The VC case for 'full stack deeptech'

Latitude Media covering the new. The energy transition. I'm Sheo Kahn, and this is Catalyst. If you make progress quickly, great. Makes it easier to raise your next round and therefore do all the other things you need to do. But if you don't, The lack of momentum will kill you. I would rather need more money, but make time my friend than treat them as pure trade-offs. Coming up, an ode to the full stack deep tech startup.

Catalyst is supported by Fishtank PR, an award-winning PR firm focused on climate and energy tech, renewables, and sustainability. Fishtank is known for generating prominent and effective media coverage for the brands they work with. If you want a PR partner that's thoughtful, shoots straight, and gets results, you'll like Fishtank PR. To learn more about Fishtank's approach, visit fishtankpr.com. That's f I-s-c-h fishtankpr.com.

When utilities need flexible capacity they can count on, they turn to energy hubs. Energy Hub works with more than 170 utilities, coordinating over 2.5 million devices to manage 3.4 gigawatts of flexibility built for the moments when utilities can't afford uncertainty.

Energy Hub builds and operates virtual power plants that utilities actually stake their grid planning on, coordinating EVs, batteries, thermostats, and more through a single platform built for utility scale, predictive, verifiable, and designed to perform when it counts. Learn more at energyhub.com. I'm Shail Khan. I lead the early stage venture strategy at Energy Impact Partners. Welcome. All right, so this one is a little out of the ordinary for me for two reasons.

First, despite the fact that I am a venture capital investor, I actually generally don't have other VCs as guests on this podcast, at least not very often. I prefer usually to talk to operators and researchers who have a much deeper knowledge than anybody in my shoes. Second, I'm not generally in the habit of having a whole conversation based on a tweet or an ex post or whatever you want to call it.

But I'm in the business of exceptions. So here's one. My friend Ian Roundtree is also an early stage investor. He leads Cantos, which is a pre-seed and seed fund that's been doing deep tech since long before it was cool. I think he may have been part of coining the term deep tech and is now part of coining whatever the next term is gonna be. Anyway, Ian posted something about his investment thesis just a few months ago that has been basically stuck in my brain ever since then.

So he said, and I quote, we invest in two archetypes at Cantos. One, full stack deep tech selling an end product or even commodity, not selling technology, or two, weird end of one, never seen anything like it before. Selling technology to incumbents, pass. Commercializing science, pass. Nth company doing the current thing.

I've been hung up on it because it resonates so strongly with my experience over the past now 19 years working with startups in the energy and industrial world, first as an analyst and now for the past eight years or so as an investor. I think it's correct. But it also describes a world that honestly excludes the vast majority of startups in the space. So it's worth unpacking. Thanks, Shale. I'm excited to finally have you on after many times of having podcast worthy conversations with you.

Not in front of a microphone. Um, and I rarely do this, but I wanna talk about a tweet, an ex post, whatever, that you posted uh a few months ago and have pinned since then. So presumably you like it as well. But it just like has stuck in my head ever since then. And I just I I could not agree more strongly with it. I think it actually I wanna unpack it because I think it contains some insights into like

the types of companies that really seem to succeed, startups anyway, within this world of deep tech, hard tech, whatever the hell we want to call it. You list out the two types of companies that you do look forward to invest in. And we're going to talk about those two types of companies in a minute.

But after doing that, you list out three types of companies that you pass on that you don't invest in. I want to talk about that first, actually. Um, so the first one is companies selling technology to incumbents, which I I wanna harp on because In these deep and hard tech categories, in industrial categories, let's say.

Um, I feel like that's the majority of companies actually. The majority of startups have a technology that they wish to deploy and the markets that they are entering are comprised of big incumbents who control a lot of the infrastructure. And the distribution and so on. And so the obvious thing to do is try to sell them the technology. And so you do see a ton of that. Tell me if you disagree. Why is that like um a blanket pass for you?

Well, I've learned this like many things the hard way. You know, I I started cantos, it'll be ten years in this spring. Um and the It seems like the easy thing to do because it it scopes down your startup to, oh, we just we're gonna use this amazing technology and we're gonna like, you know, productize it in the simplest way, and that'll make it easy to sell to

said big customer. Um and then it turns out that there's just so much institutional inertia, sometimes cultural, sometimes it's actual switching costs of having to rip out whatever they're using now to move to your product. Such that it ends up taking a lot longer than you think. Um, and this has implications for capital intensity too, but you know, we we can we can address that later. It's it's more that It slows you down when speed is so critical to a startup.

I think speed is actually the key. Point about this, because you know, the there's another piece that's not described in here, which people talk about a lot, which is like, okay, your thing, whatever your thing is. Especially if it's gonna displace some current thing, has to be 10x better, whatever that means. Yeah. Faster, cheaper, et cetera. So but let's say you do have a thing that is 10x better in whatever way, you would think.

