Mike Schroepfer's Gigascale Capital raises $250M to back deep tech and physical world startups

cutting this short. Uh a little bit of technical difficulty, but thank you so much. I'm sure you'll back be back soon

together and good luck with the journey.

We'll talk to you soon.

Awesome. Cheers, guys. Great to meet you.

Let me tell you about MongoDB. What's the only thing faster than the AI market? Your business on MongoDB. Don't just build AI, own the data platform that powers it. And our next guest is already in the waiting room. We'll bring in Mike from Gigascale Capital. Mike, how you doing?

Doing awesome. How you guys doing?

Former CTO of Meta, by the way. Uh, thank you so much for taking the time to join us. Uh, tell us a little bit about yourself and the fund, the strategy. I have so many questions.

Yeah, so I've been in Silicon Valley 25 years, been an operator, joined Facebook Meta 2008, scaled everything from data centers to Oculus to Instagram to building the Facebook ad research lab.

I uh stepped away from that job. just everything just every important project there

a lot of stuff

a lot of stuff I mean with a lot of great people so I don't want to really take credit but um I I saw this this change happening in the world a couple years ago which was I thought we were moving from atoms or bits to atoms just like the physical world mattered again that is the marginal cost of software went to zero the thing that really mattered is how much stuff can we build and how are we going to build it and so I started a venture capital firm to sort of prove out the thesis I I you know want to prove it first by investing my own money to make sure sort of I knew it was working and then you know our company started working the question isn't do they want to buy it it's how fast can they make everything uh because they're making it sort of better faster cheaper than the way we did it for the last I don't know 150 years in a lot of cases and so we've brought on a set of institutional investors and raised more money and we're just getting going on this journey

how much did you raise

250

250 quarter bill quarter bill uh what what were some of the what were some of the standout companies uh in that first kind of like angel fund that you mentioned that that made you think you were on to something.

Well, there's a there's a bunch of them. So, Panalasa is building ocean data centers for inference. I think you had G on here a little while ago um and uh is one example. Um Heron Power, which is a more recent investment, but it's just on a tear. They're they're saying like, "Hey, the way we brought energy to data centers is using 100-y old technology. Literally, we've got a new idea. We're taking the power electronics that are in your Model 3, Model Y, and we're putting them on the grid and we can deliver you power faster, better, cheaper than than anyone else out there. They're shipping products next year. Um, you know, we've got Radiant uh nuclear is building a, you know, container size micro reactor gives you a megawatt and a half of power for 5 years with no refueling. And they've got Form Energy that said, "Hey, you know, these data centers, you you have power swings, you might have renewables which are on and off. We're going to build a new kind of battery that lasts 100 days. So, if you're dependent on solar and it's cloudy for 3 days, no problem. I got 4 day battery. Um, and that battery is super cheap cuz it's made out of it literally like rusts and unrusts iron. It's like the basic chemistry of it. It's the cheapest possible material you could use to build a battery. Um, it's big and it's heavy, but who cares because it's on the grid. So, again, you're just like rethinking problems in a different way and using modern technology and you come up with a solution that that works for everyone. How are you thinking about publiclix versus privates? Because there's this massive boom and I'm so excited about Doug and Radiant and nuclear and bringing new solar on the grid. And there's so many different uh key technologies that it feels like we're in this crazy moment where like it's just hard to pull forward these foundational technology efforts in the private markets and startup world where yeah you're doing something completely new. it might take 5 years. And so in the meantime, it's just call up G Vernova and and 10x the amount of, you know, natural gas. And that that is like a technical debt that I think we're going to have to pay at some point. And I'm wondering about like is there a crowding out effect? H how do you stay sane if you're a startup founder who's doing something really cool, but you're watching some public company go 50x in the public markets? And uh and I'm just wondering about like where like the is the time compression is the time scale uh a factor one way or another. Is it a benefit or curse or both? Double-edged sword. How are you thinking about that?

