392. Cutting the 1st and Largest Check into OpenAI; Assessing Generative AI Moats, Monopolies, and Market Opportunities; and Why Now is the Right Time for ClimateTech (Samir Kaul)

392. Cutting the 1st and Largest Check into OpenAI; Assessing Generative AI Moats, Monopolies, and Market Opportunities; and Why Now is the Right Time for ClimateTech (Samir Kaul)

Samir Kaul of Khosla Ventures joins Nick to discuss Cutting the 1st and Largest Check into OpenAI; Assessing Generative AI Moats, Monopolies, and Market Opportunities; and Why Now is the Right Time for ClimateTech. In this episode we cover:

  • Taking Large Technical Risk Than Large Market Risk
  • Assessing Generative AI Moats
  • Investing in CleanTech
  • Investment Lessons & Best Practices During a Technology Revolution

Guest Links:

The hosts of The Full Ratchet are Nick Moran and Nate Pierotti of New Stack Ventures, a venture capital firm committed to investing in founders outside of the Bay Area.

Want to keep up to date with The Full Ratchet? Follow us on social.

You can learn more about New Stack Ventures by visiting our LinkedIn and Twitter.

Are you a founder looking for your next investor? Visit our free tool VC-Rank and weโ€™ll send a list of potential investors right to your inbox!

Transcribed with AI:

[00:00:00] Samir Kaul joins us today from San Francisco. Samir is a founding partner and managing director at Khosla Ventures, where he focuses on health, sustainability, food, and advanced technology. Samir led the firmโ€™s investments in Cadre, Guardant Health, Impossible Foods, Mojo Vision, NanoH2O, Nutanix, Oscar, Quantumscape, and Ultima, among others. Prior to co-founding Khosla, Samir spent five years at Flagship ventures where he launched and invested in early-stage biotechnology companies. Samir, welcome to the show.
[00:00:30] Well, thank you so much for having me on the show. Yeah, so the, the intro, you know, you’ve got quite the accomplishments in your background.
[00:00:36] Can you walk us through some of those and kind of your path to venture capital? Sure. I thought I was gonna be a scientist actually, depending on how far back you want to go, but I thought I was gonna be a scientist. So I did my graduate work in biochemistry and stopped a PhD program to go work with Craig Ventor in the early era of genomics in the mid nineties.
[00:00:55] So in nineties, like 97. And that was [00:01:00] a really exciting time. And it’s so funny how things have changed. You know, back then we were doing DNA sequencing by actually developing film. It cost over a billion dollars to do the human genome, but we were at the forefront of this new era and that was always something that was really exciting to me.
[00:01:14] And I decided to skip the science path and I went to business school to learn how to be, to commercialize and go more into business. I hadn’t actually seen a. P and l statement of balance sheet. I don’t even know what venture capital was when I went to business school there. At business school, I met my first post-business school boss, Newbar Fayette, who started Flagship Pioneering, which is probably best known for starting Moderna, which, you know, I have personally have four or five jabs of in my arms, and I learned a lot from no bar on how to combine business and startups.
[00:01:50] And so what I did there is. Find really talented professors and start companies with them and be there, kind of have enough scientific credibility because of my [00:02:00] work in genomics where I, I had four nature publications and a science publication, so they took me more seriously on the scientific side and convert those into businesses.
[00:02:09] And we did that with four different companies, one of which went public, two got acquired and one, one went out of business. But they were all at the forefront of technical breakthroughs. So the first single molecule sequencing company. The first real synthetic biology company were some of the things that we did back then.
[00:02:26] The first renewable fuels company, not to make ethanol, to actually try to make long change, harder carbons from fermentation. And Vinod actually invested in three out of those four companies. And when he decided to officially start a fund after he left Kleiner Perkins, he asked me to join him and another gentleman, David Wyden, to be founding partners at Csla and really move out to California from Boston and focus on.
[00:02:52] At that time what we thought was gonna be the next big revolution, which was in renewable energy and sustainability. And so that’s what brought me to Costa. [00:03:00] And we were a small fund back then, and I continued this finding ent, finding entrepreneurial professors and starting companies right out of the university, and then broadening my experience to make actual larger later stage investments as well.
[00:03:14] And now, And since, you know, we’ve grown from a small, a hundred million dollar, primarily Vinod money fund to, you know, approaching 20 billion under management and doing a wide variety of companies. So it’s been quite a journey I can only imagine. And hopefully we’ll get into some of those specifics today.
[00:03:31] But, Can we start off with kind of the thesis at KLA these days? You know, what, what’s the stage orientation? What, what sectors and what themes are you leaning into Smith? So the strategy has not changed from the day we started the firm nearly 20 years ago, which is to be bold, early, impactful, and to focus on technology that disrupt large industry.
[00:03:51] Uh, so that’s very little exception to that rule. We are always gonna look for large, large markets. We’re gonna look for [00:04:00] technologies that are. Very high risk, but very groundbreaking and, and let me talk about that for why that’s an important strategy. So we don’t like to take market risk, and the reason is that it’s very hard to predict product market fit until you get out there.
