, like whatever it is, you just, you have these little things that you learn through repeatedly doing your process that inform and make your next process way better and the end product way better. That messy data, like you can't really predict that data. There's no way to understand and guess what that data's going to be. And actually underneath it is where the most interesting areas of the world are. And this is amplified when you go to science. Yeah. If you're making a salad or cooking a steak, there's only so many messy things that can happen that will make a big difference, right? Adding salt at a different time or some, something like that. Changing the temperature range. In atoms, in the world of science, that is the entirety of scientific discovery. Our understanding of the universe is based on exactly that concept. How can we unearth and understand things that before we did it, we did not understand or knew existed? That is the scientific discovery process. So the amount of variables, you know, salt, temperature that exist in going into that messy data is enormous. We are most excited about that dataset, and that is the exact dataset that we build with the company. And I think to Nathan's point in the beginning, That's why everyone is excited about this area, and I'm sure you can speak to it. That data isn't captured today, and people are capturing it for the first time in very, very important industries. When you smash that into artificial intelligence, there is a serious output from that. Nice. And that's what we're seeing too, like across the AI sector is at the end of the day, the fundamental approach to AI now is entirely dependent on the data you've accessed. Yeah. And there's so many areas where those datasets just don't exist. You look at kind of pre-radical, pre-let's pull this into the real physical world. You have the Materials Project, you have all of these approaches to how do you generative AI what could potentially exist and the vast majority aren't real, right? Like you really have to bring it into the physical world to test those understandings of like, where are the boundaries of what our systems understand and what they don't. And that's the other interesting thing about the way Joseph's talking about how you upend the scientific process. Humans are— it's a very hard problem to take an enormous dataset and find idiosyncrasies that can result in groundbreaking breakthroughs, right? Like you look at the people that have punctuated that throughout history and it doesn't happen very often. And that kind of scattershot approach to scientific innovation has gotten us to where we are today. But once you can start to parallelize that process, like the amount of breakthrough takeaways that start to come through on a monthly, even weekly basis can, you know, that compounds in a flywheel that it's impossible to even imagine what that looks like looking backwards. Mm-hmm. Quick shout out to our sponsor, Defense Unicorns. This one's for the problem solvers out there. They're hosting Warhacker, a first-of-its-kind hackathon built for the defense community. No buzzwords, no slide decks, just hands on keyboard solving real mission problems with real code. You'll be side by side with developers, engineers, and innovators from across government, industry, nonprofits, and academia, all hacking for the warfighter. It's happening June 16th through the 19th in San Diego. Got a real-world problem to solve or want to join a team that does? Learn more at defenseunicorns.com/hatchit. So, Joseph, something that you've hit on here with your messy data, I like to think of that as tacit knowledge., right? So knowledge that's hard to document and to describe, but something that a practitioner who has a lot of reps can easily identify. So we're just sitting on my baseball theme, you know, MLB pitcher can tell me how to throw a pitch all day long, but I can't throw 95 miles an hour off the plate. But just the feel in his fingers, how, you know, how he uses his legs and everything, that's that tacit knowledge., and you're, you're automating that, or it sounds like you're automating that, uh, uh, or that's part of the, the, the goal. Um, how do you capture that though? When, so you, you have this messy data that you're generating, which I agree is an incredibly valuable dataset, but when there is a real insight that you've discovered, how are you, uh, what's your approach to capturing that so you can leverage it? Uh, great question. So we call that, uh, scientific intuition at Radical AI. It's like the scientist version of what you just mentioned with the picture. And what that intuition is, is experience, right? I've been a scientist for 15 years. In that 15 years, I've done these things. And so I have ideas, I have intuitions about why something is going to happen the way it happens. I might even have learnings or like real experience on watching those things happen. And so we've actually built a platform where we can capture those intuitions and we have this backend infrastructure that we've built that we can go through and I could talk to you about what that means. But there's a really core piece of this where we have annotation. And while, while our AI is analyzing the results of our experiments, which it does do today, our scientists do the same thing. And so they'll go in and look at an image and they'll make an analysis on what they think they see. And the AI will do the same thing simultaneously, and then we'll compare the two. And that delta, that difference, there is the intuition. Why is the human scientist saying this, Why is the AI scientist saying that? What is the difference between them? How do we rectify that? How do we bring those until they're one? That is exactly what we do as a process today. And so this is why it's really important to think about running fast discovery. You need to do that closing of the gap a lot of times. Like, like you, the more times you can do that, the more intuition you can capture., and therefore the smarter your AI scientist becomes. And so again, I think Nathan used this, this word that we love to use, which is flywheel. We, we deeply believe it's a flywheel. It literally feeds itself, and that is the way it should be because now you start to compound knowledge instead of disperse knowledge, which typically science today. And so that's scientific intuition. That's a core part of our approach and something that We work really hard to capture. The second answer is actually new techniques that allow you to capture it as well. I'll give you a perfect example. And Nathan's seen this live. When we make our materials, we work in a field called