Yeah. I mean I fundamentally left a big platform to bet on startups. Yeah. So I think you're going to you're going to get my belief that it has never been a better time to startups. And I'm seeing this this firsthand. I mean you think about the speed at which I can build tools um you know using AI using robotics in the physical world um you know even just using platforms like stripe and ramp and like all these it's just like the amount of time you said monkeying with things before is just like free you can have these tiny 6 10 12 person teams have incredible um unbelievable impact and you throw off all the all the you know dead weight of doing it the way you know you've done it even for the last 10 years. you started a company 6 months ago, you would you would form it differently than if you started it 12 months ago because like that's that's prior to you know 47. Um and it's just just an incredible time and we haven't talked about this just like a immense well of talent coming out of SpaceX and Tesla who are just like showing up to these old problems and saying like oh we know how to do this like 10x better, faster, cheaper starting with with new technology. So, you know, there's this analogy recently about when electricity first came to factories. You know, we used to build factories with like a central like steam thing that would spin all these like shafts and you built your entire factory around the structure. When they first electrified, they just rip out the steam thing and put electric motor, but everything else was the same. And productivity gains didn't really happen until you just rebuilt the factory saying, "Actually, I can put an electric motor at every workstation. I can completely reorient the way that the factory works." This is how I think it's happening in robotics now. we're dropping in a robot where a human used to be and you're like wait a second if I just like took a building and said actually my goal is to feed the robots and keep them 80% utilized how would I build this factory and that's a very different structure than like replacing a person with a robot in terms of it and that's where I think you get 5x 10x wins

uh are you worried we're going to run out of problems in the real world for startups to solve no I'm kidding uh I do I do uh

no one's ever asked that.

No, I mean it is it is fun.

Return the funds. There's no more ideas.

Yeah. Yeah. Just return capital. Return the capital.

No, but but uh any any intuition around like the you know the next big big category and sort of atoms like is it is it just robotics generally? Like how how do do you have a do you have a uh any ideas around like timeline of robotics and the application of them across different industries? because this is one of those, you know, categories that I think it's been it has been happening. It's real in certain industries, but it's been promised for so long. This time around does feel a little bit different. But I'm wondering I'm wondering how you view it. Um just because if you get the if you obviously if you get the timeline wrong, uh you can get the bet wrong entirely.

Yeah. predicting these things is you know we we have two bets in fusion and the joke fusion has been the you know the joke it's 10 years away every 10 years for a long time but this time it's it's it's different um but uh the so robotics I think is quickly going into that zone um I'm really bullish on industrial robotics um and we already use robots in fact car factories all the time um and so the question is can you move into like slightly less well ststructured things like taking a box if you look at a lot of like factories and wareh houses like there's a surprising amount of like unboxing and boxing that happens. I get a box, I get a box cutter, I cut it open, I pour all the stuff out, I get rid of the plastic bags. It's all done by humans now. Um, that's a really hard thing to automate with like a cuckoo robot like cuz it's just every box is a little different and like the bag gets stuck and all the rest of it. And so like those are the sorts of like things that I think are going to be really cuz they're still in a controlled environment. I'm not in my house. I don't have a dog running by. You don't have to worry about the robot tripping over my pet. Um, so I think the home is like like I'm excited about it. I want a robot to fold my laundry and empty the dishwasher. Can't wait. But like, man, you're playing on hard mode to like start in the home. I've shipped home electronics. Like people drop stuff. They spill wine on it. Like there's all sorts of things that happen in the home that I just don't have to worry about in a factory. So I think you're going to see it in factories and warehouses first. Um, and then you're going to sort of see it go out from there.

Spilling wine on your Oculus or your Quest. That's a wild time. That's something you didn't you ship millions of them. So, yeah.

Oh, yeah. No, no. This is the the the thing that created all the gray hairs is like you do all this work making this awesome product and then you're like, "Oh, we have this thing called a drop test. Like the consumer takes it out of the box and it's supposed to go on my face. Instead of putting my face, I drop it on the floor and like does it break or not?" And so, we spend all this time like dropping them and putting these high-speed cameras to watch like which part breaks and then go reinforce it. It's just like

that a lot of weight and that hurts the product. Yeah. Yeah. And the bigger issue with robotics is like then you have like you know there's not a huge risk of like if I drop the Oculus

does uh maybe it does some minor damage to the floor and the device itself, but you add these like big humanoid form factors and suddenly it's like the dog died because like the the robot, you know, just fell on it. You know, this is uh or it does like really material damage to an appliance or something like that.