Great, it's ten X better than the current thing. I'm gonna go sell it to whoever buys the current thing and uses the current thing and I should win because it is that much better. And I think in the long arc of history, if you had infinite time and infinite cash to burn to get there, you could win that way by selling to incumbents. Uh, you can't if you are time-limited on startups.

Speed. Yeah, and and there I mean to like add color to that, the the main reason that startups are uh time limited is because you raise serially in like, you know. Whatever you raise 18, 24 months of runway. And so let's say you raise 24 months of runway. You need to start fundraising, let's say six months before you run out of money. So you have 18 months.

To do whatever you need to do to unlock the next series of fundraising. And if your customer's sales cycle is longer than 18 months, you are de facto debt. And so you just you have to make progress faster. And you know, we could get you know this is maybe a problem with venture capital as a whole, but the game on the field is you need to make discernible progress within 18 months or you don't raise your next round.

And a lot of times sales cycles do take longer. And there are there are understandable reasons for this. You know, the the stakes are very high. Um, in in some industries, lives could be on the line. The customer wants to do their diligence and they're using something Good enough today, question mark. And you know, maybe they go golfing with the sales rep from that company, and you have to overcome all this.

to get your product in there and start to scale enough such that you can raise the next round and you get to live to fight another day. And one of my very first investments at Kantos in 2016 was a startup that had incredible machine learning and interoperability software that it was selling into the automotive industry, into industrial robots.

absolutely best in class. And we would hear this time and time again from their reps at Big Automakers. And the company died despite having some flashy investors and incredible talent behind it because Ford and Toyota's sales cycles were too long to raise the next round.

Yeah, I mean the auto OEMs are like notorious for having the longest sales cycles of of anybody, even for stuff that is not right. That's it, that's software. You think it'd be faster sales cycles there. I mean the other dynamic too, before we so without spending too much time on this piece is um is is pilot purgatory, right? Like big companies love to pilot something.

And the timeline between pilot and full commercial rollout is never certain. It's not well defined in general. And so again, just from a time perspective, you get They're literally different teams. Like I always encourage entrepreneurs to understand the counterparty's incentives. And if they're Like whenever someone's behavior i in a business context doesn't make sense to me, I try to learn how they're paid. And usually the answer

It comes to the fore that way. Um, and so if you can just understand people's incentives, you can navigate around that and sometimes quicken the sales cycle or assess whether you are pursuing a sales cycle that is uh compatible with your startup in the first place. I will say I've learned this lesson the hard way as well in, you know, big industrial categories like the chemical industry and the mining industry and so on.

It's easier said than done to not do this. So we're gonna get to the flip side of this, which is like if you if you're not selling your technology to incumbents, then you have to be full stack and that has its own challenges. But I think you and I are aligned that like, It despite those challenges, it may be the only way to succeed at a venture pace and a venture scale.

Yeah well I mean spe speed is I would argue the most important thing, but there's also like ultimate value capture and there have been a bunch of cases whether it's selling some, you know, amazing chemistry into the battery supply chain or um selling really advanced software built for an industry or a a subsystem that goes into a larger plant or something where you have objectively superior technology. And let's, you know,

say it's 10x better than whatever there is today. Um and they acknowledge that this is great and they want it. And they give you an offer to pay something that is way lower than you think it's worth. And when you push back they say Sorry. Yeah. Yeah. Yeah. Like if you're just not in the pole position, then you're you're not often going to get remunerated for the value you're creating. Yeah. I mean, speaking less to the value capture and more to the timeline thing, I think I'm thinking about um

A company like Seal andano, which is run by a friend of mine, Gene Burzhewski. I'm not an investor there, but um close to Gene, who's amazing and like a total rock star CEO. You know, they're doing silicon anode battery materials, and uh it is clearly better.

Everybody I think agrees it is clearly better. And they've had to go, they're selling to auto amps predominantly. And it's been and they're getting there now, but it's been, I don't know, 13 years or something like that. And like it's a very rare founder who can keep a company alive and funded that long for something like that. Yeah. Definitely.

Okay, let's let's move on. So selling selling technology to incumbents is hard for all the reasons we described. Second category that you set is a pass, commercializing science. I guess I want to hear what you mean specifically by that. Um

Yeah, I d I mean with w without getting to the semantics of like, do we call it hard tech or deep tech or frontier tech or whatever? Um Uh there is a category of company that I alluded to before, which is you have incredible technology and you think therefore it is incredibly valuable.

But that's not actually how business works. Like the only point of a startup or a business in general is to create profit margins above and beyond its cost of capital. And technology can be a very interesting means to that end, but it is only a means to that end. And just because your technology is mind blowing and you know is cited in a bunch of papers or whatever, does not necessarily mean it is valuable in an economic sense. Um, and so if you are like

You know, commercializing your PhD, for instance, you can often look like you have a hammer and you're just looking for nails. Whereas in a business context, you have to reverse it. And you kinda have to be obsessed with a problem and completely agnostic as to the solution.

