Pars Talk with Shahin Farshchi

Doctor Shaween, Ph.D. is a partner at Lux Capital. Shawin empowers entrepreneurs aiming to accelerate humanity towards a fantastic future through fits of engineering. His investments, his recent investment at Lux Capital include Nirvana, the deep learning company recently acquired by Intel Planet Labs, which is launching the world's largest fleet of Earth imaging satellites Plethora, which is rolling out a fleet of robotic machine shops Flexlogix making chips that can reprogram themselves relativity space launching 3D printed rockets mythic, bringing powerful deep learning to edge devices and Zoox, what comes after the automobile. And these are just some of Shaheen's recent investment. And please help me in welcoming Shaheen. Thank you very much.

Great to be here. I'm going to tell you a little bit about myself and my journey which led me to being where I'm at right now and my firm, Lux Capital, what we do and a couple of areas that I find exciting. So first some history on myself. So I grew up here in the Bay Area. My dad was a student at Berkeley and so I grew up in Albany, which is a small town near Berkeley. And when he graduated his first job was in Cleveland. So I went to Cleveland and my mom wasn't too crazy about the snow, so we came back to California and she became very homesick, wanted to move back to Iran, so we ended up moving to Iran. So I had a little bit of a reverse immigration. I moved to Iran as a teenager which was interesting. So I stayed there for a few years, did high school there, did a year of college there and then I came back, started all over again, went to college, ended up at Berkeley, did ECS at Berkeley, always had a passion for cars and computers and space. But given the dot com bubble days because I did college in the late 90s. So I decided to do computer engineering and get rich and hopefully buy a car, a nice car. But then unfortunately I graduated in the bust. And so it was how many of you were here during the dot com bubble kind of crash era? Okay, a lot of you. It was pretty depressing. It was so depressing in fact that I went to Detroit in pursuit of greener pastures. So I got a job at GM and so I went to GM and I realized the hard way how, how terrible it is working for a big slow moving midwestern based company. So disenchanted I tried my hand at the automotive kind of technology space. I came back to California, did a PhD in electrical engineering. Like many grad students, I decided to start a company around my research. I thought my PhD research was the most interesting Exciting thing in the world, lo and behold. It was cool technology, but it was a terrible business idea. So I was making these miniature wireless vital sign monitors to be able to track patients in hospitals. And hospitals loved paying for them, but they didn't want to pay enough so that we come out positive. And so anyway, learn the hard way that it's really important who your customer is, what the market is, what the sales cycle looks like, and what can be the underpinning of a viable business. And I had none of those, unfortunately. So shortly after that, I was thinking about my next startup and my friends were saying, hey, you should get to know VCs before you need their money. That's the best time to get to know a vc. You shouldn't pick up the phone, call a vc or email a VC and say, hey, would you write me a check? You should say, hey, here's what I can help you with, and build a relationship with them. And they're not used to people actually offering them some type of assistance. So I reached out to a bunch of VCs who are not in California and said, hey, I'm in California. Back then, semiconductors were hot. This is like mid 2000s. I said, hey, I'm a chip guy. I'm in between companies. I'm thinking about my next startup, but would be happy to help you guys find founders and due diligence on chip companies in California. I have a footprint across northern Southern California. I went to Berkeley undergrad, UCLA grad school. So I have a lot of contacts up and down the state, and I'll send you guys deals. And many VCs said, yeah, sure, send us deals. But then Lux Capital, which was a very young fund at the time, said, hey, you know what, we'll give you an email address. We'll give you a business card. I'm like, holy shit, this is amazing. So I said, absolutely, sure. And so after that, a few months into it, they said, hey, you know, we'll pay you. I said, okay, sure. I was basically getting a small amount of money every month and going to conferences and shaking people's hands and sending deals their way. And then Fast forward to April 2007, about a year and change into it, they closed their first institutional fund. It was a $92.5 million fund. And said, hey, do you want to join us full time? And I said, yeah, sure, I'll try this VC thing out. And here I am 10 years later. So how many of you know what venture capital is? Okay, okay, so about half of you. So Venture capital is basically an asset class. That's the technical way of thinking about it, but in reality is that it's a means of funding a young company, a startup company. And the biggest, I think the biggest misconception is that if your company isn't suitable for venture capital, then it's a bad business. I think that's the biggest lie out there. The reality is that the vast majority of great businesses make absolutely no sense for venture capital. You can think about it as a, as a Olympian or as an athlete on steroids. Like you put someone on steroids, sure they can lift up a barbell, but it doesn't necessarily mean that they're going to be a great lifetime athlete by pumping them with steroids. And so the same applies here. Startups that make sense for venture are startups that with a modest amount of capital can inflect many times over. So you put in 10, 15, 20, 50, $100 million, you generate many, many billions of dollars in enterprise value. And that just doesn't apply to the vast majority of great businesses out there. It only applies to a small fraction of types of companies, depending on the nature of the market they're selling into, the nature of the customers that they're going after, the nature of their technologies. So many, let's say locally local oriented companies, companies focused on small communities, don't scale the way a venture backed company would. Many companies that are selling into say the healthcare supply chain probably don't make sense for venture capital. A lot of companies that are making, that are offering services don't make sense for venture capital. And I'm sure many of you in this audience can build great companies, not raise a dime for venture capital and become very rich and successful. So venture again is a method of financing. Companies that are fall under the very, very small category of being able to inflect largely with a modest amount of capital. And I can talk about a few examples later. So the way VCs work is that they raise money from large institutions. So endowments, foundations, trusts, well, not trusts, but endowments, foundations, public offices, public family offices, you know, universities. And the goal these institutions, their goal is to put their money in different asset classes. So they invest in public securities, they invest in real estate, they invest it, they invest in natural resources, they invest in bonds. A small portion of what they of their, of their asset allocation goes into private equity. And then a small portion of private equity, which is privately held securities, goes into venture capital. So you're talking about single digit percentages of their overall assets under Management go into the venture capital asset class and they expect that tiny, tiny piece to generate massive, massive returns for them. So they expect double digit percentages of IRR for that little piece that they invest. Because they can get 8 to 10% in the public markets and 3, 4% with, you know, bonds and these other asset classes. They look for that large return, which is why VCs are very much focused on investing in these companies that can inflect in a massive way. And so we invest that money over the course of typically a 10 year life cycle. So first five years of spent investing the money and the next second five years is spent harvesting those investments and we invest, we raise funds in multiple vintages. So right now Lux is investing out of his fifth fund. Our first fund was a very small fund that was basically one guy's money back in the early 2000s. Our second fund was the fund where I joined, the 92.5 million dollar fund which was closed in 2007. We closed another one in 2011, another one in 2014 and 2015. The current fund we just closed earlier this year. Any high level questions on venture before I get into it? Make sense to everybody. Question in the back. What is your deal with your investor? Oh, that's a good. Okay, so what is our deal with our investors? So the way it goes is that we make the final call as to where we invest the money that they've committed and then they get a share in the returns. So for every dollar that we generate in returns for the investor, they take 80 cents and we take 20 cents. So if we raise, for example, a. So our current fund is a $400 million fund. If we double that, then we take 80% of the 400 million. Sorry, we return. So first we give the $400 million back, obviously the original form that goes back in its entirety. And then of the remaining $400 million that we've returned that we've generated, $320 million goes to the, to the investors. And that remainder $80 million goes to us. The managers question in the front.

