Ep.8: Lux Capital: When AI and Biology Collide With Josh Wolfe, Founder | VC

Welcome to the Just Raised podcast. I'm your host, Joe Sweeney and today on the show we are talking to Josh Wolf, founder and managing partner of Lux Capital. Lux is a deep tech VC. They invest in aerospace and satellites, biology, drones, 3D printing, and are incredibly good at identifying cutting edge science and backing the founders of taking it out of the lab and onto the market. I've had the privilege of talking to a few of Lux's founders and as a fan of all of these hard science companies that Lux is backed, talking to Josh was a lot of fun. I hope you enjoyed this conversation with Josh Wolf. All right, Josh. Well, I've been excited to have this conversation for a while. A really interesting place to start might be with the intersection of biology and computing and tech in general. Can you break down why you find that intersection so interesting?

I think most of the really interesting breakthroughs happen at the interstices of different disciplines. So pretty much everything we fund is some multidisciplinary mix. And often it's because you might have traditional biotech over here, we have a bunch of companies that do that, and you might have AI or machine learning, and we have a bunch of companies that do that in different verticals. But then it's the intersection of these things. And so I think we got a real first taste and one that was very successful with Recursion, which was founded by Chris Gibson. And he was basically taking AI and ML techniques with computer vision to look at cells, almost like we would look at satellite data, but was looking into like traditional ELISA enzyme linked immunosorbin assays with these 96 wells that would look and observe phenotypically like what happened to the cell when you introduced the drug, which was sort of contrarian and non obvious because everybody was looking for some genetic or biomolecular marker. And we instead basically said no, if you take this technique, you can find the aberrant cell and then that's indicative of some disease state and that could lead to a drug target. So I think that turned us on to the fact that nobody in pharma was really thinking about using those tools, and nobody that was using those tools was thinking about applying them to drug discovery. And so there's lots of companies that we've since funded that are at the intersection of healthcare and biotech and tech and traditional software. And so that ranges from everything now from cutting edge microscopes. Like a company we have called Icon, which was actually developed to be able to see out into the far celestial outer reaches and became the Nobel Prize winning Work of Erica Betsig and is doing subcellular real time video imagery of what happens when you introduce a drug into a cell and you can basically see how the organelles and the cell is metabolizing it. And so that's pretty crazy. And then we have companies like Benchling that are doing software for lab information management systems and electronic notebooks for scientists. And that's scaling like crazy. And then you have companies like Stratios that are basically automating entire scientific labs so that instead of people in white lab coats at the wet bench tightering and doing experiments, they basically are on iPad. And in the same way that they would program Pro Tools or Logic or GarageBand in a time sequence where the instruments, you know, musical instruments here, the instruments are scientific instruments ordered in a sequence to run an experiment with perfect high fidelity and reproducibility crisis which plagues science. And all of that is just the intersection of bits and biology. And so there's a real huge bane to tap there and lots of scientists and engineers that are coming off of that. One of the amazing entrepreneurs that came out of Recursion is an entrepreneur who started a company called Invada, which is one of our most recent companies. This one happens to be very cool because I'm really interested in just the intersection between the real world, call it objective reality, the natural world, and synthetic one that we create with our technology. And it turns out a lot of the biological drugs that we take, aspirin, this comes from 3500 year old discovery from the willow bark that people realized through trial and error basically of chewing lots of plants, going back to the Egyptians and Sumerians, that if you chew this, you felt better. It was a pain reliever. And sure enough, the active compound in that became what we today know as aspirin. We just didn't understand the mechanism of action and how to synthesize it back then. We do today, same thing. Five, six years ago, Nobel Prize was one for artemisin, which is the key component active in the drug for treating malaria. Discovered really by the first Chinese citizen to win a Nobel Prize in medicine. To you, Yu, which was basically going back to 400 AD and finding old literature and writing that would help to reveal nature secrets in that case for treating malaria. And in the 70s, the Chinese had a program to basically say, we got to figure out how our Chinese soldiers that are dying in Vietnam fighting the Americans from malaria can survive. And that led to this. The point of all this is that now we're using in a company like Invada cutting edge AI and machine learning techniques that are used in search for the Internet, let's say with old literature that has all this interesting ancient sort of discovery from tribes and people all through history. And they will come up with the largest chemical map of natural substances that are basically druggable targets and drugs. So yeah, we love that intersection.

I saw that primer just announced what they've been working on within the last few days, which is essentially natural language processing to read through millions of papers and be able to pull out, do scientific research for the scientists. On that note, with benchling primer hugging face, a lot of these companies that are applying both AI and automation to the lab. The AI part I understand, but in a lot of ways the automation seems like it's 20, 30 years behind the manufacturing industry. Is there a reason why it's taken so long for automation to reach the lab? Four or five years ago, when I was working in a lab in school, you had people manually mixing chemicals for the first couple years of their PhDs, which is pretty wild.