[00:04:16] And that it’s very hard to know how long the adoption cycle’s gonna be. How. Mo how long it’ll take to train a salesperson, what the right commissions should be, what the right quota should be. And I learned earlier in my career that if you’ve got something new that’s only marginally better, even though it might make sense, it’s very hard to adopt.
[00:04:36] So we, we always feel like it has to be a no-brainer. So like with impossible food, it has to be. Equal cost, equal taste or better, and, and plentifully available. Otherwise, people aren’t gonna buy it. You know, the, the markets are gonna be way too small and I don’t want to take adoption risk. So if it means keeping it in the lab or keeping it in r and d for a couple years longer, that’s fine.
[00:04:58] And that’s especially [00:05:00] important in times like this where financing for most industries is very tight, because I can tell you what the burn of all my r and d companies are. Within 5%, it’s between 250 to 300 K an engineer per year. That’s what the burn rate is. What you can’t predict is how long it takes for sales adoption and on these other things I talked about, and that’s where burn can get outta control.
[00:05:24] So you don’t want risk introducing something that’s maybe 20% better. It’s gotta be a no-brainer. Um, otherwise you’ll get stuck. So no-brainer from the standpoint that if you build it, the customers will be there. If you build it, it adds, it’s so valuable to the customer that the customer cannot afford not to incorporate it.
[00:05:44] It doesn’t because it nice to have, because you look at it now, many enterprise software companies are gonna have certainly reduced growth, but maybe even down in earnings and down in revenue. Because a lot of small bus medium business customers are going out of business or they’re restricting [00:06:00] what their spend is because it becomes a nice to have.
[00:06:03] But if you have something that, that that manager can’t afford not to buy, Then it doesn’t matter as much if it’s a down cycle or not, those companies are gonna adopt it cuz there’s no other way. How, how do you think about technical risk, business model risk, execution risk, right? If you’ve, if you’ve taken market down to zero, how do, how do the others factor in?
[00:06:25] Well, I think to, to be honest with you, it’s, it’s very market and technical risk are like a seesaw. So the more, in my opinion, the more technical risk you make, you take. In order to have a product that, that’s so good to make a product that’s so good that isn’t already out there, you’re gonna have to, you know, do what I like to call, take the fiction outta science fiction.
[00:06:44] So a lot of things, when I start investing them as a seed, I probably am not doing my job right. If most people say, don’t say, Hey, that’s science fiction. And then my job over the next five, 10 plus years is to make it where that fiction is out of science fiction. [00:07:00] By doing that, you take all lot of the market risk off, so now you’ve got elevated technical risk and that I’m very comfortable with.
[00:07:06] We talked about why in terms of burn rate, but it’s also because that’s where our expertise as a firm is. All of us have advanced degrees. I’m technical ods, technical swan, the rest of, and we have a. Team underneath us there, a number of PhDs in material science, ai, physics, multiple medical doctors. And then, then, because we’ve been doing this for 20, I’ve done this for 20 years, vedo for 40 years.
[00:07:31] Our advanced network is so broad and deep that any company that comes to me, a qu. I did a call earlier today with a quantum computing company specializing in finding new materials for. The industrial space, I did the call. It seemed interesting, and I had five people I’ve worked with on and off for 20 years who can now do deep detailed diligence on that.
[00:07:54] And because we’ve worked together, I can interpret, you know, the people that are over, I know who’s more [00:08:00] optimistic, who’s more pessimistic. I know how to titrate their feedback into making an investment decision. And oftentimes they go on to advise the company, which is a win-win for everybody. And if we decide not to invest, we’ve also added, you know, the, the, the founder or the CEO will be dis disappointed, but we’ll have gained a little bit of help from our due diligence, which is, I think a nice thing.
[00:08:22] So Mayor, you were the first check into OpenAI and it was, I believe, according to, to our research, it was the largest check ever written by at 50 million. Of course hindsight’s making you look pretty smart at this point, but walk, you know, take us back through the decision making process there. Um, what were the key risks and, and what were the factors that maybe, uh, gave you pause or almost prevented you from getting it over the line?
[00:08:48] So, yeah, it was the largest first check, initial foray into a company that we’ve ever made. So that always causes more consternation around the table. It was also, it wasn’t clear to a lot of [00:09:00] people that AI was gonna have this massive effect on society as it is today, right? This was multiple years ago, and you were a startup in an area that was obviously gonna need a lot of compute and storage and cloud infrastructure that was gonna figure out a way to have to compete with the Microsoft, Google’s, Amazon’s, IBM’s, et cetera of the world.
[00:09:22] So there was lots and lots of risks. Okay. And it being the biggest check ever, and it was also, I won’t go into the details, but more of an unconventional structure at the time we made the investment, but uh, we were long believers in AI and what AI could be. Obviously we, we couldn’t have dreamed of the rapid adoption, the rapid growth, the rapid monetization.
[00:09:43] I wouldn’t be truthful, but if you could predict that. But we were big believers in that. We were big believers that it would. Affect not just the software industry? It would, it would have, it would, its tentacles, would streak to many, many verticals. We also thought [00:10:00] it was important for this not to be developed just by large companies or by, or, you know, by potentially hostile nations, in that we needed a startup also involved in this area.