structural metals. We actually have to blast these things with a super hot torch, plasma torch to melt them. Like we, we take these metals and we make 'em a liquid and then we use that to make a part. Well, when you're doing that, like human scientists have really good intuition on the right way to like melt them down. Like, I'm going to go to this side first and I'm going to like scrape the wall with the torch because I want to make sure I'm melting all the exterior alloys that are inside or elements that are inside the alloy. There's like this whole process. So how do you quantify that? That's really hard to do. Actually, we asked and the scientist is like, I don't know what to tell you. I, I, I'm like, I'm looking at it and when I see that an element's not melting, I, I'll go blast it like with more power to melt it. I don't, there's no way for me to tell you the subscription process on how to do that, but computer vision can learn that. And so we just filmed it. We filmed 10,000 iterations of it, and then we trained the computer vision model on those iterations. And, and just like LLMs, They don't actually learn grammar in English, right? They don't need to. They just learn how we actually write and from that inherit language automatically, inherit, excuse me, grammar automatically. Our model did the same thing and watched what our material scientists did when it was melting down this pool of metal. It associated what the end state was with what the path it was, what the path the scientists took to get there. And now it can replicate that process and ironically can do that process better than a human scientist. It will not waste, you know, 5 passes at the top when it knows it just needs to go to one in the bottom for the end state. And, and we're working on that technology. We're continuing to develop, but that's an exact example where you have this messy data, you have this intuition. There is no language. There's no format to put it in to just give it to someone, to give it to a model. And so you gotta build technology that can actually interpret this information and then use that to its benefit. And that's live and running today. That informs all of our experiments and is an active part of our flywheel. Our process is a perfect example of, you know, trying to capture the way you put your fingers on a baseball and, and, and why when you throw it that way, you know, you get a certain direction of, of movement. Yeah. I want to dive a little deeper into this lab here shortly, but I want to jump real quick to you, Nathan. You know, from the investor lens, you know, you've seen, you know, big platform shifts like cloud and crypto and AI. Does this feel like when you hear, you know, when you see, you know, a company like Radical popping up, does this feel like a similar inflection point but just more applied to the physical world? Yeah, great question. The thing we're so excited about here is if you step back And look at the longer historical version. Materials are fundamental to scientific progress, right? You look like we actually characterize human eras according to the materials they've mastered. We had the Iron Age, moved into the industrial era with steel, the mastery over these. And we're arguably now in that kind of semiconductor era, right? The Nvidias, the TSMCs of the world have mastered how do you take silicon and do phenomenal things with it. And I think that these next inflection points are fundamentally materials problems. Like we spend a lot of time thinking about hard science theses. And where those go on a decade-long timeframe. Deep dived quantum through nuclear fusion. And fundamentally, some of the most ambitious projects there are held back by materials innovation. And so we kind of built the intuition around it without ever really deeply thinking about the material science sector. We just ended up bumping up against it a number of times of like, okay, this is where the fundamental physics starts to break down if you want to make this breakthrough. And so each of those ecosystems, I think, are underpinned by materials innovation that just hasn't happened. We did a phenomenal job of that in the US through the '40s into the '60s. There was a lot of effort, a lot of energy in this kind of sequential process that Joseph talks about. And we now have these tools to parallelize processes and expand the scope in multiples, magnitude, orders of magnitude. And that's really kind of this era we start we think is happening right now is we're moving back to hardware in force, right? Like everyone is starting to realize the built world matters more than we gave it credit for the last 20 years in venture. I think if you look at kind of the venture ecosystem from the '90s into the mid-late 2000s, it was really entirely focused on software. How do you have low CapEx? How do you get high leverage? Moving away from how do we just do just-in-time production? How do we not have to deal with building real infrastructure. And I think in mass, we're starting to realize, and this is kind of a core thesis of ours, is that the AI sector will actually be the most interesting and compelling, not in the format it's gotten to here, but in the way that it affects the physical world. And so that's where we think materials is really going to create a fundamental inflection point. And it unlocks a number of, like you talk about these large thematic industries, there's a whole number of those categories that will be unlocked through materials innovation. This is an incredibly important point. This is why Infinite was someone we wanted to work with so aggressively. And I'm going to take a swipe at some VCs here. I was a VC for 3 years, so I can before everyone gets all up in arms about it. Most VCs are shallow in their approach. They just tap onto hype trends and When they do that, yeah, you might get lucky when you jump on a trend, but it's actually about first principle analysis. And I think what Nathan just went through is exactly this. I talked to someone, to an investor. I'm like, where are you based? Silicon Valley. Oh cool. We're working on materials. Oh, are materials important? Is that the right area to go? That's funny. Didn't you just say you live in Silicon Valley? Silicon Valley. I mean, literally your area is named for a material. And It's this funny, this disconnect, and you don't need to be a rocket scientist to take, you know, some time to do a first principle analysis and understand where does this innovation come from. So in semiconductors, like why was Moore's Law a thing and why have we reduced node sizes and what is the output of that being and where has that come from? Oh, there's