Yeah, probably forgot, you know, you know already obviously the LP concept. You didn't mention the LP concept.

Yeah, yeah. So the foundations. So the technical term for the foundations, endowments, family offices, high net worth individuals, corporations, the people who put up the money are the quote unquote limited partner for the reason being that they don't make the investment decisions, but they have a share in the profits because they're putting up the money. Question, what's the difference between VCs and hedge funds. Okay, so this is getting complicated. So hedge funds, this is a tough audience. So hedge funds typically invest in public securities. And the word hedge comes from the notion that they are market neutral, meaning that they invest long, meaning that they invest in the expectation or the anticipation of the markets going or the value of that security going up, and they go short with the expectation of a security going down. So that's how they hedge. So the biggest difference is that in venture capital, we're only going long, we're only going in one direction. We do not bet on companies going down. We have no hedge, basically, we have no safety net the way the hedge fund managers do. And although some hedge funds do private investments, they typically focus on public security. So they can do the hedging. I mean, it's almost impossible. Well, I don't know for a fact, but then as far as I know, there is no tried and true known way to short a private security, because private shares and companies are typically not liquid, whereas with hedge funds, when you're doing public securities, these securities are liquid. So you can go long and you can go short and hedge against each other. Does that answer your question? But the hedge funds do have the same structure. They do have a limited partner general partner structure where the limited partners put up the money and they take a fee and they take the percent, the carried interest, 20%. That structure is very similar to venture capital. Good questions. What's the risk for venture capitalists if the business fails? The risk is infinite. So the risk is all. Well, not infinite. Sorry, I want to say this on record. The risk is the amount of money that they put into the company. So all the money that they put into the company is at risk. So the risk is limited to the entire amount. So as a limited partner, all of your money is at risk. All of your money can turn into a zero. And that's what you do. That's why these large institutions who commit to venture capital only put a small amount of their assets into venture capital and they put the majority of it into public securities, real estate, hedge funds, these other kind of less risky asset classes. But then in return, they expect greater performance for that greater risk that they're taking. Question. And the risk of the venture capital is a time and brand, Right? The risk. Venture capitalists that they basically invest on the money of limited partners, the risk is basically their brand and their time. If there's that cut out. Yeah. So for the managers, for us, first of all, we put up some of the funds ourselves, so we have Skin in the game too. Is that have to or it's. It's typically the case. I mean, there are, I guess there are some out there. I don't know of any, but there are some that probably could get away with it. But, but normally the managers also put up the money themselves. So we also have money at risk. And yes, it's also the, the reputation of the firm that's at risk. Although the expectation is that you. That's why most VC firms are formed as partnerships. So partnerships have a process for making investments. And typically if that process is well documented, deal memos are drafted and there's a relationship, there's board representation. If proper governance and oversight takes place and an investment goes to zero, then the reputation of the firm is protected. You can have a company do very well and the reputation of the firm not be so good because of the way by which they made the investment. Lack of governance. So it can go either way. Question, Is there a percentage success percentage that is considered like a successful venture capital? So the expectation for the asset class is to generate three to five times, I would say two to five times a multiple on that, on that invested capital. And if you do the math on that, two to five times puts you in the kind of IRR range over 10 years of, you know, low to kind of, you know, mid ish, might say 10 to 25% IRR. And then if you do the math, if you do anything less than that, if you're doing 10%, then like, why isn't the endowment putting their money in the public markets? Why are they locking themselves up for 10 years and taking all this risk for, you know, for the same return that they would get by investing in an index fund, for example, you know, so I would be mindful of that. So when entrepreneurs come to me and say, hey man, invest in my company, I can double your investment. Thanks, but no thanks. Because every investment we make is with the expectation of returning something of the order of the fund. Because if you do the math, the reality is that the majority of the investments will not return capital. That's just statistics. So if you make bets with hopes of maybe doubling your money, then you're setting yourself up for failure. You have to make investments with the expectation that every single Investment will return 10 times your money. So if the math works in your favor and you have a couple of good outcomes, if a couple of those investments actually do get there, then you can return a couple times your fund. If a couple of them generate 10 times and the remainder of them not do that. Does that make sense to