I think some of it is mass customization of experiments versus like mass production of objects. And so if you have an assembly line, you can vary that assembly line by different colors and types of technologies, whether you're manufacturing an iPhone, but there's still a huge amount of manual assembly, you know, many of those sophisticated parts. But when it comes to more creative endeavors, you just have a lot of humans involved. Historically, and only I think in the past decade have you really looked and had this sort of euphemism of dull, dirty, dangerous and arguably repetitive jobs where you would say, well, could a robot do that with better precision than a human can? And instead of replacing the human, which obviously there's no incentive if you are at a university or you are a line and staff scientist at Merck in R and D, you're not thinking, how do I be more productive and replace myself with a robot? Your higher ups might be thinking about that, but if instead you're like, how do I augment my job with technology so that it's man and machine, not machine instead of man? Then I think you'll see widespread adoption. But yes, for sure, manufacturing took things that typically the machine very clearly had a superior advantage to basic muscle. And here it's much more cognitive load of creativity and scientists who might have spent 12 years getting a PhD and have very discrete knowledge about how to mix things. The other thing which is interesting as it relates to science is in manufacturing, you want perfect fidelity and reproducibility. Literally in Everything that you're making, you want the next iPhone to look like the nth iPhone. You don't want any real variability and error in science. A lot of the great discoveries through history came because the following words typically followed some mistake, which was like, huh, that's funny. And so you actually want to introduce, almost stochastically, a little bit of error and randomness into the process. Process. So you need to trade off like perfect reproducibility and high fidelity science in experiments with the possibility of error, which often leads to something that you didn't know before. But yeah, no, it's a really exciting time to see high tech, the throughput of high tech, the computational capability of high tech applied to quite literally analog biological and chemical sciences.

It feels like the superpower that you have at Lux and your whole team has is being able to identify the these emerging technologies and actually find the people who are bringing them out of the lab and commercializing them effectively, which is really hard to do. There are a lot of really interesting technologies that get built in a lab that don't become billion dollar companies. Is there anything that you've learned as far as pattern matching or about when a technology is ready and how to find the team to actually bring it into the commercial sector?

It isn't universally one way or the other. Meaning it's not like sometimes we have an instinct around technology and we sort of set, okay, these are the people to do it. Nor is it, you know, we find the people and then, well, okay, let's see the technology. It's just sometimes a mix, right? Sometimes, I mean, quite literally. Before you and I are talking now, we had a team present and actually the technology was quite interesting. The team was very green. And so we immediately then started thinking, okay, who can we surround them with as advisors or directors or future hires that can sort of compliment experience that they might be lacking so that they don't make the same mistakes that others made in the past. In other cases, we have a certain instinct around a directional arrow of technological progress. So I always talk about this idea of the inevitable meeting the impossible. Inevitable are these arrows of progress where I don't know who the entrepreneur or the technology or the company is, but I have a very high degree of confidence that that's the way that it's destined. And particularly when that pairs with external perception that that particular technology is impossible. It's a perfect setup because we have very high conviction because of the read of inevitability from being students of history, of technology development and the rest of the world, or at least our competitors and venture or growth equity or whatever it might be are basically like, oh, that'll never work. So therefore the valuations are low and if we are right, the expected returns are very high. A lot of it is pattern recognition and we often get it wrong. Sometimes we think somebody reminds us of somebody else and turns out they're totally different. But you're really looking for somebody that is persuasive. Persuasive to us. That we leave the room feeling a sense of greed and ambition and awe and excitement. That we're excited to share this with other partners who might be in future syndicate or people that we do with, recruit to the board or people that we would pull out of big companies and say, you got to join this company because it's going to behoove us socially. It's going to look good for us to make an introduction to somebody because these guys are going to really impress. In other cases it's assessing their ability to tell stories, raise money, recruit. In other cases it's just the ever growing Lux family, which when we first started was very small because we didn't have a lot of portfolio companies. Now there's 150 plus active companies. And if you take the top 10 people each of those companies, from the CEO on down to their lieutenants, you've got over 1,000 people that at any given point in time there's one in a thousand chance that anybody might be starting a company in a given year or of course three years, you have three people that might give you new companies if they know two people, then suddenly the network is 2,000 people, potential entrepreneurs. So we're constantly just treating our companies as well as we can and being good partners to them. So that when they or somebody that they know and respect and really like and think is smart is about to start a company, then we want them to say, hey, you got a call, Dina or Adam or Bilal or Brandon or Shaheen or Peter or Josh. And different people might want to work with me. Some people, maybe they don't like that I take shots at Elon Musk, Tesla and his truthfulness. They might not want to work with me. Other people might love that. So we have a whole suite of different and diverse personalities inside Lux that might be best matches for different entrepreneurs. But we look everywhere, we look scientific labs, we look within our existing companies, we look at friends of our existing companies, we look at syndicates that we've invested with that are showing reciprocity and showing us investment opportunities. After we brought them into something. And so I always say it's not like a bus schedule. You never know when your next deal is going to come. And we might have a week where we've got seven investments that everybody wants to do and we might have seven months where there's absolutely nothing we want. I would say the one common theme and we all have different lenses through which we look at this. But for me I've got this quote that chips on shoulders put chips in pockets. And I really like people who have some burning fire inside, almost like the Ironman nuclear energy that's just like doesn't matter what success they achieve, doesn't matter what financial rewards they reap. They are sort of broken inside in a ultimately pro social positive way because it's a source of progress. And so I like finding people who might be the only minority in a mostly white homogenous town, person who was adopted, the person from a really broken family, a person who had some physical affliction or lisp or some handicap that they might have overcame. And generally that is just the great propulsive force for progress through history.