Um and I find that there are a lot of deep tech founders, especially PhD entrepreneurs, um, and and and particularly when they are commercializing their own thesis, then They they kind of have it backwards and and they're waiting for people to tell them how off awesome their technology is when really you have to do a little more work to go show it.

Um and so that's part of it. The other part is the timelines where I would much rather have a novel application of a existing or relatively new technology rather than wait for something to become market ready in the first place and Yeah, I've lost a lot of money just like waiting for a startup to advance through Technology readiness level. So I feel like there's two pieces there to what you're describing.

One piece is the is this a hammer in search of a nail kind of a problem? It's like somebody who's attached, they're they're personally attached to a particular technology that it turns out actually might not be like product market fit is elusive. And then the other is more like a TRL scale for lack of a better term problem of like if you're investing in like

break through science that hasn't been sufficiently proven, then it just takes too long to get there. Let's just spend a minute on each of those. Um

On that second one, on the like, it takes too long to get there. How do you about how do you think about things like fusion, for example, nuclear fusion, right? Which is like anybody who's investing early in nuclear fusion, and even if they've they were doing that ten years ago or today, like it's it's sort of inherently Breakthrough science that hasn't been commercial.

I don't know. The short answer is I don't. Yeah. That's just not the general thing I would do. I mean I used to spend a lot of time thinking about fusion. I would like, you know, do deb I'm not technical at all, but like interested enough that I would like debate people on, you know, is it is it DT or is it PB11 or whatever? Like, you know, I I I loved like getting wonky like that, but it turns out uh i in in my business context.

Um, I don't have the time to wait for that to become commercializable. And there are some very interesting initiatives to bring Fusion Energy about. It it is it is expressly the type of science that we at Kantos do not invest in. Yeah. Although I... I really hope for humanity's sake, uh, we pull it off. And then on that first one, the sort of hammer in search of a nail thing, I I also have seen

I've seen that a lot. Like one I sort of somewhat recent example for me is we we ended up um incubating a company called Voya Energy, which is still kind of in stealth but but is known to the public. And they're they're doing they're using m metals as fuel basically, um, in a particular way. And we kept running across we we were look as we were incubating it, we were working with the founders on it. Um

you know, we were taking we're doing an electrochemical approach. Um, and we kept looking around and finding a few other companies. We're all using combustion. And I was trying to figure out, and it seems clearly worse to me for a variety of reasons to do that. And so we were asking ourselves a question: like, are we missing something? Why? Why is that the way that everybody is doing it? And a lot of it comes back to there's a particular professor who is a professor of combustion who has.

been the one to create the diaspora that exists such as it is today around this particular area. So it's exactly that thing of like, this is a combustion person. That is what they do. Um, and so of course that is what they are going to spin out of their lab, and that's where the startups are going to come from and the researchers and so on. Do you see this a lot? That problem in like, you do a lot of stuff in bio world. Is the technology and solution in search of a

problem statement a bio thing often? Oh. Oh man. Um vr I should start by saying we're doing a lot less of it now, haven't we? Times of difficulty. Yeah. But um we we do we do very much believe that like in our lifetimes the intersection of computation and biology will probably be one of the more prolific areas of innovation. Um it it has been

Tough to invest there for a while for a variety of reasons. But yes, there's a lot of this like commercializing science. Now that is an area, interestingly, where The capital markets have really figured out how to underwrite to l let's say technology readiness levels.

Yes, I find this to be frustrating not being in the pharma world, right? Because I look at the pharma world and I'm like, oh, like drug the the process of taking a drug from very, very early stage through to the market is like really well understood. The financing seems really straightforward. Why can't we just like replicate that across a variety of different markets? It just doesn't seem to work the same way.

It's just, it's not standard enough. I mean, I think I've thought a lot about this. It's just not standard enough. Like you need enough. In that industry, you can kind of look at, okay, well, we're in we had this kind of readout in our phase run one trial and we're in this indication. We know this indication has this many patients, and usually you pay this much for it, and we can work out like

you know, big pharma when they acquire something, they've like run the discounted cash flow analysis on it. Not not like based on technical milestones. And it's just it's not we don't have a process that is as scientifically and regulatorily um defined and widespread enough that an entire capital market can develop around it. Now, that said, even though there are dedicated investors for biotechnologies,

It has been a really tough area for public and private companies. For a while, like a third of publicly traded biotechs were trading below cash um because they just couldn't, if if you didn't have a team that understood this, then like a broader universe of investors couldn't access that asset. And So even when you have very developed capital markets and dedicated funds that have been performant over decades, it's sometimes still not enough to weather certain storms. But I still think

In principle, at least like leaning in this direction on some other deep tech stuff would be interesting. You see this occasionally in other places, not to bring up fusion too much, but one that comes to mind to me is Pacific Fusion, which raised a quote unquote billion dollars out of the gate, but it's actually like a staged series of capital infusions contingent on milestones. It's like a billion dollars lined up if they can achieve