everyone, by the way? So don't go to a VC and say, hey, I'll double your money. They're going to show you the door. All right? Okay. So to tell you a little bit about what we do at Lux, the vast majority of VCs in the valley focus on the left. They focus on companies that are doing social, local, mobile, find a date, find a parking spot, rent your couch, rent a car, get your meals delivered. Those kinds of businesses do have that network effect, kind of viral growth aspect to that them that makes them very conducive to venture capital. You know, you throw some money into Facebook ads, you know, you, you do something special and unique, you grow your company very quickly. You know, you do growth hacking, quote unquote. Still don't know what that is. But anyway, you're able to generate a massive company with a modest amount of venture capital investment. And the majority of the companies out there, the majority of the VCs out there focus on that. And when they evaluate these companies, they look for very well defined metrics. How many users do you have? How many of your users come back next week? How many of them come back every month? How many of them click on these things? What's your ltv, what's your ltv, what's your cap, your customer acquisition cost, what's your LTV over cac, what's your cap over ltv? What's your MRR daily like? There's all these metrics that they use to measure you. And different VCs have different thresholds. When you hit some threshold, an alarm goes off and they make an offer to lead your Series A or to seize your company. So we are not one of those firms. We focus more on the rights, which is high impact technology investing. We like to say that we invest where capital is scarce and barriers are high. And typically those barriers are technology related barriers. Now I'll tell you more about what areas that I like in a second. So this is Lux Capital. When I joined back in 2006, I'm the guy in the back. So we're a bunch of young guys. We took ourselves seriously, we wore suits, we hung out with old guys to look more sophisticated. And we were doing a lot of work in the materials and nano space, nanotechnology space. And since then we've taken the suits off and we take ourselves less seriously. We've grown the team and we're doing a lot of more interesting things in AI and computer vision. We're doing a lot of Manufacturing, automation, robotics, space. And so, you know, just to give you a couple of examples, you know, we invested in Planet Labs, which is launching a constellation of Earth observing satellites. I'll tell you more about them later. We invested in Zoox driverless cars, I'll tell you that. Tell you about that more too. We invested in a company that was founded by a Nobel laureate that's investigating the connection between our brains and our guts and trying to find medication that would address neurodegenerative disease. Through that, we invested in Nirvana. And that picture is a chip from that was just announced a couple weeks ago where Nirvana's technology went into an intel chip. I did a rocket company. They're doing 3D printing for rockets. On the bottom left you'll see a really interesting technology that is able to overlay any kind of vehicle on a rolling chassis. So imagine producing an advertisement for a car and having just one car be a variety of cars in production. So when you're filming it, you film one, that little slice in the middle, you'll see the actual car, which was basically a rolling skeleton, and be able to put any body on it that you'd like for future incarnations of a to say, for example, a commercial. And so basically what this company is doing is bringing post production and production together. Right now, post production is a very manual process and they're doing this entirely in software. So I'm going to talk about three areas that I'm really excited about. One is driverless cars, the second is space, and the third is the brain. So first of all, driverless cars, it's really amazing how quickly over the past three or four years, driverless cars has gone from science fiction to zeitgeist, pretty much. And pretty much anyone who could have anything to do with driverless cars is getting on the bandwagon. And you've had a handful of startups raise an ungodly amount of money and some of them get acquired very recently. So all the automotive manufacturers, all the ride sharing companies, all the semiconductor companies, the consumer electronics companies are all getting into this game with the expectation of riches at the end. And why is that? Well, first of all, the hardware has come a very long way over the past 10 years. So all of the hardware pieces that are required for autonomous car to be able to see its surroundings are almost dirt cheap right now. So radar, which cost thousands of dollars 15 years ago, is now $100. Acoustic sensors, cameras, all of these have become completely commoditized. And so you can build the eyes and ears of a car for less than $1,000 of a driverless car for something on the order of $1,000. Now, except for the LiDAR. And LiDAR is what gives you a 3D picture of the world. And there's now probably 50 companies, and we're investors in one of them that are trying to reduce the cost of a LIDAR device. And so I'm very confident that the cost of all of these devices are going to make it very cheap and easy to build a driverless car. So that's the hardware and the software has advanced dramatically. This picture, believe it or not, is from 2011. And this is the intersection. This is an intersection in near downtown. This is Castro and the Central Expressway in Mountain View. So it's a very complicated intersection. You'll see a base map that they've had a priori and they've centered the car, they've localized