That's amazing. I've talked to a bunch of your portfolio founders like Chris at Hadrian and Varda. They all say stellar things about Lux. On that note, all of those companies are in the space industry. Starting with space, you've funded a couple satellite companies like Planet and also a couple satellite, at least one satellite components company. And then also Varda which is a cutting edge use case and Hadrian which is supply chain. So you're really across the stack. I mean that is the stack from supply chain up to everything except for launch really, which is kind of dominated by SpaceX at the moment. Although love Rocket Labs and pretty interested in some of the others. I'm curious where you think the most interesting places to potentially invest in the future of space space are. Are they cutting edge new technologies like space tugs? I just talked to a fantastic company, Atomos, or is it more along the lines of data and actually software platforms for these companies in space? What do you think of when you think of the most exciting areas in that sphere?

I think it's all of those things certainly, as you noted, I think SpaceX and Twins rocket Lab will in the future dominate launch. We do have investments in launch with early investors. Relativity, not a significant later investor, but our partner Shaheen was a strong table founder early on and then. Yeah, I've joked that if you took the railroad industry going back 150 years and basically turned it vertical, you would have the playbook of the sequence of serially, how it would evolve. So first you're laying down the tracks. In this case, that's the rockets and the boosters and the bike paths and the launch capability. Then you've got communication and commerce. And that's a combination of the actual trains and payloads they're carrying and who they're transporting for and what the economics of those are. We've got a company that I co founded with Bill Gates called Kymeta, which is no moving part, satellite antennas. That's going to be a really big deal as everybody is launching constellations. So SpaceX has Starlink right now that's static, they can't move. You've got Amazon, that acquired Facebook's satellite group O3B and OneWeb, which is closely affiliated with SoftBank. And so the number of assets that are going to be launched up into space to rain down bandwidth and the means to capture that bandwidth on the move, which is the most important thing because there's a lot of static terrestrial sources for bandwidth. But being able to do planes, trains, ships, boats is killer. So I think that company, Kymeta will end up being the winner in that space. Hadrian was basically looking and saying, you've got all these mom and pop shops, I mean, thousands of them, that are basically literally doing the cutting edge work for these cutting edge space companies. And there's got to be a way to build the machine that builds the machines. And Chris is just an incredibly badass entrepreneur. And we were really thrilled to get to partner. And back in Varda, this was an idea and we sort of looked at this and said, you know, is this something that is ready for its time? Part of us was like, this will be willed into existence, no pun intended, Even on the founder, Will, who's just awesome. And Will and Delian were just amazing at putting us together. And the people that they've attracted are absolutely top notch. And this is real. And they've hit milestones and exceeded expectations. They're going to be doing a new closing, a new financing imminently way faster than we expected. And so, yeah, manufacturing in space, launching space, communications in space, positioning and thrusters. And then a new area that my partner Bilal is starting to look at pretty significantly is information security in space. And so that's everything from, you know, the communication relays of very sensitive information, photos, videos, et cetera, and how that's encrypted and stored, certain things that are almost like cold storage where a particular satellite bus might be performing A mission, whether that's for a government or a company, and you don't want to actually transmit that so it can be intercepted, but it might require another satellite that only when it's in proximity has a protocol to be able to exchange that information. So space is really a space race. It has geopolitical implications of very significant proportions. You've got GPS system that is vulnerable. You've got a Russian counter system called glonass. You've got a lunar base that we wanted to do with Russia and Russia said, no, we're doing with China. Russia's pulled out of the iss. You're going to see independent private space stations and other sovereign space stations. China has already launched theirs and starting to dock. So space is going to be wild. It really is not just the final frontier, as Sci fi once called it, but just the next frontier with really important commercial applications for pretty much everything. Now I am actually very impressed with everything that is happening at SpaceX. While I'm critical about Tesla, I'm very not critical about SpaceX at all. But I do think that the mission generally that Bezos has envisioned of manufacturing things in space and bringing them back down to Earth is certainly the world that quite literally I would prefer to live in rather than like us going to Mars. Now the push to Mars and extraterrestrial life and other things like that is important because of the ancillary tertiary things that will come off of the discoveries from all that which even happen with our own space race. But certainly in terms of the near term, like next 20 year practical commercial billionaires, I think a lot of it is going to be how do we do things near Earth orbit and bring it back as opposed to trying to go explanat.