X, Y, and Z over time. And I like that concept. I want to see that. I mean, you're an investor in like Saw Yugen, right? Um If you had looked early days to what Solugen was doing, so this is in the chemicals industry, could you have lined up a series of technical milestones?

ahead of time. Like could you have done this a priori and then said, okay, we're going to sort of pre-define the capital roadmap here? Or was it just too much uncertainty? Were unit economics too much of an open question? Like what stops us from doing that? Yeah, I mean well in in in Soul Agent's case it was special because You know, we we met at Y Combinator Demo Day and they had like max out their credit cards to buy fifteen thousand dollars worth of

Home Depot components and had a small bioreactor using their technology and were making hydrogen peroxide at at that time and selling to like float spas in the area. Like, you know, they they had thousands of dollars of revenue and it was small, but you could at least like they were already commercial. Yeah.

So it's just it's just been scale up. Like that I I would define, you know, they they had this cool science out of their their PhD work at um at MIT. And then it was just like scale up from there. It was more engineering. after the seed round. Um, and we could we could do the T E A and we knew that if they got to a certain scale it would be this profitable and then there were some

question marks around which chemicals can we expand to and like how much of this is going to be specialty versus commodity. But um that was a little easier than like, trust me, trust me, I just need 200 million dollars and then we'll make money.

We see a lot of. Um and occasionally works, right? It has. I mean, to be clear, like I um my opinions come from my own experience and are in the context of like my firm and my background and my capital base. It doesn't mean it's the only way to make money. Um okay so that that's two sort of tough.

Categories, selling technology incumbents, commercializing science. The third one I don't think we need to talk about a whole lot because it's kind of self-evident, but you said like nth company doing the current thing. I share that view as well. It's just It's much, much more difficult for a variety of reasons to get a venture grade outcome if you are in a really crowded market. Doesn't mean like some somebody breaks out of those crowded markets sometimes.

But the bar is so, so much higher for you. It's harder to raise capital. to attract talent, it's harder to to separate yourself from the pack, basically.

Well this gets this gets back into the commercializing science thing a little, which is um th th there's a type of startup that I have invested in in the past that's like Well here's all the technical reasons why our technical We're the best because X, Y, and Z. Like we could create a chart and it's going to show all of our competitors and we have the full X and they have the partial. Yeah. But um sometimes you do legitimately have something that is better than the incumbents, like um

you know, i in uh I'll I'll pick on one of our one of our own portfolio companies. We invested in a a company that's doing um amazing processing of right radar technology, like mind blowing can take off the shelf data from your run of the mill radar and resolve more or less a three dimensional image. Um they they call an RF camera, like mind blowing. cheaper than a camera, can see through fog, in some cases see through walls, put this thing on cars, on drones, has has

so many implications. Um, and there's, you know, a a much better funded company incubated by A B C called Chaos Industries, which um I personally think has worse technology, but they're doing a lot better because they've productized this. They're very good at selling it. They're very good at marketing it. They they have a lot more capital and they're landing contracts with the government left and right.

Um, and you know, my hope is because we've invested in a much better technology, we can catch up. But like, I've seen this movie before and Sometimes having the best technology doesn't win you right. It gets back to the old trope adventure of right, like a combination of team tem tech and timing. But like you can have the best tech and you don't have the other things going for you.

It isn't it isn't sufficient. It may be necessary. It's not actually, I don't think it's always necessary either. It depends on the situation. It's certainly not sufficient. Yeah. Well it gets into like the the intersubjective nature of capital markets where like you can be objectively. Right. But if everybody else is wrong for long enough, you are de facto wrong. And they are de facto right. And so you have to account for it's it's tempting for

th self-styled intellectuals to sit and think that like, oh, I've just outsmarted everybody. But Like the the market can afford to be wrong longer than you can afford to be right. Yeah, that's the other point, right? Like how how what does it take to prove your rightness? How long, how much money, etc.? I think back on um You know, the the heady days of all the solar uh technologies and the synfilm technologies predominantly in the like late. Two thousands, early twenty tens.

And um, you know, the the I think the common story that everybody tells about what happened there is that You know, VCs invested billions of dollars into a bunch of different synfilm technologies. And then meanwhile, China scaled up crystal and silicon, which is the dominant technology and cost got so cheap that Uh there was no way to compete. And that that's true. It is also true that in principle, some of those thin film technologies could be cheap.

Right? Like they could still be cheaper then. uh crystal and silicon is even at a silver cost today. But to get to the point where they would be cheaper requires a lot of scale and a lot of capital. And as crystal and silicon prices were crashing, nobody was willing to fund that effort.