[00:10:14] Yeah. And then finally we’ve known Sam Altman for nearly two decades. And you know, he’s pr now everyone knows this, but it was, it’s always been clear to us that Sam is a once center generation entrepreneur. You know, he’s super smart, super hardworking, incredibly high integrity. And you know, when you have someone like, you know, a Sam Altman or a Max Leche or Jack Dorsey, or you know, the number of others that we’ve had, the.
[00:10:43] Great privilege of working with you. You know that you’ve got something special and you, you make that investment. And so those were all the things that went into, into the decision making. Samir, do you think this is a winner? Take all Mark. I think there’s very few markets this [00:11:00] large that are winner take all.
[00:11:02] I mean, just look at, you know, people are comparing this to the iPhone only because, not be, and I don’t agree with that in the sense that this is much bigger than the iPhone. They just can’t think of anything this big. Right. The and, and, and what’s the right analog? Is it the internet or is it the internet?
[00:11:19] Is it the semiconductor? Is it, you know, there it is. It is certainly printing press. It’s certainly of that. Yeah. Printing press, right. Guggenheim. Right. It is certainly of that scale and it’s, that’s exciting to be right in the middle of it. Amazing. Was it unanimous decision so, What I, from my rec. So we have an interesting model at our firm, which I think is an another important talking point.
[00:11:43] Look, the venture business is really interesting and once you understand this, it becomes, it’s an incredible unlock in your decision making process in that you only lose one times your money. Yep. If you then, if you succeed more than 50% of [00:12:00] the time as an early stage vc, you’re not taking enough risk. So if you accept those two truths, which now we’ve demonstrated over two decades to be that way, it’s an amazing unlock.
[00:12:11] And when you make a decision now to get those a hundred x 50, you know, 50 x, a hundred x, et cetera type returns, it can’t possibly be something that’s so obvious that everyone around the table. It is gonna be jumping up and down about it cuz there was, you would be, you’d be slicing off all of that alpha that you, that you have in that investment.
[00:12:32] So we have a no veto policy. So it is really more, it’s more is everyone, is there one of the managing directors, there’s four of us do. One of the four managing directors strongly support this investment. And there’ll be lots of discussion, there’ll be lots of pushback and we actually score it. As ones, twos, threes, and fours where one is a slam dunk and four is a boy.
[00:12:55] I really wouldn’t do that. You might want to think twice orric about it. This, [00:13:00] because things that are non-linear are not gonna be obvious and are gonna be very hard to get consensus on Google wasn’t a consensus. Yes. At, at Kleiner Perkins when it was done, you know the, the stories that Vinod tells us.
[00:13:13] And so was everybody a one? No. But there wasn’t a, a, a, a veto on it, and I don’t think you could have everyone be a one on anything that’s said so non-linear or it wouldn’t be that non-linear. Mm-hmm. The other thing that’s really important, Nick, is that when I know that I’m not gonna veto a deal and I can’t veto a deal, I’m gonna pay a lot more attention and I’m gonna feel much more comfortable giving real feedback to my partners because I’m not torpedoing the deal.
[00:13:45] I’m not repeating the investment. I can give real, real feedback. We had a discussion yesterday about a life science deal that Vinod was very supportive of, and I’m not, but we ended up having a really helpful 30 to 40 minute conversation about what I thought they should focus on, what could [00:14:00] work, what I saw as the pitfalls.
[00:14:02] And it’s a really good learning experience for the junior folks in the room because in the end, this is a mentorship business. If we just rubber stamp everything and don’t have these kind of discussions, the next generation won’t learn. So was everybody. There’s a, was everybody a one? No, but, but, but, but, but everybody was supportive.
[00:14:23] Well, and I guess if there’s risks that can be abated, whether it’s through the network or through certain strategies, then it’s helpful to surface all those absolute, absolutely. And what to focus on first, because this is, this business is also about, it’s a risk management business in the sense that you want, if you’re gonna fail, you want to fail quickly and cheaply and so on.
[00:14:44] These technical companies, at these deep technical risks, we do a risk matrix where we say, Hey, company X, What are the two or three binary risks? If this doesn’t work, the company is dead, and let’s focus on addressing those [00:15:00] first. Let’s retire as many of those as possible as quickly as possible before we hire sales and marketing and do like these, you know, other things because it’s all irrelevant if these one or two risks aren’t addressed, and you do that as quick.
[00:15:14] And I think that because we are technical, because we’ve been doing this for decades and because we have this advanced advisor network, I think we’re pretty good at identifying which risk to focus on first and, and how to hire the best people to address those risks. When, when it comes to those people, how do you know that they are the best people and that it may just take a different well, Like a different set of folks with a different lens.
[00:15:40] And part of the reason I ask is cuz the, the most successful product I ever developed was sort of a parallel spectrophotometer that could analyze compounds in drinking water at nanoscale. And um, that was a like a three-year development cycle. We had a team of chemists and engineers and it was very difficult.
[00:15:58] It was funded by a [00:16:00] corporation. All through every phase, but at one point, the entire chemistry team had to be replaced. And I remember thinking at the time, if this was, if this was venture funded, this would’ve gone to zero. And fortunately, you know, there was permanent capital there that supported it all the way through and it was a banner success.