Yeah. I mean, besides laundry and emptying my dishwasher, I wanted to like make me breakfast, right? But like now we got a robot like operating with hot oil, you know, on a stove with a flame. Like, what could go wrong?

That's going to be wild. Uh solar. Uh I feel like there's, you know, the the Elon Musk of space is Elon. He did electric cars. He's working on robotics. Uh he's been a little shy about solar. uh I mean he's optimistic about solar but he he hasn't actually scaled up American manufacturing capacity of solar. We have a number of founders you know if you have Palmer Lucky in defense and uh there's a number of nuclear founders who have sort of taken up the mantle to go on the press tour that's necessary to get folks excited about nuclear. I haven't seen that in solar as much. Am I just not paying attention? Is there some sort of fundamental constraint why solar is not going through as much of a fast takeoff? We've talked to the CEO of T1 Energy. That's sort of this, you know, acquisition changing the management. It's very exciting, but it's not the same, you know, startup seed round, you know, visionary founder all the way through this really broad vision. And I'm wondering if that's going to happen, if we're early, if it's already happening and I'm not I'm just not aware. What are you thinking about solar generally? Yeah, I I it's such an interesting thing because solar has gone through the biggest cost decline of anything I've ever seen and you know it's 99% cheaper than it used to be. Um it's mostly manufactured in China. It's a scaled manufacturing problem. That's not a great problem to go after as a startup where it's like manufacturing scales what determines

compete China

against China like good luck. So so I think that that that explains a lot of it. Um, what I do think is going to happen is if I'm just trying to make a, you know, 400 mm by 300 mm solar panel cheaper than China, good luck. But like for a long time, we treated these things like they were like precious minerals. And so if you look at a solar farm, you like lay down all this steel and you put up all this framing and all the rest of it to like put this panel on. Well, when that panel was the most expensive thing on the field, then like yeah, that makes sense. That thing is now the cheapest thing in the field. All the rest of the stuff, the like steel frame cost more than the panel. And so I think what we're going to see and I'm excited about a couple of startups in this who are saying like let's rethink how we actually deploy solar from a form factor from an automation perspective and like just go after installation cost and assuming the cells themselves are basically free. It's basically glass

um like what would you do differently and that again I think this notion of like constraints have changed is what creates a new opportunity. So that's to me what I'm most excited about in the next wave of solar.

Yeah. And even from uh you know obviously there's a lot of geopolitical considerations with the US and China but at the end of the day like a solar cell is not the same as buying like a full computer or something that can have spyware on it. Like it is just a chip there that that it's you know where would the spyware be exactly? And so I think I think you'll see more like hybrid approaches like this like what Whimo's doing where they're buying Chinese batteries and Chinese steel frames but then they're putting all the software and hardware on board. So correct. It's a little bit uh more uh you know less politically dicey I think. Anyway, um last question on fund size, fund structuring. I feel like a lot of these projects are capital intensive for the best possible reasons. Uh how are you thinking about actually participating over time? Do you think you'll be doing SPVS to maintain ownership? Because some of these projects are probably going to be pretty capital intensive. That usually means dilution. Have you thought about that? And do you have a strategy in place?

Yeah, I mean so the first thing we have been doing is just going to work to get these companies fundraised. So we have a giant network of of other so Pentalosa is a great example just raised a huge round from from Peter Teal. Um we were very early investors in that. So the first thing is get these companies successful then you have the options. Um in terms of how we take advantage of that access and option is is something we're going to figure out over time. Um you know I think this is the beginning not the end. uh we are overs subscribed in this fund so there's lots of opportunities for us to of other other vehicles that makes a lot of sense and uh and get going but I wanted to keep it you know I give all our entrepreneurs advice stay focused so let's get out the gate stay focused prove that there's a bunch of early stage to your first question are there opportunities for