a lot of advancement in silicon and the different playing we do with semiconductor fabrication that allow these new capabilities to come from that. And I think this physical world thesis is very, very important. I think you're seeing a lot of activity in areas where they're shallow at their surface. And I, I do think there will be winners there, by the way, because the physical world is so disruptible from technology because it's been so long since it's been disrupted that there will be low-hanging fruit picked off. Absolutely. That will definitely be the case. But if you want to build a category-defining company, if you want to build a Bigger than that, human altering company. Like you need to do these first principle analyses on where is there a huge opportunity to revolutionize the way that we typically do something. And I think you see the biggest companies in the world focused on this right now. AlphaFold 2 in science was, it was done by Google DeepMind, was a really important moment because they took this perspective. They're like, yes, like AI is really impactful and we've built it for games and We're not going to solve an easy problem. We're going to solve protein folding, which like has been elusive for the industry for 50 years. We've needed massive GP or compute centers to do this. They did it in like months with GPUs because they started bringing experimental data in and actually use these machine learning techniques to solve the problem. That's exactly the kind of first principle analysis that you really have. I probably have pitched, I don't know, hundreds, maybe thousands of investors since I've started company. I don't know the exact number. And there are very few investors who after the whole pitch look me straight in the face and go, right, but like, what about Mars? What, what are you thinking about materials for that? Which is exactly what Nathan asked me when I pitched him kind of some of the problems we're working on today. What about the future of semiconductors and silicon?, you know, RTAP superconductors. What about things like that? You know, those are, you're talking about, again, human altering technology. These are technologies that change the way humanity moves forward today. The same way OpenAI and Anthropic and, and XAI, SpaceX and Tesla, they change the way humans are pushing forward and experiencing the world today. That's the company that you should build in physical AI, in my personal opinion. I think You need to find people. If you're on the entrepreneurship side, you need to find people who understand that, who do their own analysis, and they come to the conclusion that I'm interested in your space, not because it's sexy, not because it sounds cool, not because there's a lot of investment moving into it. I'm interested in your space because I've done a first principle analysis and netted that if you don't succeed, if someone doesn't build your company, the world will not move forward in this respect. Therefore, I want to back the company that's going to succeed in doing that. Really important perspective if you're going to play in AI and atoms. Yeah. Yeah. And you've got a pretty sexy list of VCs backing you as well, Joseph. And so, you know, Nathan, you're in good company. And Infinite was something that I also was really intrigued in. I went down a little bit more of that rabbit hole. And while we're on the topic of, yeah, let's call it, you know, investor value. Talk to— tell us a little bit about Infinite Labs, because I think there's something there that is really important to note when you're thinking through what's going to help this founder beyond the check, right? And so tell me a little bit about Infinite Labs and then that hands-on layer to support the portfolio. Absolutely. So I fundamentally believe that a lot of the best VCs were entrepreneurs to start. Like, you understand the journey, you understand the trials and tribulations. There's a lot more empathy for that founder journey, which is a very, very difficult one. Right. And coming from that background, we are inherently builders at heart. And so we put together Infinite Labs to support our companies kind of across the range of design product wherever we possibly can. We like to build on top of companies to the degree that it's useful and we can help create ecosystem. And back to kind of Joseph's point on that synergy, like the most interesting companies And founders we end up running into are ones that open our worldview. Like we think a lot about these theses deeply. We like to study them at that first principles level. But every now and then, and it's pretty rare, you meet a founder like Joseph that opens up this whole sector and is categorically thinking about it different. It's this very contrarian approach to something that largely, you know, across the industry, across the entire United States has been forgotten, right? Like the government stopped funding this to a large degree. Like we relegated material science to a number of large companies that had just very deep problems in a very, very specific vertical, right? You look at the Raytheons of the world, SpaceXes that have a material problem. They work on that themselves. But very few people have thought about how do you start at that base level and redefine the way we do that from the ground up. And that's become a very big theme of ours as well, is like the founders that kind of open that view. And in some degrees too, we actually really like when a founder contradicts our priors, right? Like something that we may have learned historically through the venture space of, oh, this sector is really hard to do, or there is failures here, et cetera. Like it actually ends up being a structural advantage as an investor because you now have the crowd that decides to like move away from that altogether. And that happened in mass with when you look at hardware and just the physical world and venture, et cetera, like so much just focused on the software side. And even today, right, you see that with AI, right? So much is focused inherently just on that kind of software layer of LLMs and you have these kind of memetic companies that grow around it. And that ends up leaving this tranche of individuals and companies that are doing phenomenal things that you get to identify and you actually almost, you meet them and you feel like everyone else is missing something. And back to the conversation me and Joseph had when we were first on site with them, we spent an hour going deep down the rabbit hole of how wild this can get. And that's where it gets exciting. You start to think about What kind of material sciences advances do you