[00:16:18] But I’m, I’m curious, how do you fail fast? And then also think about that r and d component and continue to, uh, invest in something that. The r and d is much more complicated than originally anticipated, so I wish you had called me. That sounds like a fun project. We should, we should have probably worked together.
[00:16:36] Then. Look here. Here’s what I would, what I would say you asked your first question was how do you know who’s the best people you don’t? I think if you succeed and two thirds of your hires. You’re in the top decile. It’s just hiring is so hard and you don’t, you can do the interviews, you can do the references, you can do the blind references.
[00:16:56] I still think the best [00:17:00] hires are probably only right 65 to 75% of the time, and you don’t know for at least six months. And so the key then becomes is when you know that the person isn’t the right person, is to remove them as quickly as possible in a, you know, in a very professional. Kind way. But the longer you keep the wrong people in, the worse it is because they hire, they influence and everyone knows who the bad apples are.
[00:17:24] And if they, you know, it, it’s discouraging and demotivating to those who are the, the stars. That’s one. Second, we do a thing called, I talked about the risk matrix. How we identify the key risks and how to, and the binary risks and how to address those first. In parallel, we do something called a gene pool matrix, where we say, let’s, you’re a spectrophotometer company, for example, right?
[00:17:44] We’d say, okay, let’s hire, let’s find a couple of different companies which have had, have addressed complex optical problems and let’s, let’s say, so we identify two such companies. Let’s hire two people from each of those companies. Then you’ve got the physics and engineering side of [00:18:00] it. Let’s. Go hire two or three people from a couple different companies there too.
[00:18:04] So you’ve got multiple different components of the key risks, multiple different people from multiple different companies all coming together and you throw ’em in a, in a lab, and all those different perspectives. Hopefully they’ve all failed enough different ways that together they can identify which ways not to fail and what things could and couldn’t work.
[00:18:25] And there’s enough diversity of opinion that. You get to those right answers. So, for example, otherwise if you just hired a bunch of people from Becton Dickinson or Carl Zeiss or something like that, you’re not, they all have been trained to think the same way. You wanna bring people in from multiple different places, and that’s where the innovation comes in.
[00:18:45] So for, you know, Airbnb wasn’t a bunch of people from Hyatt and Hilton coming together, right? Right. Square wasn’t until Keith Rabbi, my former partner, was the first person from a payments company to join Square, and he was a 40th employee. [00:19:00] You know, Uber wasn’t a bunch of Hertz and taxi cab company. So those innovations hap at some point you do need to hire people who know the what, the regulatory and the do’s and don’ts, et cetera.
[00:19:11] But, but at the innovation phase, there’s. You, you know, there’s this famous Orson Wells movie that we played, a clip from our CEO Summit where you know, Orson Wells says, I succeeded because I didn’t know enough to know this was so hard. And I think that’s that’s key in everything that we do on the innovation side now, as to how do you keep funding and keep going?
[00:19:35] Well, if it becomes a function of market size. You know, is the market still big enough where it’s worth waiting five more years, six more years going after it? You know, we’ve got a number of battery companies that have been at it for a long time, but you know, everyone’s electrifying. Their cars and batteries are still the bottleneck and there isn’t a clear winner.
[00:19:55] Is your team outstanding and are they sticking around or are they leaving right in the [00:20:00] end? They know better than anybody. They’re in the, they’re in there 60, 80 hours a week, so if they’re staying and they’re hanging in there, That’s another really positive sign. And then it becomes a financial thing is, you know, how much money are they spending and do we have the wherewithal to keep funding it?
[00:20:15] But we tend to be, if the prize is big enough, if there isn’t a clear winner. And if the team that we have faith in is sticking around and still working hard at it, we tend to be pretty patient and it’s, and it’s worked for us. Tamir, what is the best resource you’ve come across with regards to hiring?
[00:20:34] You’ve talked a lot about talent and how hard it is, and even the best, you know, two get it right, two-thirds of the time. Any, any resources in particular that you found very helpful? Look, there’s search, I’m, I’m not gonna be an advertisement for search firms, but they’re obviously a few that were most comfortable with and, and, and I think it’s more a function of they know us.
[00:20:52] You know that they know who we would like, they know how to position people with us, et cetera, et cetera, as opposed to them just having a better Rolodex. I’m not sure that’s the [00:21:00] case. And then it just, you know, this is another big advantage of the deep tech space. Having been in it for 20 years. You have this accumulating advantage that we’ve probably known a lot of executives and people at these various companies.
[00:21:15] And I, and I think the. The personal ref referral network. If there’s a CEO or someone that I know and trust who’s worked with me for a decade, if they make a recommendation, hopefully that’s a, that’s a, you know, a higher value than just going on LinkedIn or getting the initial list of 20 candidates from a search firm.
[00:21:32] So I think it’s being in the space. And, and the other thing I’d say is that I always look for failure and how did that person, so I don’t have a problem if someone. You know, left Tesla, which then became a massive success and joined Fisker, which did not become a success. That’s, in a lot of ways, that person might be more interesting than someone who just stayed at Tesla because that person has taken risk.
[00:21:55] So then you have to ask those questions. Why did you leave Tesla? Why did you like about Fisker? [00:22:00] And then maybe for me at least, the most important question is what? What did you learn from your time at Fisker? What were the mistakes that were you made in making that decision? And what were the mistakes that were made there?