And so, you know, you you may have been right, at least some of those folks may have been right. Um, but yeah, it means nothing, essentially, in the in in the grand scheme of things, because they were never never able to prove it. Um Yeah, I I used to kind of think that you could like if you had an amazing enough technology, you could you could sustain years of no commercial progress, assuming you could fund it.

such that um because you were taking no market risk, i.e. it was, you know, patently obvious that your technology was better, um Once you came to market, you could jump up the commercialization curve faster. And rather than needing to start small, you could just start with a giant contract because your stuff is so awesome. But it it one of the things I didn't see. There is the the learning curve matters.

And once you start doing a thing, you tend to get better at it. And if you're commercializing something at small scale, you're kind of working out all the kinks that if you're if you're, you know, sitting in your lab making your technology better. then you bring it to market, you only get to learn then and it's kinda like Too late?

And so I I like I actually like startups that start small and they just get that little compounding and learning every little like sales motion, getting to know your customer better and internal operations and how does technology hand off with with manufacturing and and closing that um design for manufacturing loop and all those little things that like if you if you're just developing the technology, you have to like punt them.

And then do them all later and it's maybe too late and you've got all this technical debt or something. Um, this compounding is um a wonderful thing. Are you tired of overpaying for big name PR firms but not really knowing what they're delivering? Is your comms team wasting time reviewing lengthy messaging briefs and decks?

Instead of engaging journalists or producing content, are you wondering why your competitors are getting pressed and you aren't? Fishtink PR is an award winning climate and energy tech, renewables and sustainability focused PR firm

Dedicated to elevating the work of both early stage and established companies. Whether you need to position yourself as a thought leader in between project announcements or translate complex ideas and technologies into tangible, compelling stories that resonate with the media, Fishtank can help. Check out fishtankpr.com. That's f-i-s-c-h fishtankpr.com.

Virtual power plants are becoming a reliable way for utilities to manage capacity, but enrolling devices is just the start. What really matters is confidence, knowing those resources will perform when dispatched. and being able to prove it from the control room to the living room. Energy Hub's platform handles the full picture, from near real-time forecasting, locational dispatch, and the kind of rigorous verification that holds up when regulators, grid operators, or leadership asks,

Did it deliver? Easy enrollment creates momentum, proven performance builds trust. That's why more than 170 utilities rely on EnergyHub to manage over 2.5 million devices, delivering 3.4 gigawatts of flexible capacity. See what that looks like at energyhub.com.

All right, so let's talk about what to do. We've mostly been talking about what not to do. So you you have two archetypes here. I want to spend most of our time on the first one, which you described as full stack deep tech selling an end product or even commodity, not technology. So say more. What is full stack?

Um so so when I say full stack, I'm subtweeting a uh twenty fifteen post from Chris Dixon at Entries in Horowitz, who's largely been a crypto investor, but um back in twenty fifteen he uh described a full stack startup and was um largely alluding to companies like Uber and Lyft and Airbnb, who rather than trying to sell software

to the taxi and limousine industry or to the hotel industry, we're were leveraging software to influence the real world while not necessarily building hardware themselves. And I thought that was really interesting and largely apply that same thinking to more industries and the types of technologies that we invest in where you are building some software, but also some physical components.

There's nuance to this, right? I mean it's it's sort of a similar full stack is similar to vertical immigration, right? Like it but it's not exactly the same. It it there's there's heavy overlap, but it it it connotes the combination of technology and non tech product or services specifically. Um

I think the key thing for me actually more than like defying full stack is the selling an end product or even a commodity, not selling technology. Whatever the thing is that your thing unlocks, like go if I'm vertically integrating in a direction, it's downstream. To start. At least. Right, sell to the end customer the thing, whatever your technology makes better, figure out what the end custom who the end customer is for the thing that that made better and then sell that product. Um

But that is expensive. Like let's be clear about the trade-off there, right? Like, cause I think it is, it is, it's like definitionally more capital intensive to do that. Do you disagree with that?

No, there is this is clearly the the downside of um of going full stack is that you you typically need a little more money to do it. Um And like the time is money adage um uh it is appropriate, but there's a multiplier on time, as we talked about earlier, for startups where there's such an importance around making progress quickly when layering in the intersubjective nature of capital markets that like if you fail

Like if you if you make progress quickly, great. Makes it easier to raise your next round and therefore do all the other things you need to do. But if you don't, The lack of momentum will kill you because it'll become multiplicatively harder to raise capital the less. progress you make. I would rather need more money, but make time my friend. Treat them as pure trade offs.

Can you give me like a canonical example for you of like a a company either you are or are not involved with that is full stack in a category where they could otherwise be selling tech to incumbents? Yeah, so I mean

A great example is mining. Um, and there's a few companies in that in this space. Full disclosure, one of our largest investments is Earth AI. Um, uh, but you also have Kobold, which is rates a lot of money, and Vera I and there there's a couple of others in this space. So Mining is an industry that is massive and there were just not that many startups in.