need to colonize Mars? What types of material sciences or what would happen in the world if we get to our tap superconductors? Like that's too big of a problem to attack right away, but it's the kind of thing you layer into and you end up in a space as a company that no one would ever have imagined, right? You look back to NVIDIA, right? Like when they first started out, there were so many people competing in that sector and they just kept compounding and making bigger and bigger decisions and bets off of that in a way that no one else really In some ways you could argue Jensen is a big reason that AI exists at all, right? Like huge bet on the sector and the stock market hated him for it, right? Like he actually had to do it with all of this pushback from Wall Street analysts, et cetera. And that's, I think, where the most interesting things happen, right? It's like small number of people that decide to take humanity's priors and throw them out the window and start from the ground up from a first principles based approach. We've seen that happen in spades with Radical. Yeah. Just false idea. If you haven't heard Jensen on, I think it was Joe Rogan. I think it was his, it was his interview with Rogan. It was an awesome story. Yeah. About the guy that almost didn't come through with the check, where we would be without that, you know? Without one. And then they, and then I think, I forget the exact number. He says it in the podcast so people can find it. Couple billion, he tapes out a new chip. And no one wants it. Like there's just no interest in the community to actually use this new, this new GPU. The first, I think it was the first DGX, DGX-1 or something like that. And then Elon tells him, I have a nonprofit that like could use it. It'd actually be a great workflow. And then when he's telling the story, he is like, are you kidding me? Like I literally just spent billions on this and, and I'm all for helping out the nonprofit, but like that's not going to be the buyer. Lo and behold, that nonprofit is of course OpenAI, which totally kickstarts more investment in this machine learning or deep learning ecosystem. But that, that's such a good story. And I think one that's more present is robotics. Yeah, everyone's talking about robotics and AI now. 2 years ago, 3 years ago, pretty forgotten field. They're hard. It's hardware. Robotic as a service was supposed to be this interim business model because SaaS VCs wanted predictable revenue for robot companies, and that's not the way they work. And it was like a terrible thing for the industry in my personal opinion. And then there were some people who took this contrarian approach that Nathan described and said, well, what if you take these technologies and apply 'em to robotics though? And Nvidia was paying attention and the robotics— Amazon was paying attention. And now you have serious private companies that are building wild technology on the robotics side, from humanoids to pick and place all the way through to like full automation spectrum that you see like AWS and other companies investing in, Nvidia investing in. Now obvious, you know, I, again, 3 years ago, hard, no business model, scary. Are you going to make money? And arguably going to be one of the largest fields in the world today with where we're moving towards an AI and autonomous-based future. I think it's that type of conviction that you have to have where you just view an industry differently than the rest of the world use the industry today, that is the alpha. Like that is the entire point of going to do it. If everyone already viewed the industry that way, why, why would you go do that? If everyone was already building a GPU to accelerate ML workflows, why would Nvidia have so much success with that? They might have had success, but it was that no one saw that and they saw that and took a bet on that. And then from that, catalyzed the most important technological revolution in human history. That's why they're one of the most valuable companies, if not the most valuable company in the world today. They've earned that right and they deserve that right. Yeah, that interview on Rogan is fantastic. If you haven't listened to it, check it out. Sean, did you have a— I think I saw you had a question popping up. Yeah, there's— I mean, I completely agree with what both of you have said about basically the American innovation R&D system being fundamentally broken. Over the past several decades. We had this amazing progress after World War II, but then there seemed to be sort of a stagnation by about the '70s and certainly through the '80s and '90s into today. One reason though that R&D public funding, government funding for R&D is always large is because of this kind of idea that, oh, there are some things that don't have a market, that aren't going to generate a return, but we still have to fund them, fundamental science and research because Eventually down the road, we'll have a happy accident and we'll create the next NVIDIA, we'll create the next ARPANET or something like that. It seems to me that both of you are kind of on the opposite side of that where you think that no, you can do pretty fundamental research and knowledge discovery, scientific discovery, and also make a really effective business case for it, right? To where you can generate capital appreciation, you can generate real cash flow. What gives you that conviction? I think, Nathan, this is probably a question first for you as an investor. But what's the signal there that, that makes you think, hey, you can actually make this into a profitable enterprise and also hit that scientific discovery piece as well? Yeah, great question. I think there's kind of two parts to that, right? Any really large problem, you need to step function, right? You need to think about who's the initial user that will pay for this thing out the gate. And fortunately, there are a number of users, like especially in structured metals, you look at high entropy alloys, There's a list of companies there that really need those advancements. They want them, they're willing to pay for them. There's a clear pathway to revenue there. And then on the flip side, and that becomes like that initial revenue generation arm, right, is how you get to reinvest in R&D and then again, step function to larger and larger, more ambitious and more fundamentally expensive endeavors. Right. And I also think there's currently a very big push within the government that's flipping that script, right? We're rethinking about how we do R&D at a large scale. And the government is starting to really step back in as a willing partner to underwrite some of these advances. And