[00:22:10] Because I. Every startup’s gonna have mistakes. My goodness. I mean, you know, Facebook did a massive down round before they went, public square went public at half. Its last private round valuation. And these companies are, no one’s crying about those investments. Now they’re fantastic investment. Tesla went bankrupt nearly three times.
[00:22:26] But the learning you get from there versus someone who went to Stanford and Google than Facebook. I mean, just there, there’s no way that person can be super valuable in a startup, especially in the early days when failure is just around the corner. Yeah, I mean, something we’ve noticed is, you know, sometimes the success stories like the Teslas create more false positives.
[00:22:46] So you don’t know if the person was there and was a key driver of the success, or if they were along for the ride. Well, that’s where the network becomes important, because we know who the key drivers of success at Tesla were, and we have to ask them, what [00:23:00] did you think of this person versus someone who just kind of rode along for the ride?
[00:23:04] Right, so I’m not sure Tesla’s ever had a non turbulent year, an awesome company, great product, and Elon is clearly a entrepreneur. You know, of our generation. But yeah, that, that, I don’t think that they really haven’t had smooth sailing ever. So, no, that’s for sure. One more question on Deep Tech and then back to, to AI and open ai.
[00:23:25] So, you know, deep Tech is, has some different. Follow on financing. Mm-hmm. Dynamics than other categories. You know, how do you think about reserves and being offensive? Defensive, making sure that you’re protecting your positions when sometimes, you know, follow on financing can be more shallow or sometimes technical milestones, you know?
[00:23:47] They don’t come in a predictable path, and it’s not just about bookings, it’s, it’s about a lot of other factors and and risks that need to be addressed. Yeah, I don’t worry as much. Those are all [00:24:00] very good points. The flip is the deep tech markets are massive. You know, they’re just massive. So you know, the odds of having like a.
[00:24:07] Two or $300 million valuation software company that gets acquired by Salesforce or just Low. So a lot of these, like the sales product market fit stuff we talked about earlier is not as much a mystery. You know, if you build it, you know you’ll scale revenue. The risk will really be can you build it and what’s the cost structure and can you scale it?
[00:24:27] Right? We learned, uh, just like in.com 1.0 where a lot of companies failed miserably and people thought that. Dot com was done or the internet was done. It was a fad. You had similar people say the same thing to us after the Cleantech crash, 2011, 1213. Whenever, whenever people wanted, declare it. And I think by being in the middle of that and having all those scars on our back, we have learned a lot of lessons.
[00:24:54] And one of the things is that we’re entering industries where the incumbents [00:25:00] are welcoming these renewable technologies. So the in Cleantech 1.0, there are a lot of investments made in biofuels. And you know, Exxon and Chevron don’t give a shit about renewable or the environment or sustainability. They just don’t, you know, they spent more money in Washington shutting those technologies down than they did actually fostering innovation to improve human life, improve climate, and it would’ve improved, frankly, their balance sheet.
[00:25:24] Look what’s happened to Exxon. You know, engine one came in, put on three board members. I mean, they could have avoided all of that. If they just had a little bit more innovation and forward thinking vision 10 years ago. So I think that’s also what’s made venture so successful in biotech and in technology in general.
[00:25:42] And, and now in, in the drug world, you know, pharma companies have become much more, Hey, we’ll take your drug into phase three and commercialize and market it. And the biotech arms companies have become basically outsourced r and d arms for the pharmaceutical. So it’s a very. Cohesive, kumbaya type relationship.
[00:25:58] Everyone’s happy. You’ve seen that [00:26:00] in tech. Salesforce, Google, Amazon, apple, Microsoft. They’re constantly buying our companies and we all are happy cuz they’re, they’re fostering that innovation. And now you’re seeing that in a lot of areas in renewables as well. You know, whether it be in carbon capture or in battery technologies, in distributed power, you’re seeing a lot of areas where the large incumbents are investing, are doing distribution deals are becoming customers.
[00:26:27] You know, we just, we have a very interesting hydrogen company that has multiple strategics in their last round that are investors and customers and it’s a huge advantage. So we’re, so that becomes, A very important part of the investment criteria is that, second thing is there are more venture groups around the table, so that obviously always helps.
[00:26:47] But the most important thing is, is who’s working at the company. And I’ll tell you, when you go to campus, you know, I’ve spent a lot of time at Stanford or my alma mater, Harvard or Michigan, the new students who are, you [00:27:00] know, the future of the next wave of technology, if it’s not ai, they wanna work on social climate.
[00:27:07] Sustainability technologies, and some people want to do the combination of both using, using AI for sustainability. And that’s kind of the best sign for us in terms of where to invest is, where are the young people who have, you know, who have a lot of the horsepower and uh, where do they want to work? And they’re, and, and they don’t want to go to the large traditional tech companies.
[00:27:27] They want to do things and, and that are more meaningful. Is, is that the biggest change that makes Cleantech work this time around? I think that’s, I know that’s one, that’s one, yeah. I, I don’t think a new brand new area can really work without a lot of kind of learning since we’ve had those learnings. I think for most educated people, the debate on climate change is more clear now than maybe it was a decade ago.