One of the companies that I just talked to is pretty committed to a nuclear powered space space future which really requires smaller nuclear reactors, also getting nuclear material to space, things along those lines. And I know you've invested in nuclear before. You've told the amazing story of starting a company focused on nuclear cleanup and then nine months later being picked up as the major contractor for the Fukushima disaster and selling it for like a billion dollars within the year. And given your experience there was. What do you see in the future of nuclear technology? Do you think that's moving? I mean, it hasn't really moved much in the last 60 odd years. So do you think that's a space that has potential?

Of course it has potential. The only barrier is actually not technological. It's human and political. I'll correct you we didn't sell for a billion. It was a little under half a billion dollars. That company Kurion, which was focused on nuclear waste, should have been a billion, but still a phenomenal outcome in the case of nuclear power in space. I mean, I'll be a little bit flippant here, which is that a actual reactor to me does not make a lot of sense because I think the dangers of some sort of explosion on rocket launch and nuclear waste raining down, you know, like that would just be very bad. And the gains that we would get from having power from nuclear in space would be far inferior to the ultimate form of nuclear power, which is the sun. And that is an actual form of fusion. And you can benefit that nuclear reactor by just having really efficient solar. And so that is how most things in space are powered. You can certainly have ion propulsion systems to move stuff around. But unless one were over like several decades building part by part nuclear reactor on another planet, I wouldn't be building a modular reactor and shipping it on a rocket with risk that it's going to blow up. So I would be a little bit skeptical about the merits of that. But nuclear, domestically, we've got 104 reactors here, we've got 103 nuclear subs. The nuclear navy is a very part of our military and the geopolitical balance as it relates to China and Russia. And so you have years and years, I mean, decades of very high performance and really smart people. You did have a labor scarcity and shortage really starting in the late 70s, 80s, as a lot of nuclear trained people from MIT went to quant shops and hedge funds instead of going into the industry. We've not quite had a formal moratorium, but certainly new builds have been slow. And then at a time when a zeitgeist around zero carbon power ought to have said nuclear is the greatest thing for us to fight climate change and global warming, you know, instead the progressive left sort of went hard for solar, wind, biofuels, ethanol batteries, electric cars, some of which are good, it makes sense. But they cannot provide non intermittent baseload power as you saw in Texas and Southern California and elsewhere. So at some point China will outbuild US nuclear. They will have a cleaner economy. Today, they don't talk about it, but they're building more coal than anybody. But in the near future they will. And we'll look and say, my gosh, how did that happen? But the biggest thing that's needed is not a technological breakthrough. There's tons of reactor designs. Many of those reactors have new And a half century of great safety performance. Three Mile island, which happened here in the US was actually an engineering proof positive that safety systems work. Chernobyl was a certifiable disaster. Fukushima was a disaster primarily not because of nuclear meltdown, but because of geological. But I think that the most important thing that's needed is a psychological change and people need to rename nuclear. It's just too taboo. So I've thought about elemental power which includes solar and wind and naturally occurring uranium that is enriched. And you would have carbon neutral high gigawatt power baseload that could supply power per million people. But I'm very long element of power as a citizen.

It's fundamentally a branding problem more than a technical problem. On the technical side, do you think just more modern, smaller reactors will be built?

Yeah, and here again, you know, a lot of people have broke their pick on the problem in fusion per se. I mean there's a lot of companies and there's a lot of, you know, alternative thorium power reactors. But yes, certain small bed modular reactors, all kinds of modular reactors that instead of building a gigawatt power plant, 30, 30 megawatt array kind of thing will be cheaper, I think more likely to happen around military bases, potentially another sovereign that wants to be able to show that they've arrived as a nuclear in the US the problem has been a little bit of the catch 22 of you need NRC regulatory approval. But to get NRC regulatory approval you need a customer. But a customer won't do it until you get regulatory approval. So that'll sort of be caught in purgatory for a very long time. And I still think like the Westinghouses and Arribas, Toshiba's et Als historically will be the ones that prevail in the future.