the car on this map. It's the red Prius that you see at the intersection. And they're able to pick up all of the cars and the pedestrians and light posts and they can read the traffic signals. And this is from six years ago. And so imagine how much it's improved since then. And so a lot of this is attributed to many open source tools that are available now and a lot of work that's gone into this, a lot of labeling that's been done, a lot of that data being available broadly. So now you're seeing scores of companies, Companies like comma, AI Drive, AI TuSimple, Auto X, trying to build the brains, the software for these autonomous cars. So how does this all come together? Different people have different opinions on this. And my opinion is that in order to really enable the driverless car, the vision of a driverless car. When I say driverless car, I don't mean like autopilot, where you can take your eyes off the road. I mean a car with no steering wheel. And to realize that you have to reinvent the car from scratch. And the analogy that I like to use is the dawn of the automobile itself. How many of you guys remember in the mid-1850s when you read about the history of the car, the cars basically look like carriages with no horses on them. And these were coach built coaches, effectively, that tinkerers would attach motors to and would call them cars, call them automobiles. And the original Oldsmobile curved dash was one of those, the original Daimler was one of those. And these cars were great for rich people and you know, novelties at the time. And it wasn't until Henry Ford partnered up with the Dodge brothers. And by the way, Henry Ford tried to start car companies twice and failed. This was his third attempt. So if you're an investor in Henry's second company and chose not to fund him the third time, you would have been an idiot. Which is why we'd like to back entrepreneurs even if they miss the first time or second time. So Henry Ford, third time around, parted up with the Dodge brothers, and together they came up with this new vehicle architecture which survives to this day. And so it was rethinking the automobile from scratch, which made the concept of the automobile practical and mainstream. And so if you look throughout the decades, the automobile has pretty much stayed the same. And so my hypothesis is that to really realize a driverless car, you have to reinvent it from scratch. Why? Because the current car is designed around the driver, whereas an autonomous car needs to be designed around the passenger. Autonomous cars, well, traditional cars only go in one direction because we have eyes in the front of our heads. We don't have eyes in the back of our heads. Current cars only have the front wheels turning, unless it's a sports car or a race car. But typically cars only have the front wheels turning because the suspension of the car would not make it, you know, not be drivable if all four wheels were turning. And we would do, like these craft style cars. But then if a computer is driving the car, then all of those encumbrances can be removed. So you have a car that's bidirectional. If you have a car that can laterally go into a parking spot, you can have a car go down a narrow street and then come back out that narrow street instead of having to do a turn. So that's what Zoox is doing. So Zoox is an excellent example from multiple perspectives. First of all, they're taking a completely vertical approach. So they are doing everything from designing sensors to designing software, to designing a completely new vehicle design. They're creating a new operations network for running these vehicles and offering an app for passengers to be able to hail these vehicles. So it's a complete soup to nuts. You basically pull out your phone the same way you say, hey, like shine, come pick me up. Where are you? You have a chat interface with Zoox. Vehicle will come pick you up. You sit inside the car, you put on your music, you change the lighting, you change the interior atmosphere controls, whatever the climate control, and you have a very, very unique experience as a result. So it's not just an Uber with no driver, and they're basically Doing this by assembling what amounts to basically four companies put together. It's not just like one company trying to do one thing. And given the nascent nature of this technology, in my opinion, this is the way you do it. And this is probably the only company that's doing it this way. They'll have to raise a ton of money to realize this vision. So far they've been doing really well. Another aspect that's really important to touch on here is the nature of the founding team. I like to say that I look for founding teams that have cognitive diversity, meaning that they view the world in different ways and compliment one another. So in this picture here you have the two co founders. On the left you have an Australian creative artist who started four design firms and had this vision for an autonomous vehicle driven urban environment and basically tracked down the captain of the Stanford driverless car team who did his PhD in computer vision and sensor fusion and convinced him to partner up with him to start Zoo. So you have this technical genius who is known as a God in his field and can attract great technical talent and you have this creative guy who can attract excellent creative talent and together create a once in a generation company. And so not all my companies have that kind of profile, but those who have done very well. In the case of Nirvana, you had a neuroscientist coupled with a PhD electrical engineer who made chips that mimic the brain. And that's why intel decided to get them to buy them. So that's autonomous cars. Any questions on this topic?