And I thought that was interesting because it's such an important industry. This is definitionally a commodity industry. Um, and it is increasingly important for semiconductors, for, for defense tech, for um, for uh electrification of the grid, yet we're not seeing much innovation there. There had been some startups along the way that said, Hey, we're gonna use satellite imagery, AI to help

people mine better and know if there's a deposit near your existing mine or something like that. And um it turns out it was very hard to sell that. technology into mining companies and and two, uh they weren't willing to pay much for it. And so uh the company we invested in, Earth AI, um, and I and I believe Kobold may have had a similar journey, but I although I won't speak out of turn. Um I said, Well, hold on, like How much is that?

Is it to acquire these mineral rights? How much is it a drill? Like, let's just hire some geologists and go apply for the rights. Um, Earth AI went out and bought rights and drilled their own hypotheses and we now own deposits.

that uh we think are very valuable. There's a lot more work to be done to to prove them out. Uh you've seen Kobold go out and acquire a deposit. They're using AI to um uh maybe find and definitely mine a little more efficiently. Um And this is one where like if you had taken the typical startup path of just selling software.

Maybe you make a software license, but if you are willing to go out and take a little more operational risk and need to raise a little more capital to be sure, then you can end up owning little literal gold in the ground. And that's a lot more interesting to me as an investor than, you know, a mining SaaS company.

The other category that I think uh of as you describe this, that I think it's equally applicable to is the like AI for materials discovery category, where there's a bunch of companies doing that now. We're like My question to every one of them when I've talked to them is like, okay, let's say you can you build your magic model and you can use that model to discover novel materials. Are you gonna license the model and the technology to a big company? Are you gonna

try to, you know, sell the IP around the material you discover? Or are you gonna do what you probably should do, which is make the material? Um, whatever you discovered, if it's better. You should just make and sell that thing. And that ultimately that's your business. You're not an AI discovery company necessarily. You're an AI enabled materials company. And some of them are doing that. But it's kind of the same story as I think you're describing in the mining industry.

Yeah, totally. I mean the one of the one of the um uh the initial thread that I started pulling on that led me to my preference for full stack hard tech startups was I invested in a um a construction software business that was doing um AI for planning. Um, and there were a couple of these out there and we were sort of selling against some of those other companies and the general contractors are really slow to adopt. And I was like,

Couldn't we just like buy a failing general contractor? Like just do this better. Um and and it turns out it's not just the technology that that improves. Every single time we've invested in a capable team or watched a capable team go after an industry with a lot of inertia, um, there are a thousand little things that can be improved upon that if you're not taking that on and you're just selling into the industry, you're not going to help innovate those areas.

You mentioned even commodity thing. Let's talk about commodities for a second. A lot of the world that I treated, I'm like, you know, focus in energy and industrials. And and like when you boil that world, energy in particular, down to it. core constituent parts. It's mostly commodities, right? electrons or MMBTUs or whatever. Um and notoriously commodities are difficult markets. Commodity sectors are difficult markets to build startups in. That's like a

It's like a trope. Um, I I think there are a bunch of ways that you can do it, but you you said in here even commodities. So like what is it that makes a commod a company who is gonna end up if they go full stack, they integrate downstream, they're gonna sell a commodity. What is it, what does it take for that to be attractive to you? Yeah, I've I love I love commodities. Like for for a long time it was used as a pejorative in venture capital. I I love when a company is selling a commodity.

When the technology that is embedded in their full stack business actually gives them a cost advantage, um, that's incredibly powerful. And and taking a step back, like what we do as investors is in a way. Price rip. And so if we're thinking about the types of risk that a startup takes, there's there's technical risk, there's execution risk, and there's market risk.

and you you always have execution risk. So let's focus on the other two. In hard tech and deep tech in the physical world, you're typically taking a bit more technical risk than some of our peers who focus on maybe software and AI applications. Um and n not foundation models, a lot of technical risk there, of course, but um If we're taking more technical risk and we are on balance

Basically paying the same prices at seed, series A, series B, whatever, as companies that are not taking more technical risk. Um, then unless you are offsetting that with a equivalent decrease in market risk. You're probably just going to make worse investments. Like those these are the the physics of finance.

Right. And so why I love commodities is because it offsets the technical risk we're taking with less market risk. Because I know if I'm selling a commodity, i.e. my product is molecularly identical. to the competitors, but I have a cost advantage. I know you can sell that. I love when like in my memo I can just look at the spot price of something on a liquid market. And I and I know that's that's

the the mark that we're shooting for. Whereas If you're um if you're developing something new, um, or you're having to convince a customer that they they need it, um, even if you know it's objectively better, then there is inherently market risk there, and that just makes it a Riskier asset.

I've heard Vinod Kosla say something very similar, like he'll take technical risk, but he won't take market risk. The other thing that I would say here about commodities is like I think Yeah, but he invested in impossible foods, forgetting that there was market risk there.