[00:27:49] I think the IRA bill that passed is a, is, is another big, big sign of that. I think the. As tragic as the war on Ukraine has been, I think it’s woken people up [00:28:00] to, you know, energy independence being not just a neat thing, not just a thing for the environment, but crucial for political geopolitical stability.
[00:28:09] I mean, if we knew what was happening now, there’s no way those nuclear plans would’ve been shut down all over Europe. And I think that these are all things that have happened that, that a lot of them are bad. But you know, if you look forward 20 years, we’ll be. Reasons why these technologies would, will have advanced is because there’s a heightened sense of awareness and priority to them.
[00:28:29] What, what about on the, the science side? Are there key hallmarks on the science side that one can look for in, in the clean tech segment? I know in, in other sectors, in other segments, you know, certain tipping points with the science and the technology can. Then create large markets and large opportunity, you know, are there some key things that have happened on the climate side?
[00:28:52] Yeah, like, so it’s not as clear, like for example, like on on the healthcare side, you know, genome sequencing is much cheaper now. Right? That’s a huge [00:29:00] unlock. You know, if you can get a human genome for a hundred bucks or 250 bucks, it can dramatically change healthcare. It’s amazing. And on the tech side, The continued cost curve of semiconductors and cloud costs are, you know, dramatically important on Cleantech.
[00:29:16] Those things can help them too, but you, it’s, I think it’s much more adoption. So if you take the fact that Tesla, you know, completely 10, when we, when we first did Cleantech 1.0, I doubt more than one minute. Or one page of every board book of every automaker was dedicated to electrification. Now, I can’t imagine that at least half of the board meeting isn’t talking about electrification, so that stimulates demand now for that, right?
[00:29:43] You’ve got the war in Ukraine, so now all of a sudden you’ve got over a dozen credible nuclear fusion. Companies funded, right. Including one of ours, which we helped incubate called Commonwealth Fusion, which we’re very excited about. That would’ve been unheard of to do a fusion company over a [00:30:00] decade ago.
[00:30:00] Just wouldn’t have happened. So I think it’s much more those type of of events, which I’ve catalyzed it as opposed to something. Yeah. The closest thing I can think of is, you know, now you’ve got massive demand for copper and lithium. If you’re gonna get to a real. Renewable world and electrification in the automobile industry, you’re gonna need massive amounts of copper and lithium.
[00:30:21] And I think when you have that, I believe in free markets. I think the people are out there mining, copper and lithium. There’s more available and the costs become. You know, batteries are gonna continue to be a less and less as a percentage of the overall cost of an auto vehicle, of an automotive vehicle.
[00:30:37] One would hope so. I mean that, that, that’s an area of technology that has not progressed as fast as mining. I would have expected mining batteries. Batteries themselves. It’s a really challenging problem. And automakers are incredibly slow adopters, right? You gotta get a samples, B samples, C samples. You could be designed in multiple years in advance, but yeah.
[00:30:59] But. [00:31:00] Again, in 2010, if you had said by 2030 all automakers would be going electric and having, you know, I think that’s still 20 years is still pretty quick in, in the scheme of automotive. Certainly for an industry that large and, and that, and that uninnovative. Right, right. It’s, it’s one thing, inertia, it’s one thing for, you know, people in.
[00:31:25] In the tech world where they’re constantly looking for disruptive innovation to adopt these things, versus the automakers, which, you know, they made a whole movie about it, right? They killed the electric car a couple decades ago. Cause it just, it wasn’t convenient, it didn’t fit. We are looking at a lot of really cool mining technologies because it’s such an important market and it’s also bad for the environment, the pollution, the contamination of the tailing ponds.
[00:31:49] So again, innovation, you know, when there’s a market, innovation starts now you’ve got a lot of really. Talented molecular biologists working on biomin technologies, you know, accelerating the [00:32:00] metabolism of bugs that naturally reside on hard to mine oars and, and feed their metabolism and leach copper or lithium or metals faster.
[00:32:13] I mean, it’s amazing. Wow. Wild. Amazing. Well, while I’ve got captive audience here, I gotta ask you more about sort of generative ai. Sure. And you being in a first check in open ai. So in, in a world with a number of these competing LLMs now Samir, and lots of different applications, what do you think open AI’s moat will be?
[00:32:34] So first to market is really important here. If you look at how much momentum, I won’t go to the specific numbers, that’s really up for the company to share, but it’s. Breathtaking, how many million subscribers they had in the first couple months. How many downloads they had in the first few days of their app being launched, you know, their revenue traction.
[00:32:58] But we’re not even, we’re talking. I mean, it’s [00:33:00] amazing. It’s never, I mean, forget me, I’ve been doing tech for 20 plus years. I know’s. Been doing it for 40 years. No, nothing is even close. So, When you, when you have that massive advantage and you’ve got so many users, you have the ability to design these plugins and design the primary experiences for these large LLMs to get work done.
[00:33:22] You are kind of the conversation piece. Like everyone talks about OpenAI, they may try other things, but OpenAI is the first one and the more people that use it, the better it becomes cuz it’s a self-learning dataset for these people to learn more and more from. If you say who’s the leader, you say OpenAI, well what’s one of the risks is overregulation.