It's more of a general question than just this topic. One of the main questions is always, is the market ready for this kind of technology? Right now a lot of people are asking what happens to jobs? What happens to people's jobs?

And.

How much of the, how much of your focus is on when you're doing your research and how do you focus on that part of this?

So this is not just related to autonomous cars, this is automation. Because we do a lot in automation, robotics. So the question is how do we feel or how do we look at the possibility of annihilating jobs when we're making investments in and robotics and automation? We're invested in a company called Visor, which is automating the process of tax preparation. So what happens to all the tax preparers? We have a company that's automating machine shops. So what happens to all the guys that, that you know, handle these tools, that use these tools? What happens to the drivers when you have this autonomous car company? And so the truth is that yes, you are destroying jobs. But if you zoom, if you zoom back from a macro perspective, if you look at economies that have adopted technology, that have adopted automation and compare those with economies that have not, you'll see that companies like, you know, name your underdeveloped country which has not adopted technology and automation typically has more unemployed people. And if you look at the economies that have adopted automation, have adopted technology, typically have less unemployment. So that's the kind of high level macro views. The question is, well, what happens to people whose jobs are destroyed as a result? And I think that's where education comes in. In my opinion, our educational system is screwed up. I don't believe in the notion of it just doesn't make sense to me why you should spend the first, you know, 25, 30 years of your life, depending on what you're doing, getting educated in something or getting tools that will be obsolete probably single digit years after you graduate and you're laden with debt for the first 10, 15 years of your career because of that. So I think our educational system has to change as we adopt more and more technology. So long as we are constantly being trained as technology changes and we don't have people that are unfortunately disenfranchised as a result of their jobs being automated and this younger generation being prepared to take over the new jobs to do the new jobs that are created as a result, that needs to change. Does that answer your question? Yes. Question just what are your perspective on the regulation side of it? Not everyone wants to adopt that, let alone the regulation. And then the conflict between like self driving car and the cars with drivers. Have you done any lobbying or things? Well, I don't know how to lobby, otherwise I wouldn't be standing here, I'd be on top a little. But I, you know, so I think there is a, I think there is a, there's a codependence in my opinion as it relates to this. So for most, for many kind of automation, robotics topics, it's very one sided. It's the technologists trying to push regulators to adopt the regulation. But I think automotive is unique because regulators know that tens of thousands of people get injured and killed in accidents, auto related accidents every year there's this crazy statistics out there that more people died in car accidents than they did in Vietnam whatever year. So like it's pretty crazy and everybody's aware of this. So the regulators feel it incumbent upon themselves to do something about this. And automation is one way. In fact, many people view it as the Way to make our roads safer. So the question isn't like, should we adopt autonomous cars for regulators, the question is when, when do we know that autonomous cars are safer than humans? Well, luckily humans set the bar very low, right? So, so they don't have, they have to be good, but they don't have to be that good to be better than me. Not that I've killed people, but you know, statistically anyway, so, so, and that's a very difficult question. I think that's the hundred billion dollar question in my opinion, is that when do you know when autonomous cars are indeed safer than human drivers? And the RAND Institute did like the study where they, they calculated how many miles a car needs to be driven until you absolutely know for a fact that that AI is safer than a human. And the number came out to like, you know, 4 billion miles or some ridiculous number. And so the question is, well then how many people need to die until you actually feel confident that the car is safer? I think some of that really needs a change in attitude towards consumer products right now. When we buy a consumer product, there has to be a guarantee that it's safe. When you buy an iPhone, you know that it's not going to explode in your pocket, right? But then with autonomous cars, we know these autonomous cars are going to kill people. They're going to be running over, running over kids and dogs and like, you know, that's something that we have to accept. So will we as a society accept imperfect autonomous cars knowing that less people are going to be killed and so, and the same is going to go for the robotic assistance and whatnot, you know, like say robotic doctors, like right now doctors cause a lot of deaths in our healthcare system. So when are we going to accept computers treating us, knowing that computers are also going to screw up at some point. But right now we are more tolerant to people screwing up than we are technology screwing up. And so I think over time we're going to learn how to choose to be able to better measure technology versus humans, knowing that both will make mistakes and regulators will come up with, well, I guess after this cultural shift, then regulators will be more comfortable allowing technology to take over human tasks again, knowing that the technology could also be imperfect, knowing that as a result people will be saved through the adoption of that technology. So to be more specific, there are a lot of laws that are being put in place that are promoting autonomous cars. In fact, a lot of them are faster than when I was what I was expecting. So in Europe you have laws that are allowing for pseudo autonomous cars so people can take their hands off the wheel and like the Tesla and not be alert all the time. I guess the solar sitting, sitting behind the wheel. So I think we'll see