that's extremely meaningful. Like that helps a lot of these companies out, right? You look at Radical, they were picked as part of the Genesis mission, which is an enormous feat, right? They're in the company of giants there. Almost everybody on that list is over $100 billion. And, you know, Radical, you know, punching above their weight in a very big way and showing the world like, okay, we can, make these hard science advancements in a way that no one else is doing. And so I think we're at kind of a very special moment in time where you see that government side coming in to support. You have a lot of the commercial players who are also really high demand signal, right? Like really need to break through these material science boundaries for the sake of their own business, right? You look at SpaceX, reentry and reentry level reusability on the material side is such a large problem. Like they think a lot about that. And so you see this across industries. And that was also like, again, back to one of our core, you know, realizations there. You look at the quantum world, you look at the fusion world, you look at the amount of money that's going into those sectors. And yet very little amounts being spent on how do we actually make those materials breakthroughs that will allow this unlock, right? And there's enough capital generation and enough like focus from large corporate strategics, etc. That are really funding these endeavors that I think we're at this moment in time where the ecosystem realizes how important this is. And that's really what gives Radical the potential and the optionality to one, have capital access to undertake these pretty ambitious endeavors. And two, that flywheel just keeps compounding. You have to do it that way. Like to pick one of the greatest, in my opinion, one of the greatest companies in history, SpaceX. And what's the mission? The mission is Mars. It's always been Mars from day one. Why do they sell internet service then? Well, one, it's an incredible business as we've come to learn. But two, it's actually an important stepping stone to get to the admission of Mars. You know, when they started the company, never was it like, we're going to build rockets privately so that we can launch things to provide internet to people. I mean, they might have, I don't know if they had the idea, wasn't on the founding team of SpaceX. They might have like had an idea of like what they could do in space, but Never would you kind of assign their value to Verizon or like Comcast, right? Right. Like expanding, you would never do that. That's not, that's not what their value is at all. So how do they do that? Because they're very good at defining where they're going and what the impact of that is. And they're very good at defining what steps they have to take to get there. Exactly as Nathan said. And as an entrepreneur, that's your job. I actually think one thing people conflate, and I explain to them typically, like people that are going to start a company, is they're like, oh, you can just get funded to do research. And I'm like, whoa, whoa, we don't do that at Radical. I mean, we do do research, but we are constantly pressure testing our ability to capture value, not just create value today to feed the long-term mission. And Nathan said this perfectly earlier. Look, we want to go after RTAP superconductors. We, we've said that from the, from day one, if you build the largest experimental dataset in the world, you can tackle problems like an RTAP superconductor or a topological material or a novel battery chem, like, you know, pick your end application. That's super exciting. But you need to find the right path to get there. If you're a billionaire, you can self-fund it. If you have good networks, you can find ways to raise capital for it. You can bootstrap it, I imagine, in, in some form or fashion, but you, you, you must have this ability to think about what, what am I going to do that is accretive to my end goal and differentiate around that. At the same time, you must not be afraid to take risk. That accretive nature is not focused on, oh, it's a, it's a safe bet. So like, that's a safe bet, and then therefore we can go after the crazy risky bet.. It just doesn't work that way. And again, if we come back to the same example, just for simplicity of Nvidia, they had a very good business on chips, right? They were a good company. They were making money. By no means was moving into a GPU, like, not a risk at all. It was very risky. It was very research heavy. I believe that Jensen says he kind of got this idea on going to conferences, research conferences, and like paying attention to what the community was working on, what the research community was working on. That's about as fundamental as you can get from a research perspective, but like a deep understanding on the roadmap of like, where does it compound? Where do we go next? And then how do I commercialize it? What's the business around it? And I think this is really, really important. And we have great people around the table, Nathan included, that constantly pressure test us on this. Joseph, love the idea on RTAP, dude. Yeah, you would. That's like a, That's like a $10 trillion opportunity, if maybe more, if you figure that out. What about now as well? What are we walking through? And we're always thinking about ways to assert value. Alloys and high entropy alloys, as Nathan said, is one way we think about that. Self-driving labs are one way we think about that. Novel ML architectures are one way we think about that. Size of dataset is another way we think about that. There are these mechanisms that you must be paying attention to. As an entrepreneur and as a founder, that if you're going to go after some of these ambitious use cases, you must solve and look no better than the biggest, best companies in the world. NVIDIA, SpaceX, OpenAI, Anthropic, Tesla, like pick your company. Figure, if you want to go into the humanoids, that boom, Supersonic announcing that they're actually selling their turbine to power data centers so they can get the supersonic flight. Like the best companies in the world are very aware about how to build a tractable business, excuse me, but also not give up on their mission that they have that north star with which they're going to move towards. That's what makes a sexy but exciting early-stage company in my personal opinion. Well said. Yeah, let's zoom out for a quick second here on the long game. And Joseph, you had said that the biggest bottleneck isn't discovery. I was reading through some stuff. You said it's manufacturing. So even if we discover the perfect material tomorrow, what