[00:33:42] So who’s gonna be at the forefront for helping to set, you know, Sam’s all over the place as he should be and he’s, you know, and he’s being very, I think, responsible. And not a maverick and how we should be dealing with this. You get it when you’re first, you kind of under you, you got, I think, a really interesting lens into what all the use [00:34:00] cases are and where it can apply, and I think that gives them a really, really huge advantage as well.
[00:34:06] And then of course, so you’ve got all these awesome things that come and look what they did with less than, I don’t know, 400 people, you know, they, they, they, they became the first credible threat to Google in two decades. Uh, with what? 400 people and Google has what? A hundred thousand, 70,000? You would know better than me in such a short period of time for just one application search.
[00:34:30] And they can go after so many different applications in healthcare, in, in insurance. I mean, there’s education, there’s so many different applications. So, and they have a massive partner in Microsoft. So, you know, you know, there, there’s be lots of competition out there. But, you know, I think. Again, Sam being a one center generation entrepreneur, the partnership with first of all, convincing, and there was an article about this, I think in the information a couple days ago where it talked about [00:35:00] how Microsoft had to kind of swallow their tongue on not invented here to do this deal with OpenAI, you know?
[00:35:06] And, and they did. And I think it’s a massive win-win for both of them. You know, for opening out, I got a ton of capital cloud, all the power of Azure and all the help you know, that Microsoft can do. It’s a huge partnership and it’s something that’s gonna be very, I think a huge advantage for them by locking up one of the large, large players.
[00:35:27] And, and from a competition standpoint, who do you think is the largest threat? Right? Is it one of the large US tech companies? Is it a big Asian tech company? Is it one of the venture backed startups? Is it open source? It’s all of the above. I mean, I think it could be, you know, obviously Google’s the obvious one.
[00:35:43] Open source is an obvious one. You know, it’s hard to know what’s happening in Asia, and I think we have to be. Very careful not to be too proud of our advantage, but you know, keep that advantage, grow that advantage. And, and you know, one of my all time heroes, Andy [00:36:00] Grove’s book, right? Only The Paranoid Survive and the company is certainly embracing that.
[00:36:04] No one’s doing victory laps there yet. And I think all of those are legitimate and potential competitors. Samir, you know, if we think about. Three different types of AI companies. You could have like an AI first company, maybe it’s infrastructure. You could have an AI enabled company. Maybe it’s application layer, or maybe it’s a company where AI isn’t a primary differentiator or driver, but they’re using it as a feature.
[00:36:29] Maybe a chat bot on an e-commerce website. You know, where do you spend most of your time? I think I know the answer, but you know, how do you tease out sort of the winners at the layer that you play? So we’ve got companies that we’re really excited about in all of the above. So we have AI applications, like fully automating jobs.
[00:36:48] So QI is one that we talk about a lot, which should augment, reduce the number of primary care visits that you should have to go to. You know, automated Poly AI is another [00:37:00] company that can fully automate jobs. You could augment humans like Lexion on the legal space or you know, Firefly for note taking in conferences and in conference calls and, and in meetings that are really valuable.
[00:37:14] There’s a lot of money to be made and a lot of innovation happening in the AI infrastructure place. So, you know, we in, in hardware, you know, you see what it, what’s happened with Nvidia, right? It’s breathtaking. So there’s, you know, I think there’s gonna be a new innovation in chips and inference chips.
[00:37:29] We’ve got exciting companies. There, we’ve got companies that are, you know, basically missing pieces to accelerate AI capability beyond just LLMs. So what’s next after LLMs? And I think that’s another thing that’s what’s really exciting about being at the forefront of these things is that you see like what’s possible before the rest of the space.
[00:37:48] So we do a lot, we do a, that exercise a lot where it’s like, Hey, okay, let’s assume LLMs work and let’s assume that they get what’s next. And so we’ve made a couple of investments in thinking about that. [00:38:00] And then the dev cloud, you know what’s gonna be the dev cloud for ai? And we’ve got a couple of bets there.
[00:38:05] One, we incubated with one of our partners, Nikita, called Neon, and then another company called Repli with a, you know, terrific entrepreneur named Amjad. So, uh, there’s a lot of different areas and in terms of teasing out the winners, it’s, it’s gonna get harder and harder. There’s so many companies, everybody’s chasing it.
[00:38:22] Valuations are getting bananas, and so you just have to stay disciplined. I think a lot of the insights we have, it was one of the thesis around the open AI investment too, was like, Hey, you know, with Sam there and with all of, with the early start, we thought we’d make a lot of money. Uh, but we also said by having, having early and access to all the really talented engineers at Open ai, we’d get a great network of people that once they leave, they’d maybe would come to us to inve for investment.
[00:38:51] Yeah. And as we’re evaluating companies, we would have really good, smart people at the forefront who could help us with due diligence. And that’s [00:39:00] really important. Like, you know this, you know how to evaluate potential entrepreneurs where we’ve done really well. Like we were the first money in Square, you know, we were the largest venture shareholder when I went public.