regulation sooner than you think. But that regulation is absolutely required for companies like Zoox to get into business. Question in the back. Should they contact directly or through a third party introduction? So that's a good question. I mean, I get many cold emails. And if you're sending a cold email, make sure that the first three or four words in your email resonate with that individual person because otherwise you just get buried in the noise. But a more effective way to, to get in front of an investor is to either, if you don't know them, to say, hey, here's what I'm doing, here's why. I think you'll be interested in one sense, preferably if you can find a way to that investor through someone that the investor has done business with. So the highest signal path for an investor, for a founder to get in front of an investor, is to go through one of their founders. So if you go to one of my founders, like one of these guys says, hey Shawnee, you better talk to Jack, like Jack is doing something really interesting, then I have Jax, Jack's has my full attention basically at that point. And then like lawyers and accountants and all those people, that's the signal starts coming down. Other investors, that's when it starts going down. But you can cold email, but make sure in that cold email that the message is very clear and crisp and resonates with the individual investor that you're targeting. Question. What is your vision on like having something like design to the market? So, because are you trying to, for example, sell this driver as a license to the big companies? Because this kind of thing is lots of like capital investment infrastructure. Yeah. So, yeah, so this is not for the faint of heart. I mean, and this company's going to have to raise like many, many hundreds of millions of dollars to make its first dollar in revenue. My personal view, as I was saying earlier, is that at this point in time when the technology and the regulation and everything is very nascent, you have to go vertical. Think about the dot of the computer. Like back in the 50s, I'm sure IBM was bending its own metal and printing its own circuit boards and fabbing its own silicon and writing its own operating compilers and writing its own operating system and writing its own IDEs and design environments and creating its own programming languages and hiring its own salespeople and selling to banks or the military or whoever. And so I think as far as driverless cars are concerned, we're in the same era. So if you were Bill Gates in 1950 and say, hey, buy my OS, people would have said, huh, what are we going to do with this? So I think here we're at least 10, 15 years away from some kind of a standard where you can create some module or some piece of the software or maps or simulators or communications technology, like communications protocols for cars, and expect somebody to be able to buy it from you for a good amount of money. I mean, normally, since the technology is so nascent, everything is so tightly integrated that like, getting, integrating something into this stack probably is now more difficult than just building it from scratch. Does that make sense? That's, that's my personal view. I may be wrong. I may look like an idiot, like, you know, two five, two three years from now, but there's a lot of companies, a lot of very smart investors who have the opposite view, which think that, oh no, we can build the software and, and build the simulators and the maps and all that, and license it to companies like GM and Ford and Daimler and Lyft and Uber and whoever or whoever wants to run a fleet of cars. But I personally don't have that view. Question back. For example, when you're deciding on whether investing first, second, third phases or whatever, do you on your team have people that have specialties in different areas? Like, for example, they can give advisement on what if they think it's a viable tech or if they're biomed? Is that how that works? Yeah. I like to say that we're jacks of all and masters of none. So I don't think I'm qualified, qualified to look deeply. So I did my PhD in electrical engineering. I know analog circuit design and MEMS. But if somebody showed me this MEMS accelerometer, I wouldn't know if this thing can be, can be produced, manufactured, and I wouldn't know if the thing would stick after it comes out of the oven. So like, I mean, I would, I would. But the thing is that we do know people who are specialists. So for the domain that I invest in, which is typically hardware, whether it's driverless cars, whether it's space, whether it's human machine interfaces, whether it's computer vision, like, I know the very best people in those fields who I can trust can tell me which is where the referrals. The question was about, like, where do you, who do you like? What's the high signal source of deals if it comes from people like that. So for example, I'm friends with the founder of Invensense, so if he brings me a men's deal, then I feel pretty confident that he, you know, has vetted this before bringing it, bringing this to me, or if Naveen from Nirvana brings me an AI chip deal, then I'm, I feel pretty comfortable that, you know, that this person, that these people are legit. They give the confidence. Yeah, yeah, yeah. But normally, again, another secret that isn't, you know, I guess or I guess that's not talked about is the fact that when I first joined Luxury, I was extremely paranoid and worried that technologies wouldn't work. Well, guess what? Teams make it work. So, you know, in my 10 years at Lux and having watched 50 plus companies, more 80 plus companies go through the process, like the founders, guys like you make the guys and girls like you make the technology work. At the end, it's not the technology not working that's the problem. It's people not paying for it that's the problem. So more times than not, you'll find yourself in a situation where you built the technology, you pop the champagne and it works. And you know, people are actually buying from you, but they don't pay enough, which was my case. Like, you know, like I had a gadget that worked but people were paying me less than what it cost me to build the things. And so, and then sometimes you actually are able to sell product with margin. But like as a startup you build up this software and support team that you have to pay, you know, let's say a half a million bucks a month to support. But