still kind of stands in the way in your opinion? Yeah, really good question. I love this question because this is a lot of the thesis of Radical and it actually is a perfect complement to what we just talked about on business side. So we were probably, we started the company 2 years and change ago, we were going to build self-driving labs, right? So self-driving labs, it's worth qualifying, are labs that run themselves, right? They're, they're the Waymo of science. They're not Ford Blue Cruise where it's assisted driving, right? You take your hands off, but if you want to make a left turn, you still have to pay attention, know that you should make a left on, on 2nd Avenue, put your turn signal on and turn the wheel. That's not what a Waymo does. You get in the back of a car of a Waymo, you go to sleep and you wake up at your destination. You don't care what left, right, or stop it did in between. That's what we're building. And everyone called us crazy. CapEx intensive. You don't need to do that. Just do simulation. Software's a better business model in the short term. A million things that we were told. But when we did the analysis, we realized, look, if you want to get to big discoveries, like game-changing discoveries, you need data. And if you want to get that data, you gotta run experiments. And if you want to run enough experiments to make the data valuable to make new predictions and discoveries, You gotta run it autonomously with self-driving labs. Very core thesis to the company. And so that's why we built them. And now we're at a place where we are by far, I'm confident in saying this, and I'd love to see someone challenge it, the best builders of self-driving labs in the world. And from that, there are so many places that you can accrue value, one, but utilize the technology too to get to the North Star. You know, and, and that we have a big belief and have had a big belief from day one that that needed to be where we invested in as a company. And, and we did exactly that. So we have 5 technical teams in the company I can talk about. We have actually spent on hardware and integration and building fully robotic systems that can actually run experimental science that other people haven't. And we've tied those to problems that we think are really important. And also get us to the game-changing, again, to use the famous one, RTAP superconductivity. Remember, RTAP superconductivity is like floating trains, lossless nuclear or energy transmission. It's just like Star Wars might not even be— I don't know where that comes from that. We're not there yet. We are not there yet. I'm not sitting here today telling you we're about to discover that. We're not. We are building the technologies that together we believe will give us the best chance of doing it. Whether it exists or not, fair debate, but if anyone's going to discover it, it's going to be Radical AI. Just like if anyone's going to build a deep learning chip, a GPU, it was going to be NVIDIA. That's the perspective that, that we're taking. And I think that's why the self-driving lab, to your question, was a really important investment and a focus area for us. Nice. Yeah, I love the passion, man. I just want to quickly touch on the lab itself. We'll kind of close with this. So paint the picture for us. Like, if we walked into that lab, once we get past the robotic dogs, what do we actually see? I mean, yeah, to what you can explain, right? Yeah, sure. So, and you guys are welcome anytime if you're up in New York to come see it. So what we have, we've built this entire backend infrastructure that shows what the lab is doing. It's got a digital twin, it's got all the experiments and every data stream coming from the lab is accessible, clickable, viewable from this data dashboard. So from the outside looking in, so you haven't even stepped in the lab yet, you can have a full picture view of everything going on. And we're getting very, very close, 3 to 5 months away from giving a scientist an iPad, sending them to the Bahamas and letting them run experiments from said iPad. We're not there yet. But we're, we're getting very close. And then once you step into the lab, it doesn't look exactly like a normal research lab. Now, yes, we have science tools on benches in a circle, like, like maybe a normal scientist lab would have. But the running of those tools, the loading of the samples, the movement of the samples is autonomous. You know, we have in 7 of 9 tools removed humans entirely. We have 2 more tools to go that do still use humans. We do have humans in the loop still, but that are autonomous. And so, you know, you guys will come here and we will stand outside the window and I can tell you what you want to run. You know, give me a new alloy. You can make up a combination in your head. I'll put it in the software, I'll hit enter, and you will watch that be made, synthesized, characterized, and tested for performance properties live in front of you from standing outside the window. That's the future that we are moving towards today. And We're quite close to, we are months from that. That's sick. Yeah, the lab's a beautiful thing to behold. Like when we went and visited it, just walking through and seeing science happen in this autonomous way, like it's hard to even just describe how big of a deal it is, right? Like self-driving cars, autonomous vehicles, like it's the biggest sector in physical AI now. And it solves a very simple problem, moving people and things from one place to another, right? You apply that type of autonomy to a problem like humanity's mastery of the physical world, like it's very almost hard to imagine what comes out of it. And those are the types of bets we like. Those are the types of audacious founders we like, where you're thinking about how do we push the boundary in a way that we won't even, like, we're going to surprise ourselves. Like that kind of magic. That's, that's where science has really been pushed in the, in historically and the lab. If you guys get, get a chance, highly recommend visiting. Hell yeah. I like to think, what would Walter White do with a lab like that? A fully autonomous lab. He could— I'm not commenting. I can't confirm or deny. Cool. Well, yeah, let's put a bow on the main conversation. The last thing I do want to make sure that I point out, because you kind of touched on it, Nathan, it is a big deal, and kudos to Joseph and Radical for being a part of this genesis mission. At a high level for a lot of our listeners who maybe have no idea what this is tied to the Department of Energy, what is it and what does it unlock for you guys? Very quickly, what we view it is, this is