So you mentioned China there a few times, both in going to the moon, building space stations, in building nuclear power plants, things along those lines. And one of the things that gets talked about with Chinese innovation is government funding. And I recently talked to an Air Force captain who talked about how even when he was using the most cutting edge satellites, he was running on pretty much Windows 95. So it was just not up to date. And a lot of DOD dollars are devoted to five or six big prime contractors. And I'm curious if you've seen and at Lux, if you've been able to help your companies actually break in and access government dollars. I know that you work with a number of people across the DoD, but that seems like a major barrier to just Increased investment in the space industry in a lot of these areas of really hard tech where you'd like to see more government funding but haven't thus far. So do you think there's an opportunity for startups or companies to be able to work with the government more effectively? And what needs to change there for that to happen?

Yeah, great, great set of points. I mean, a few different things. One, historically our entire culture has been relatively sort of indoctrinated since the rightly criticized military industrial complex of Eisenhower, which was like you have these large prime contractors that are feeding the war machine and you've got these beltway bandits. And so that became a very taboo phrase, military industrial complex. And it's one that I think is deserving of scrutiny. And I would actually argue that the monopolistic position of many of their primes, the Lockheeds, Boeings, Raytheons, et cetera, is reducing competitiveness in the same way that any field that doesn't have great competition potentially doesn't have the best technology. And I've seen that firsthand at the edge of the sphere, going and spending time embedded in SOCOM overseas and seen some of the technologies and the frustrations that they voiced. And I understand the source of those frustrations also as a nation been frustrated by something like the F35 Joint Strike Fighter, which is in 300 plus congressional districts. It's still massively over budget and over time they'll do a better job with the B21. I think Chairman of the House Services Committee Smith is doing a far better job than anybody has done in the past. And I think that there's a real willingness and openness to see competition from small companies. Now most small companies have had a zeitgeist of don't work with government because it's bureaucratic and slow. And that has historically been true. They started changing that. Not with the small business investment stuff that they do. The sbirs, I think those are relatively not that meaningful. They started changing it a bit with DIU and AF works and things at SOCOM and elsewhere. And SOCOM itself, Special Operations Command, whom four star General Tony Thomas and the former head commander of SOCOM is a partner here, really had their own budget to be able to do the big crazy technological things. And I think a bunch of the other forces have seen that. And so the Marines and Army and Air Force and others have basically said like, we want to get this going. You never really had super high tech Silicon Valley like company since the beginning of the Cold War. I mean, Lockheed Martin itself started in Sunnyvale with basically zero employees four years later, had 25,000 people and was developing the most cutting edge technology for the military. And so the roots of Silicon Valley were the roots in the Pentagon and electronic warfare and radar and submarines and other things. And we've forgotten that. We now think about Silicon Valley as Snapchat and WhatsApp and Instagram and Facebook and Google and Search and all that kind of stuff. But there is a return and there's very few companies that are doing that. Anduril is one. You had Palantir, which was a bit apologetic. Palantir took what some could have criticized as a cowardly position to say we make the technology, we don't decide how it gets used. Anduril has come and we're a large and early investor and said unapologetically we want our men and women that are on the front lines of war in every facet to be equipped with the best technology. And they're not selling to commercial, they are strictly selling to the US and our allies and its defense. And it was a very complicated thing internally and morally because we had to answer the question, will you be comfortable if one of the technologies that we're funding is involved in a kill chain like an explicit order that uses a technology to take out somebody? And generally our moral view on this is that if you're doing something to reduce human suffering, it's a moral good. And so we can talk about that later. But, but going back to this whole thing, military industrial complex in the U.S. taboo. You mentioned China before. China doesn't just have a military industrial complex, they have an over state mandate that is called military civil fusion. So if you develop something technologically, even if it's by the way with US investors that are investing in some of these tech companies, it gets sucked up right to the CCP and the Chinese military. And so we are competitively disadvantaged with a culture that is insistent upon that system working. And we are slightly disadvantaged here in the US with a system that is loathe to work with Pentagon we are working actively. Every time I talk like this about changing that zeitgeist because you can be as I am, anti war, I don't want war. But you can be pro defense and you want the people who are defending us to have the best technologies and not be disadvantaged.

When you are thinking about things to invest in, what are the thesises that you're most interested in right now? You've talked about mental frameworks like sci fi to sci fact and figuring out what sucks like Nuclear waste cleanup. What else are you looking at right now? What do you think is most exciting?