[00:39:12] And Outta Square came Tony Shu, who was an intern at Square, and we put the first money in DoorDash. We put first money in fair with Max Rhodes is another outstanding entrepreneur. We put and we, and, and other companies too level with Paul Aaron or Andrew Burrow’s new company and Anchorage, which is a in, in the crypto space.
[00:39:32] We did the same thing with Nutanix. We did the same. So you’ve got these multiple trees of just fantastic innovation that comes from this company. Jackie Reese’s, who ran Square Capital, were a big investor in her neobank, so. It’s one of these things that really helps. And, and that’s, that’s the advantage I think we have with our play in open ai.
[00:39:49] But we have to stay disciplined and we’ve stayed disciplined in our, there are companies that we’ve missed because evaluations were too high there. Companies that chose to go with other VCs because we wouldn’t match the valuations. We [00:40:00] have to stay disciplined. Samir Ev. Every new technology. Revolution’s a bit different.
[00:40:04] You’ve been at this 20 years, you mentioned the node’s been doing it for 40, so you’ve seen some, some major waves. You’ve seen the internet, you’ve seen mobile. Are there any heuristics or observations or investment lessons that seem to apply when you’re going through this? Rapid phase of new technology introduction and adoption that, you know, can be borrowed from, from previous eras and applied now in this new AI era?
[00:40:35] It’s a great question. I, I, I think the, I think the, probably the best way to answer that is that no one really knows where the future holds. It’s gonna be very hard to say which one. You know, how, how no one could. I remember when I got the first iPhone, it was a gift from John Doer, and I was like, Ugh, this is 800 bucks.
[00:40:56] That’s nice of John. I, I kind of like my Blackberry. I don’t wanna type on [00:41:00] this like weird screen and what are these things called? Apps. I’m not gonna pay 10 bucks for month for this thing. Like, this is crazy. I’m happy with my Blackberry, so I didn’t even use it basically the first six months. And then I was, you know, you had a lot of people who would be freaked out about putting their personal information, their banking information on the phone.
[00:41:18] Just think about it, in 10 years you’re doing everything. I mean, my entire net worth people could managing their net worth, their homes on their phone. There’s no way. I mean, and a, and so the one thing I can tell you is every expert study is gonna be wrong, right? You look at every expert study and they said, you know, McKinsey, God knows how much money they got paid to say how many cell phones they’d be in the world.
[00:41:42] Or how the forecast will be wrong, a hundred percent wrong. And, and the que only question is how wrong will it be? One order of magnitude, four orders of magnitude. I mean, remember when the camera phone came out and they were like, I. You know how many people are actually gonna pay for a camera on their phone?
[00:41:58] Well, now we all have two cameras on [00:42:00] our phone, right? They have a video camera and like, do you remember when you’d run out of, I mean, you’re not as old as I am, but I remember when you’d run out of memory, you have to go to this computer hardware store and go buy extra mi ram chips. I mean, And those, and I, you know, I’m not that old, you know, it’s amazing how fast this in.
[00:42:17] Yeah. 15 years since the iPhone, right? It’s, there’s no way. So I think that the only lesson is don’t, don’t limit the horizons. Don’t limit the the, because it’s just impossible to know. I like, I like your message before about, you know, your downside is one x your investment and your upside is unbounded.
[00:42:35] It’s a lesson we gotta remind ourselves. It’s, but it, it’s, so, you know, when, when the folks that work here, and I kind of pound that into them, you know, it’s just so unlocking, like, don’t do a three x investment because first of all, it’s impossible to know where there’s gonna be a three x investment. And the downside from three x to zero is.
[00:42:56] Is, it probably doesn’t make it interesting as a venture investment cuz you’re [00:43:00] not gonna have that kind of predictability. It must, the, the consequences of success must be material otherwise in our business you’re gonna lose. Love it. Final two here, Samir. What do you know you need to get better at? So the thing I do worst.
[00:43:19] In my job by far is I get too passionate and too emotional about my companies, and I think my return, my returns would be much better if I knew when to give up on a company sooner. I’m just not good at that, but I’m not that apologetic about that either. I think that also works both ways, but I think that’s, that’s one of the things that is, is, is trickier for me.
[00:43:42] And then finally here, Samir, what’s the best way for listeners to connect with you and follow along with Kla? I think Shenaz might be able to answer that, but I think you know, you can. You can always email me@skcoastalventures.com and our website, our Twitter handle seems is pretty active and and current.
[00:43:58] But Nick, I [00:44:00] wanna thank you. This is a lot of fun. I really appreciate you taking time to spend with us today. Well thank you Samir. He is Samir call. The firm is Kla. You know, congratulations on all the success. You’ve been at this a long time and I appreciate you spending the time to share some of your, your insights and your wisdom with us all today.
[00:44:15] So thank you Samir. Thanks. Thanks.
[00:44:22] All right. That’ll wrap up today’s interview. If you enjoyed the episode or a previous one, let the guest know about it. Share your thoughts on social or shoot ’em an email. Let ’em know what particularly resonated with you. I can’t tell you how much I appreciate that some of the smartest folks in venture are willing to take the time and share their insights with us.
[00:44:44] If you feel the same, a compliment goes a long way. Okay, that’s a wrap for today. Until next time, remember to over prepare, choose carefully and invest confidently. Thanks so much for listening.[00:45:00]