my personal perspective, not Department of Energy's perspective. It is the US's acknowledgement that science is changing and the US being, they have the most advanced scientific labs in the world via the national lab ecosystem. Needs to adopt and very quickly implement this new form of scientific discovery. And if you read about what the Genesis mission is all about, bringing advanced compute and HPC via like quantum and chips, bringing autonomous AI that can actually help design and predict, and then bring fully robotic systems that can test and confirm and build an active learning loop, or what the DOE calls a closed loop system. That is exactly what the premise of the Genesis mission is. And we spent a lot of time with the Office of Science Technology Policy on the Hill, at the Pentagon, the Department of War, the Department of Energy, and letting them know that science is changing. And if you want to be competitive when it comes to discovery and implementation of new scientific discoveries, you need to do science this way. And I think they very quickly understood that, and to their credit, have put together a hell of a roster at making sure that that change happens at what I call not government timeline. Yeah. Yeah. A very, very fast timelines, not as fast as us, fast, but, but very fast for government. And I think to us, that's exactly what the Genesis mission shows and why we're so proud to be a partner. I'm, I'm proud to be an American and I'm proud, I'm proud that the US is focused on bringing this level of innovation to the nation. When I think about, you know, I'm a veteran and so when I think about what our warfighter gets access to by focusing on fundamental discovery and the scale-up of those advancements fast and aggressively, that's a good world that we should live in. And so I think, you know, it's a kudos to DOE for putting their foot down and trying to pull this off. Kudos to this administration, the president, on pushing that perspective. And we're happy to be a part of it. We're happy to do— give our part and support the American people in the pursuit of scientific discovery in the best way possible. Cool. Yeah, we're excited to continue to track that. There's a lot of— a lot of really impressive companies attached to that mission. So Kudos to Radical for being a part of that. When Joseph and I were talking about it, he phrased it as the Manhattan Project for American AI dominance and unimaginable. That's exactly what it is. That is. And then you think about the focus, the results that came out of the Manhattan Project and also Sputnik and the space race with the invention of semiconductors, essentially. This effort could have a really meaningful impact for US innovation. Cool. Awesome. All right, let's, uh, let's put a bow on the main conversation and, and close here with our last segment. It's the 5-second scramble, rapid-fire Q&A. Try to rattle off the first thing that kind of comes to mind. Um, why don't, uh, Sean, why don't you lead, uh, off with Nathan and then I'll close with Joseph? All right, you got it. All right, Nathan, 5-second scramble. Uh, first one is, you know, we here at the Pair Program, we pride ourselves on diligence. We have a team of FBI investigators on payroll. You have a creativity section on your personal website where you have some writing and photography and other art. Not exactly common for a typical VC or typical investor. So I was hoping you could explain a little bit more about the creative part of your website. Yeah, absolutely. So in college, I got really into photography, went down the design rabbit hole. I actually studied architecture, which most people don't realize, and dual majored in business entrepreneurship. And that's actually what sent me into the technology world. So I ended up starting a design firm, then started building my own company, still building my own product. But fundamentally, we really believe in the power of interdisciplinary approaches. That's where some of the most interesting venture-backed companies come out of. And I think that's a really important thing that we keep in touch inside of Infinite. Awesome. So on that theme, what's a non-investment, non-business book that has heavily shaped the way you do invest and you do look at markets and companies? Great question. There's a book that is phenomenal called The Genius of the Beast. They talk about the boom and bust cycles of humanity into other species, right? Even the way bee colonies grow, the way bacteria grow in a petri dish. And interestingly, it has so much to do with the reflexivity of markets, the boom cycles, the euphoria we've gone through, the bust cycles and kind of those troughs throughout history and We've actually seen it happen in the open market multiple times now over the last decade. Awesome. Awesome. What's the strangest place you've ever taken an important work call? Probably in the middle of the ocean. Starlink is a game. I did a sailing trip a few months back with my brothers from the Dominican Republic back to Puerto Rico, and we spearfished the whole way. And so like absolute middle of nowhere off Isla de Mona, I took a really important call on Starlink there and It was pretty beautiful to be old. That's badass. That's great. Another, another win for Starlink. Absolutely. That's my— if you had to teach him— if you had to teach— the woods in Montana hunting elk, I was on, I was on a Zoom call from Starlink. Oh yeah, outstanding. All right, Nathan, if you had to teach a master class on something that has nothing to do with your job, what would you teach? Ooh, that's a good one. Probably first principles thinking. I think it applies to everything. It can apply to art, it can apply to action sports, spearfishing, creativity into technology. But I think that's something that should be like required course in like grade school. It's the kind of thing that really changes how people view the world and how they view their own trajectory in the world. Awesome. All right. And then our, our final question is always the same. Is there a corporate philanthropy or other charitable effort that's near and dear to your heart that you want our listeners to know about? Yeah, it's a great one. There's a couple of philanthropies we support. So my dad was originally from India and we support a school there that really caters to kids that can't afford to, um, to go to school, period. And they've done phenomenal work. It's called the Sri Sai Deepthi Foundation. And then there's another school, um, in New York City, Jesuit high school for kids that really want a more interesting hands-on education. So they actually have this phenomenal program where