It's all of those things. So, yeah, sci fi is human imagination. Everything around us, from the cameras that I'm talking on to the lights and the fiber optic cables and the pixelated screens and all that, I mean, it was once somebody's imagination, and then that imagination got crystallized by engineers. So sci fi is a Greek treasure trove of ideas that haven't yet been materialized. And so I love science fiction. I read a lot of it, I watch a lot of it. Graphic novels, movies, books, you know, and there's a lot of great people that basically use special effects to imagine technology. And very much in the David Deutsch camp, in the Karl Popper sense, that if a technology obeys the laws of physics, even if we don't understand those physics of how something might work, then it should be possible to exist. You can imagine it and it can be reduced to scientific practice, then it should exist. So sci fi is one certainly identifying directional arrows of progress and talked a lot about this publicly. But whether that is transportation, going from horses to horse drawn carriages, to electric cars, autonomous electric cars, whether it's going from a sconce on the wall with fire to an incandescent bulb to an LED light, whether it's going from spinning disks to mechanical to a solid state drive, or going from mechanically moving gimbals for space antennas to something that has no moving parts and is electronically scanning the sky, you can spot those directional arrows of progress and they relate to energy density usage and going from moving parts to solid state and form factors and all that kind of stuff. And so I'm constantly just being a student of history and looking at the way that things have trended. This is also what led us to the Control Labs investment, which was a brain machine interface that was non invasive, that Facebook bought for a little under a billion dollars about two years after we invested. And that was basically this idea of the half life of technology intimacy. You went from a giant ENIAC computer to a desktop computer, to a laptop, to your phone, to your watch, to your AirPods. And the directional arrow of progress was that technology was becoming more intimate with you, more natural. Voice and gesture would be the next modalities. Voice was dominated by big three. Gesture was basically limited to Xbox and PS4 for Just Dance and depth perception. And that led us to that. So we're always on the lookout for that. So science fiction, that reading scientific literature and just seeing where smart people are voting with their Feet and their attention motivated by their ambitions to stand out, to win status, to be an alpha male or woman by getting social currency of a breakthrough discovery. And so we're constantly reading scientific literature. I read voraciously. And so sometimes ideas just come out of nowhere. And then, yes, looking around and listening to what people are complaining about, sometimes there's things that actually suck that you don't know suck because you've become so accustomed to it. And then once you experience the new thing that doesn't suck, then you're like, how is it not always like this? So some of the themes now that are both sort of near term and practical, and maybe intermediate term and less practical, near term and practical. I do see a cold war on the horizon or already present with us and China. You have two very distinct political systems, two very distinct ideologies, two distinct Internets, one that is relatively free and open, even if it's subject to interference, and one that is closed and surveilled and censored. So you have a hard power competition which is going to be air, land, sea, space and cyber in every facet, every stratosphere, Cyber is particularly notable because there are no boundaries. The permeability is our increasing dependence on each other and networks and computers. And then there's a soft power race, which I think that there's going to be a big demand tailwind because of. And that soft power race is for prestige. And prestige historically has come from winning the Olympics. Music, movies, cultural exports, clothing, fashion, art, those kinds of things, literature and science. And so great scientific discoveries have bestowed prestige on the sovereigns and the countries and the individuals or the companies that come up with those breakthroughs. And often those breakthroughs were enabled because of some new cutting edge technology, an instrument, a microscope, X ray, crystallography, cryo em, whatever it might be. So we are funding all kinds of cutting edge microscopes, tools, hardware, as well as software that helps the productivity of scientists. Because I think the demand, funded in part by governments, will be off the charts. I've talked about things where like, you know, your phone is basically a remote control for everything good in your life, you know. But I can record with high fidelity, I can capture sound and audio and video. I can't capture smell yet. But I'm convinced that as an extension of our senses, which I believe technology ultimately is, microphones to amplify my voice, and earphones and headphones to amplify sound, and screens that can amplify vision, and cameras that can amplify and record vision. But being able to capture the salience of memories and smell. Whether that's a great meal you had, being on the beach in Brazil, visiting your old room, the smell of your grandparents home, the smell of shampoo of your first love, whatever it is. All those things I think will one day be able to be captured with the same sort of chemical signature that is digitized in the way that we can almost like sorcery. Pull out our phone in a cafe and pitch a zam and hear what song is playing. And a generation ago people would have thought that was like magic. So all kinds of things like that. And then always looking for this interstitial breakthrough where there's some scientist that reaches into the other person's discipline and is like, what if you did it this way? And like nobody ever thought to ask that question. And that usually leads to some really cool breakthroughs.

Yeah, I've seen AI applied to a bunch of different hard science fields. Do you think a lot will come of that?