your customers aren't buying enough to generate enough margin for you to cover that. So you're going back to your VCs and hey, you know, give us some more money so we can make the next generation product like it's this kind of unfortunately dilute yourself to the point where it doesn't make sense anymore. So a lot of companies fail because markets don't adopt or markets don't pay or even if they do it, they don't adopt fast enough. That is the issue with startups that many founders don't appreciate and investors don't appreciate early on, unfortunately. So you guys do your part. It's your customers that typically don't do their part. Okay, So, I mean, there's not a whole lot more that I wanted to cover. I can just power through this stuff real fast. So space is another area that I'm Interested in. So, you know, getting stuff to orbit has gotten really cheap and really easy. As you can see, there's an exponential growth in the number of satellites and spacecraft that's being launched. Another example of a great team that we backed at Lux on the top left, it's the team planet. So another great example of cognitive diversity. You have two NASA physicists, a public policy person, a bunch of nerds in a garage basically that had this vision of putting cell phone components into satellites, flying them. And you know, when I first met them again, space tech wasn't cool. And they were saying that we want to open source our satellites and give away the images. And I thought like, how is that a business? But like, you guys seem really smart. And so we funded them and now they've launched a constellation of over 150 satellites. They're capturing daily images of every part of the surface of the earth. And they're selling that to mapping companies, to hedge funds, to agriculture firms, to insurance companies. So it's turned into a really interesting business and more on kind of Moore's law on satellites. Another area that's really interesting is the analytics of these signals. So taking these pixels that are captured from these satellites and turning them into signals that are actionable. And so the analogy here is the Bloomberg terminal where all this data is gathered. Decision is made, decisions are made for the hedge fund managers to be able to buy and sell or short stocks. And so there's a lot of interesting companies we're investors in. One of the more little insight which is taking satellite imagery and drone imagery and turning it into actionable data. The last one is the brain. So this is a personal passion of mine. You know there's a lot of interesting news and government led efforts, initiative around better understanding the brain and interfacing with it. There's a lot of scientific literature around this and it's interesting how brain technology is no longer a medical device. We saw with the success of the Fitbit, people are more in tune to tracking themselves. They want to share information about their health. People are very interested in hitting peak performance, whether it's yoga, whether it's, whether it's hiking, whether it's other extreme sports. And so there's a lot of interesting companies that are doing work in the space. We're investors in Halo Neuroscience. So they have this headset that does really cool brain stimulation that helps you hit your peak performance by increasing mental acuity. So it's kind of like three cups of coffee effect wise. We have another company that, well there's other companies out there that are doing kind of brain machine interfaces. So you can have. There's one company called Neurovil where you wear a VR headset and then it can also measure brain signals and you can just play a game without using any kind of controller or any kind of eye tracking. It's just your thoughts. And there's also interesting stuff in the simulation space which I won't get into right now. I want to leave time for questions, I guess. And so yeah, I think that's pretty much it. So if you guys have any questions, I have to jet for the airport in 15 minutes, but you can contact me. That's barely legible. So my initials SF Like San Francisco at Lux vc. So drop me a line, be very concise and I'd be happy to answer any questions you may have. Question in the back. So these boys that you invested, they didn't have any traction or any revenue, how could you invest in them? I've invested in companies, I've committed to investing in companies before there was a company. So see, that's a very interesting question. So you know, it's all about comfort level in this business. So if VCs are comfortable with you, either because they know you or because they're very excited about a market or because they absolutely want to have a bet in a certain space, like for whatever reason, if they're very motivated to do something, then they will do it regardless of even if you don't have a company. Right. They'll say, hey, form a company. Like in my case, hey, start a company and I'll fund you. So like that's one extreme. And the other extreme is, oh, you know, come back when you have revenue, come back when you have a million visitors a day. You know, come back when you have X, Y and Z. And so that just when you hear that from an investor, the translation of founders should be okay, this person's not comfortable right now. And well, lucky you, there's a thousand VCs and you know, a 10 mile radius from here, 20 mile radius, well we get San Francisco, but whatever, 50 mile radius from here alone. And so you have many people to go to until just, you know, it resonates with at least one. No matter what state you are in, just go and try, right? You are eligible for venture dollars from the moment you come up with an idea or even if you don't have an idea, from the time you have a resume in your head, hey, I want to start a company. Oh, here, here's money. Starting company Like a lot of people, a lot of companies have been funded this way where people with resumes have shown up. Oh, you know, you did your last, we funded your last company. We made a shit ton of money. Okay, we're going to give you money. Take our money for your next company and figure it out. You know, go, go on a beach, sit down and figure it out. So that's one extreme. The other extreme is waiting and waiting and waiting until you're practically pre IPO. And a lot of VCs specialize in pre IPO, kind of last private dollar in investing. So you have a lot to choose from, which is great for founders. Question?