Yeah. I mean many many things have I think nothing is as simple as that makes it sound, right? Like there's market risk where you think there's not, but also commodities are not commodities in the way that you think that they are. Like I mentioned that that like electrons are commodities. They're kinda not, right? Like, yeah, an electron is is a fungible thing and and it's similar to every other electron, except where and when it is delivered is very, very important.

And so there are lots of non-commodity businesses built on selling electrons because they're able to sell electrons in the right place at the right time or both. And I think that's true of of most commodity markets, right? Like very few of them. are truly like indifferent to time and place. And that additional factor matters in terms of whether something whether it's truly like a knife fight to the bottom on cost, which is why people don't like commodities from a from an investment perspective.

I don't wanna go on too much of a tangent, but like there's an interesting corollary here to the type of real world full stack business that that we both like. Um and sort of decentralization of infrastructure. Like if there is advantage to um defy economies of scale and co-locate your your product, your, you know, behind the meter.

energy generation, whatever. Um then you are you are operating your own sub facility, but you're still taking some market risk by virtue of like customer concentration, you know, because you can presumably as a startup only afford to have so many of these co-located things and um you know if

they're with the same two or three customers and they change their mind or someone gets fired and the new gal doesn't like you or whatever. Um, that's risky. So, um I've I've invested in some of those, but you just have to be clear about like there's you know, whether the counterparty is is gonna slow your sales cycle down. Yeah. There are also interesting examples that are like companies in a commodity space with a differentiated input.

And or companies in commodity space that like their execution and ability to learn in that market allows them to enter into a new market that's non-commodity. So let's let's take a look at it. Oh yeah, give an example. Crusoe, right? So Crusoe was flare gas to Bitcoin. That was that was the start of the company. They they went and locked up assets that were methane flaring.

Wasted energy, they turned that, they, you know, put little data centers there and mine. Bitcoin. That's that's a commodity, right? Bitcoin's clearly a commodity. Um, and they did that a whole bunch. And then that allowed them to sort of like build up more expertise. They built they started going beyond Bitcoin and into cloud and they had their own little cloud managed services thing.

And they had all this experience building small data centers and then like the right place at the right time came around and to give them credit, like they fully went after it and took full advantage of it. And then now they're building You know, Abilene's Stargate campus for OpenAI and Oracle and a bunch of others, and they just raise money at a$10 billion valuation. Now they're a totally different company.

Yeah. I think Corweaves started mining crypto too. I don't know if it was because they were flaring but No, it's no it's different. I mean CoreWave was just was just doing crypto mining. A lot of crypto mining companies are now, you know, AI data center companies. That that pivot is not unique to Crusoe, but I think Everything is computer.

Yeah. But they've been able to particularly take advantage of it. What's interesting to me about it is like they were they were certainly in a commodity business. Um, they had a differentiate the original version of the business had a differentiated input, right? It was flare gas was their input to create Bitcoin and that's cheap energy. Right. If you could take advantage of it.

I mean if you have you have a cost you have a structural advantage in in producing a commodity that's like the holy grail. Right. Right. Like if I if I can make dollar bills for 70 cents. I'm a trillionaire.

So I have a different frame that I often use to describe to categorize startups. Um, and my version of it is wave makers and wave riders. So like there are companies that uh make a wave, they do a thing that but for the existence of that company, the would not have happened, certainly would not have happened in that.

timeframe or they're wave riders, right? They correctly predict a trend a few years ahead of time and they time the development of their technology or their company correctly to that trend and then they're able to ride that wave.

you know, Tesla being the canonical wave maker, like the EV thing wouldn't have happened, at least on the timeframe that it has, but for Tesla. And then for every Tesla being the wave maker, there are hundreds of wave riders, EV charging companies, battery technology companies, et cetera, et cetera, right? And it's useful for him in the first place, but it's also valuable for me because I think the the

Um the onus on the you know what matters the most is different across those two categories. There's a certain type of particularly on team, there's a certain type of founder and founding team that is required in the wave maker category because it is so hard because you are swimming against the tide. Um and you have to bend the arc of the universe to meet your your needs.

I I think it'd be similar in the full stack deep tech concept, right? Like it's it's harder to do. You need more money. You have to do more things. Definitely. You right. And so it what do you how do you think about like the archetype of the founder that can do the full stack deep tech thing? Our

preference for this type of business has also increased the onus on the entrepreneurs. Um And, you know, when we when we thought that you just needed incredible technology and, you know, sort of good enough business sense, I was not um Not that we were lowering our bar for entrepreneurs, but I was sort of looking for different things and now we flesh it out, um actually have i internally like a way of of s scoring founders along.

six axes and how we tease out each of those. Um, and a lot of them are around the qualitative. It's sort of, you know, it's talent gravity, it's it's um narrative ability. And if you're going to raise the capital required to vertically integrate a business and go full stack, then you would just have to be that much better at fundraising. And I think you have to be really honest about the fact that, you know,