Maybe. I think the most interesting use of AI for science will be searching through hundreds of thousands of old scientific papers and using NLP and machine learning to basically dissect, particularly on results and methods processes, sections of scientific papers and parsing abstracts and conclusions, and not only finding correlations between papers, almost acting in the directional arrow of progress of science itself. By which I mean you used to have single author papers, then you had like five author papers. Now your average paper that is making into science for nature has 12 authors from like six different universities. And so science itself has become way more networked and collaborative between different institutions, institutions and different people. And so you can imagine almost in the present going back to the past, using AI and ML to have collaborations between long dead scientists, because you're taking their literature and finding interesting parallels that they themselves never discovered. So that's one. But the second is as you build a repository of interesting considerations, if you accept that science itself is a process, it's not a destination, it's process. And the ultimate purpose of that is to come up with conjectures, to come up with hypotheses and ideas and speculate. And that's creativity, right? Whether that's discovering the benzene ring by imagining a snake eating its tail, or that's Einstein sitting as a patent clerk and thinking about the relationship between space and distance and time and coming up with a theory of general relativity. Computers which we thought would never be able to do creative things like write an article or generate art, or beat somebody at GO or chess, even though those are more parametrically constrained than say language are now proving that they can do creative things. And so creativity itself is sort of under the microscope, so to speak. But I believe that scientific AIs will be generating hypotheses and humans will be able to test those hypotheses and machines will be able to test those hypotheses. And I think that is like the most exciting thing that you will have robots and AIs that are effectively generating conjectures, criticizing them through experimentation, getting the output and results of those criticisms and experiments, and then proposing new experiments. And that will be happening 24 7, tirelessly, at speeds that we can never do. And so we will get more and more explanations about the way that the world works. And those explanations will over time become harder and harder to vary, which will be the expansion of knowledge. And that's a beautiful thing that gets.

Into generative design, where you look at 3D printing. You can build brackets or you can build something that is the same strength of a bracket with a lot less material. They don't look pretty pretty, but they function the same way. I recently read the book Burn, in which I think I've heard you talk about with Tony Thomas. And that book is fundamentally about technology outstripping human civilization's ability to cope with it. So mass loss of jobs, AI just wiping out huge sections of the economy. What do you think the biggest threats are on that front from a technological perspective as all of this moves forward?

Well, in the near term people do get displaced. In the intermediate term, more jobs get created almost always. And everywhere in the canonical example are like ATMs. When ATMs came, people thought you're going to have mass layoffs of bankers. And what ended up happening was the cost to run a branch became so small that you ended up with thousands and thousands more ATMs. But you also ended up with thousands and thousands more bank tellers that were now doing higher value stuff rather than just giving you change or doing withdrawals and deposits, now giving you wealth management advice, mortgages and that kind of stuff. So the aggregate number of bank tellers went up even as ATMs also went up. So the short term displacement, I do not buy the. Let's take people in the heartland that were truck drivers or were roofing people. And then there's a housing crisis and we're going to train them to be coders. That's just not going to happen. I think there's always going to be a class of people who are in society a little bit more adaptable and moving quickly and there's going to be another class of people who are being disrupted. And we as a country have a responsibility of how we deal with that. Just like some people are healthy, some people are sick, and like, what's the sort of insurance mechanisms that we have for that? At the same time, there are certain times when some of those people are going to be in very, very high demand. You saw that with the spike post fracking and people that were in the oil and natural gas industries for a while with housing, and then there was a bust, and then with solar, and then there was a bust. And so there will always be these very painful displacements. And I think just that becomes more of a political thing of how do we make sure that the people who suddenly are without economic opportunity are provided for while the system sort of gets out of a volatile period and resorts itself? The answer isn't just quote, unquote education. You are indebting a lot of people with the current system, which is really a system of legal discrimination for employers. And so that's a whole other thing about the future of college and higher education. But I think technology itself has probably always been feared. You had the Luddites that were attacking factories. I mean, you think about the advent of artificial lighting itself. People bemoaned, going back to Plato and Socrates, the written word, because it would erode our memories. I always say that if the most dangerous words are, this time is different. The most valuable words are, it will rot your brain. Because every time a parent or some politician utters those words, it will rot your brain. It is always predictive of the next $10 billion industry that has scared the former generation is embraced by the latter. Go on and make a fortune. So technology is a species. Kevin Kelly has really nailed this. You know, if you read the Technium I think was his book, or the way he describes it, it's like another kingdom. There's like mammals and reptiles. And it's a species. It's one that we should understand. It evolves more rapidly than other species do. And it's a language that people should learn. So I'm generally bullish on technology and science and progress. I'm generally bearish on human nature, which is more unchanging. We have to repeat lessons and learn them from generation to generation and help that the institutions that can both help the people that are suffering and advance, the motivated ones that want to continue to push us forward, are in place.