they do a work study where throughout high school you actually get placed with some of the top companies across the across Manhattan, whether it's finance, fashion, go down the list, and they've done a phenomenal job there as well. Awesome. Thanks so much, Nathan. Back over to Tim and Joseph. All right, Joseph, you ready? Ready. All right. If Radical AI were a restaurant, what would be its signature dish? No idea. Filet mignon, because I like it. Yeah. That's a couple of references to some steak on this conversation. I like that. What's your favorite part about the culture at Radical? Oh, culture's like the most important thing for us. We're psychotic about culture. We focus an immense amount of energy on it. The culture is probably one of the most important things you have to set as an entrepreneur in a founding team to drive after big problems. The culture that I like here, so we require 3 things. First principle, bias to action, and then like relentless pursuit, just this ability to keep going. The bias to action is important. I banned the word strategy for a period of time because I just cannot stand the whiteboard with no action. And so I just think everyone has this innate nature to get stuff done. Hell yeah. Love that. What's, uh, what, what are some of the jobs that you guys are hiring for over the next 3 to 6 months? A multitude. So mechanical engineering, both from mechatronics to sensing, all the way through to manipulation. Definitely hiring a bunch there. On the material science side, material scientists across alloys, semiconductors, and, and a couple future systems that you can kind of reach out to us specifically and ask about. Robotics, particularly on software and automation. Working on things like computer vision, as well as working on manipulation. On the machine learning side, both AI research scientists, so more looking into fundamental architectures in the machine learning side, or ML engineering, so actually applied AI and production workflows from some of the machine learning technologies that we're building. And then standard software engineering, full stack across the board. That's it, huh? Yeah, we'll give those a boost when we, when we promote the episode. That's exciting, man. Oh my gosh. Your time in the Army, you know, what, what's a, like, what's a habit that you still carry with you today from that time? Wake up early. I love getting up early. I get up before the rest of the world on one, just to, just to know I beat everyone else up. So if you're, if you're operating today, I am already beating you and you should know that. Uh, two, it's actually really important for my life. Um, I'm, I have a tough, uh, a kind of a stacked calendar. I have a wife, I have, I have a new child. They're the most important people in my life and my biggest supporters. I'm deep into fitness and, and health, so I gotta get, I have to be up early to take care of personal business, make sure I hit the gym before I'm starting work at 6:30 or 7 in the morning. So if I don't do that, I don't get the things done that are important to me. Nice. What's a, what's a material that we all use every day that you think people totally take for granted? I mean, we'll be here for the rest of this weekend. I don't want to answer that. There are silver nanoparticles in most of Lululemon's clothes to contain odor and sweat, would help wick away sweat. I love some Lululemon, man. All right, wrapping up here. Uh, who's somebody that you'd recommend we should have on this podcast as a, as a future guest? That's a good question. I have to pick my alliances carefully here. Yeah. Um, you, you nailed it with Nathan, so I was gonna, yeah, nurse you for another referral. I think Eric Ho from GoodFire. I'm friends with Eric and they're doing really cool stuff and they have a cool perspective of the future. Eric's just a standup dude, like entrepreneur, has been a co-founder of companies multiple times and we work with their team and he's just a great person. And then the other one is Beth himself, Guillaume Verdon from Extropic. I'm a deep material scientist, material scientist, not a physicist. It's tough to have a conversation with him. Like sometimes I'm like, dude, you gotta like just play it down. You gotta slow it down, dude. I don't know what you just said. He's just a ridiculously smart dude and I just love his mentality on kind of pushing the world forward. So he's another one, you know, big fan of him. That's great. Yeah. To have him permanently, you know, put it into chat and be like, dude, just explain this to me as if I'm a 6th grader. Yeah. And sometimes even that explanation, I'm like, all right, it's still over my head. I got a long night. That's hilarious. All right, last question and then we'll wrap. What's a charity or corporate philanthropy that's near and dear to you? Yeah, so not corporate per se, but really important philanthropy is a philanthropy called Krause's Coats. It's run by my father. They collect coats, hats, scarves, and lightly used winter wear for homeless. And they've been doing it in Philadelphia for, I think, coming up on 20 years now. They have done over 300,000 coats before in delivery to homeless shelters and homeless inside Philadelphia. They've also looked at expanding to other cities as well. My dad runs it entirely by himself and, and the support of local help is deeply impactful to the Philadelphia ecosystem. I grew up outside of Philadelphia. I went to high school, St. Joseph's Prep, right in the city. The amount of people that stop and say like, is, is your dad Krause's Coats? Like, do you know that he gave us coats when whatever is super important? So if you have used outerwear, lightly used outerwear that you no longer kind of utilize or don't know what to do with, call me and I'll get you in touch with them. They have people that can do distribution pickup and they're already collecting, uh, for 2027 winter upcoming. Uh, they went through all of their distribution this year again by December, end of December, January, and continue to have been working to do delivery. So, uh, really important to me and I help out there where I can. That's great, man. Yeah, I mean, you look at this winter was so cold, man. Just, it always makes you think about people that are out in the streets and just hopefully they have something. So, uh, great calls. We'll boost that as well, uh, along with the episode. Uh, Joseph, Nathan, thank you guys for, you know, talking through the, the future of science and technology, man. You guys have been a blast. Great energy on this one, and, uh, thanks for joining us on the podcast. Yeah, thanks for having us. Great conversation, guys.