Yeah. About exits. I'm in budgets and acquisitions. I'm doing very large, but I don't meant doing smaller M a day. How do we find out what kind of companies you have in that you can watch and study and you can have an understanding of it meets this. I can bring the money.

Sure. So the question is, how should people who are, I guess like advisors, M and A advisors.

I bring the money also.

Yeah, exactly. So how do you track?

How do you raise that money?

Yeah, exactly. So how do you track. And I think, I mean, VCs love talking about their shit. So if a VC isn't talking about something, if I didn't talk about something here, it's because those people in those slides would kill me if I, if I talked about it. So we love showing off, we love talking. So we love blogging, love tweeting. So, you know, look at our websites, follow our Twitters and you know, and, and you'll get as much information as we can possibly share because again, we'll get hunted down. Yeah, we'll get hunted down by the founders if we talk too much. And it's, you know, I've been slapped on the wrist before, don't want it to happen again. Question in the back. How do you see the future of deep AI? I mean, aren't you worried that the future humans would be obsolete? And do you think AI should be regulated? Good question. So are we worried? Am I worried about AI? I think, you know, folks like Elon and you know, like, they worry. I worry. I mean, I do. I am concerned as to whether or not humans really matter anymore once you have general AI. But I just think it will take many, many years to get there. I think we're hundreds of years away, frankly. Has anybody tried using transcription on their iPhones? Like I was trying to do a send an email today when I was driving and it was horrible. So you know, it still doesn't know who Shaheen is. My iPhone doesn't know Shaheen. Like I said Shaheen, I put in Shahid. And so like the phone doesn't know its own owner. So I frankly am not worried if the most valuable company on earth doesn't know my name. After I've been using Apple products for 15 years, I am not worried about AI coming in and obliterating my race.

I'm not worried starting from Dragon's Peaks onwards, Nuance etc still hasn't gotten there because the device doesn't have enough processing power. You need to go back to the cloud.

Well, this is with going back to the cloud. So this is with infinite compute power from Apple still. So. And by the way, speech is one of the prime uses of AI. So am I worried about AI, you know, warlords, AI based like Terminator. Not, not, not. I'm not gonna, I don't think my great grandkids are gonna have to worry about it, but at some point, yes they would. And I think humans will have bigger problems in my opinion and whatever in 10,000 years than worrying about computers taking over.

You had some investment on deep learning. I'm just wondering, you're not trusting AI that much, how come you're investing?

Yeah, so the question is if AI. Okay, so if I don't trust AI, why am I investing in AI? I think AI can solve a lot of interesting problems right now, but there are a ways away from taking over the world. Like if you look at computing, computers back in the 80s weren't practical for stuff like this. Right. For doing presentations, but they were great for doing payroll, for example, for a big company. So I think everything has its, its place and its utility. And I think the utility of. There is a lot of utility for AI. There is money to be made with AI, but it's just not going to annihilate us in my opinion anytime soon. That's. So I don't want you to think that I don't trust AI and I think it's stupid. I do think that AI is over hyped. I think everybody is just putting a AI, you know, suffix on their companies and their URLs to increase their valuations. I think a very small subset of companies are actually using AI tools for their businesses. And I don't think AI is valuable in itself, but I do think there are companies that leverage AI to do unique things that could be valuable. Maybe I'm wrong. So we'll see.

Questions?

So how do you evaluate the companies.

And how do you know that this whole AI thing is not overhyped?

So the question is, how do you. Let's just make it general. In any hyped space, in any space that is overheated. Let's just say, like, you know, I was investing when clean tech was a big deal, green tech was a big deal. And it really comes down to getting to know the team and getting to know the asset, what they're really bringing to bear and the business that they're trying to build. Normally, when you're doing diligence on a business, the veneer tends to come off very soon. And so, you know, who's the customer? Why would the customer pay? How is this an interesting business? You will see through all that bullshit very soon once you start asking these questions. And so the onus is on the investor and the entrepreneur to make the case as to why this business is interesting and how AI makes it special and unique. And if that answer can't be communicated in, like, two sentences, then it's no, it's a. There's nothing. There's not much there. I can take maybe at five minutes. Maybe I can take.

No, that's five minutes behind.

Oh, okay.

Yeah.

Okay. Okay, okay. All right, awesome. Thank you, everyone.

Stay tuned for our November talk. Have a great night.