One of my last questions for you, I've heard you on a bunch of podcasts, but I've never actually heard the story of how you started Lux Capital. And I think you started Lux Research before that, which was my first job. Can you tell us the story of how you actually started the firm and how you built it to where it is today? You just raised two new funds.

Yeah, we went out to raise friends and family money. And I was joking. I had a lot of friends, I had a lot of family. None of them had any money. I got very lucky. We met a guy, Bill Conway, who was one of the founders of the Carlyle Group. And whatever mood he was in that day, he decided to back us. He would go on to sponsor our first fund because we never really didn't have a fund. How do you get a track record when you have no money? How do you get money when you have no track record? In his magnanimous way, he was like, I hope you make a billion. And he asked us really important questions. How are you going to compete with Kleiner and Sequoia and Accel at the time? This going back nearly 20 years. What's going to be your competitive advantage? And why does the world need another venture funder's time? Like 800 of them out there today. There's probably 2,000 out there. And that will be rationalized at some point in the next few years. I don't know when, but we were one of those upstarts, right? And there will be upstarts that came during this current phase that will turn into great firms. And we struggled to raise. Our first fund was like a pledge fund, a little under $10 million, just build money. And then we raised $100 million, but we couldn't even do that. We ended up 92.1 second fund. And I remember every LP meeting that we had and I went gray hair in the process and ground down my teeth. And there was no fundraising. We made the reasonably intelligent decision to carve out a niche that others hadn't yet carved out, which was in advanced materials and nanotech and sort of these cutting edge merchant technologies. We actually formed Lux Capital First. Lux Research was one of the entities which was Peter, my co founder's idea to start so that we can get access to some of the big companies that ultimately would be buyers of our portfolio companies, but in particular would be paying us to reduce their own ignorance around some of these cutting edge areas about whether they should buy or build. And we built that out, Peter and I. At some point we knew it was a success because we wouldn't have been able to get a job there if we tried. And so you had really smart people like yourself that came to work there. And grateful for that, we ended up selling that years later to Briegal Sagemond, which is a spinoff from a Goldman PE Fund. And that was successful outcome for a bunch of folks. And then we went to raise our third fund that was 245 million. And then the fund sizes got larger over time, and the slope of our fundraising got easier, and we were able to become more selective about the limited partners and who were managing money for. And instead of it being just wealthy individuals, it was foundations that were doing things with the money that we would make for them that we felt really morally good about. And so whether that's immigration reform or education reform or fighting cancer or childhood diseases or women's rights or minority rights, and just, you know, it's good to know that the money that you're making or these institutions are in turn going back and forth socially for society. And what started with Pete and I grew to a team now of 10 on the investment team, a little under 30 at the whole firm or between New York and Menlo Park. And I feel very lucky. I get to do something that I feel uniquely. Sort of my personality fits, you know, I have a little bit of ADD and. Or maybe a lot. And I'm interested in a lot of things, and I love meeting smart people, and I like being involved in things that I think are not going to change the world, because I generally find that to be a BS kind of thing, but really do something that nobody else has ever done before. And I love the social status that comes from backing that. In the same way that somebody might like finding an artist or a musician before everybody else has found them. And I think a result of that, in a very intellectual, honest way, ends up changing the world. But I like the people that are really competitively driven with chips on their.

Shoulders, a healthy amount of skepticism and a lot of curiosity. I imagine. My last question for everybody on the show is just about their vision, usually for the company, but in this case, for. For the firm, Lux Capital. Where do you want Lux Capital to go?

I just want to keep doing what we're doing and attracting brilliant people to our team, bringing people to the extended deluxe family, people that are at the cutting edge of fields and making discoveries or inventions in pursuit of truth, which to me is a very meaningful thing that people look back, particularly my kids, in 20 or 30 years, they say, my God, my dad was involved in launching this and this and this. And there's a group of amazing founders the likes of which you've interviewed and talked to that view us and what we did with them and for them in service of them they're just very proud of and that people that we made a lot of money for went on to do amazing things with that money in a pro social way and that to me would be like a very virtuous circle and it's be intellectually rewarding, it's financially rewarding and nobody when they're dying I've ever heard didn't want to have more children. I feel very blessed that we have three kids and nobody ever wished that they didn't do something really meaningful and purpose driven and I feel very lucky that I found that and find people that find that the companies that they start. So more of that.

Yeah, that's an amazing vision. Well Josh, thanks so much for coming on the show. This has been a lot of fun.

Awesome Joe, thank you.