A a lot of people don't. Like the reason that I'm on this side of the table, not operating a company, is I know I don't have those qualities. Um, I have enough of them to run a small investment firm, but not to build a full stack business that's going to come at a giant advantaged incumbent. Um, and to do so, you probably need to raise you know, nine figures of capital, probably ultimately billions of capital, um, across equity and other forms of of capital. So the

Um, the type of person you need to make a wave is like you said, like, you know, bend the arc of time toward them. We'll we'll say that a founder has to have w we feel ourselves being physically pulled toward them in a meeting. Um, and if we don't get that sense, we sort of say, Okay, well maybe this isn't a a wave maker. Um and we've just found ourselves looking for people that are are both incredible technologists and also have this

uh ability to communicate things in in a hyperfluent way. They can kind of go up and down the stack of of understanding, communicate it to an expert, communicate it to the general audience, and get people so fired up that they Bend the fabric of space-time to their will.

All right, just to close out, the other category of things that you do look forward to invest in, which I also share, um you described as weird N of one, never seen anything like it before. So here's my question for you. Apart from those things, I mean I'm drawn to them too. But apart from just them being cool, like Well, how would you articulate why are the weird end of one never seen anything like it before ideas, good venture investments on balance?

to um c continue the astrophysics parallel. Like it if if you're in a um a area with other astronomical bodies. You're sort of like competing with their gravity to pull in talent and capital and and PR, um, then that's harder than if you are the the only body to reach some critical mass in an area with a lot of um, you know, a lot of s matter. Um and if you're like the, you know,

Yeah, we saw we saw a early stage company recently that I thought was weird and interesting. We ultimately did an invest where um a woman had had left Wall Street to build uh as she calls it, de beers for dinosaurs. And she's using, you know, um, some satellite imagery and and AI and a field team to go out and find dinosaur fossils, which have been selling for a higher and higher prices.

Um and I thought that was really interesting because I never heard I mean, I've been doing this for 10 years. I see like a thousand startups a year more. And I never s seen anything like this. And if it's an area that is sort of big and interesting enough, and I don't know if fossils are, I didn't go that deep, but um Then she's gonna build like the company. And when people think of dinosaur fossils, they're going to think of her company. Whereas if you've got, you know,

You and I have both invested in in long duration energy storage businesses. But if you've got like the 20th Eldes business, it's gonna be a little harder to stand out and the and the venture capitalists that you need to raise money from have probably made a bet already and a lot of them don't want to make a competing bet and it just complicates things.

Yeah, but I want to separate two things. I mean, there's a part of this is just the flip side of nth company doing the current thing, right? Like if you it's just First to a market, I guess would be another way to put it. That's a piece of it, but the other piece of what you describe is weird, never seen anything like it before, like it's something above and beyond just you are the first in this category.

It's I think to me, an enormous amount of startup success ultimately comes down to narrative. um how well the founder is able to construct the narrative that raises capital and attracts talent. But also just like if the narrative, if the story, if I can tell myself the story of what this company is trying to be and like my eyes light up. as I'm trying to do it, because it's like so interesting and so weird or so impactful if it would have such a big impact on the world.

I I'm drawn to it and that create that's a part of the magnet. that we were talking about before, that that, you know, can come from the founder, but can also come from the vision. Like I think of a company like Colossal Biosciences who's, you know, the tag I don't know anything about them apart from that they're trying to bring back the woolly mammoth. Like that, that's interesting. It's weird.

Yeah, it's cool. So yeah, I guess I just wanna I wanna clarify that it's not just being first. It's not being the soul body in that part of the universe. It is also about that part of the universe being Exciting, I guess. Right. It does and it's not valuable just because it's weird for us. Weird is a filter for might be unique, um, and and and unique in a valuable way. Look at this. defense tech wave. And now it's like the hottest area outside of AI.

Uh, and that was more or less created by Andril. And I'll give a lot of credit to Palmer, one of the four founders of that company, for like completely reversing this trend. Um and That is I don't know if I would have called Andrew. so big and has spun off all these other defense tech unicorns that um yeah when I first heard about it I thought it was weird. And and and weird sometimes means controversial to be clear. Like to to lean into these areas you you

necessarily are going to have to disagree with people. Um and if you're not comfortable doing that, then it it's going to be difficult to build a weird, unique end of one company.

Alright, and this was um unsurprisingly a lot of fun for me. Thank you again for joining. Of course, yeah, well this is just one of our normal conversations we hit record this time. That's right. Ian Roundtree is a partner and the founder of Kantos. This show is a production of Latitude Media. You can head over to latitudemedia.com for links to today's topic. Yeah.

Latitude is supported by Prelude Ventures. This episode was produced by Max Levinson, mixing and theme song by Sean Marquan. Stephen Lacey is our executive editor. I'm Shao Kahn.