Episode 77: Gigascale’s Mike Schroepfer and the laws of climate technology

Molly Wood Voice-Over: Welcome to Everybody in the Pool, the podcast where we dive deep into the innovative solutions and the brilliant minds who are tackling the climate crisis head-on. I'm Molly Wood.

This week, we’re going to keep talking about tech and climate innovation with the former CTO of Facebook Mike Schroepfer he spent decades building the tech industry as we know it before founding a climate tech investment fund.

He believes we’re at an inflection point we’ve been at before where solutions in this case climate tech solutions are about to cross that magic threshold where they become not just better for the planet, but flat-out better across the board AND cheaper than what they're replacing.

Plus as you know we just wrapped up our series on AI's massive energy footprint today we’ll talk about how tech giants have the deep pockets AND the urgent need for clean power and that could mean good news for all of us.

Let's get into it.

Mike Schroepfer 

Yeah, so my name's Mike Schroepfer. Most people know me as Schrep. I am running a climate-focused venture capital firm called Giga Scale Capital. So I spend my time investing in early-stage deep tech startups who are trying to of hasten a clean energy revolution.

Molly Wood 

Amazing. And then how did you come to that? Most people I think will likely know you as the former CTO of then Facebook. But what what brought you from there to climate tech?

Mike Schroepfer

Yeah, I mean, I've been a technologist for, wow, geez, 27 years now. Graduated in the 90s with a computer science degree, was in the dot com boom, started a company then, worked at a web browser place called Mozilla in 2006 when sort of the birth of Firefox and the open web was a big deal. And then joined Facebook, now Meta, in 2008. My joke is it was smaller than Myspace at the time when I joined, and then we got bigger. But I did a lot of fun stuff there, and we can talk about that. You know, really it was about five years ago, I started to get, you know, just really passionate about what are the problems to solve, you know, for future generations. And climate change was one of these like clear problem, gonna affect a lot of people, required a lot of investment. I originally thought that, you know, this would be a sort of nights and weekends philanthropic endeavor. And that I would sort of, and that's where I started. And I started like, look, I'm gonna direct some of my energy here towards that. I actually continued to do that work and doing a lot of it. And there's some amazing science, policy, sort of field building opportunities there that I'm excited about. But as I did that, I sort of accidentally became a climate tech venture capitalist because I kind of bumped into lots of entrepreneurs and was like, hmm, actually that's a really good business. Like that technology, know, my reading on technology is technologies get cheaper over time. And you like have this magic crossover point where like the new thing is now suddenly cheaper than the old thing. And then everyone switches. And we're seeing this in solar and the clean tech world. Obviously we saw this in smartphones and the web. now AI. And so that sort of dynamic of like investment in a new technology can sort of remove a bunch of really challenging problems. Like how do we give everyone the resources and power they need to live like healthy, fruitful lives without having to have a conservation mindset? And it's like, it's technology. And that's sort of, just kind of was accidental, kind of snowballed and I just sort of like fell backwards into, you know, I want to take the skills, the energy, the resources I have. and work with entrepreneurs who are trying to tackle these really hard problems to build sort of amazing companies that really shape the way we live in the next 25 years.

Molly Wood

Yeah, I want to ask you more about the technology piece. I would say I hate it when podcasters make it all about them and also went through a similar journey of kind of slowly realizing like, oh, this is the biggest tech and business story in the world. This is fundamentally the way we save ourselves has always been through innovation and ingenuity. so talk about like what was you just touched on in a little bit, but what was exciting to you about that as somebody who was like, oh, this is a pattern I can recognize.

Mike Schroepfer

Yeah, well, I mean, let's first talk about tech and then we'll move into climate tech. I think that the understated trend in tech is chip development. Like we've had this massive every 12 to 18 to 24 months doubling in transistors and doubling in performance of chips for my whole career. I was the last class at Stanford that we didn't have to take assembly language programming anymore because nobody used assembler because you could just use C because you could waste cycles on the CPU because they were so cheap.

And then now we're all in Python and a bunch of JavaScript and all these really, really inefficient languages that just burn cycles. And then now you're like an AI world where, again, it's like who can build the most biggest, largest compute is sort of where you're going. So this is this underlying trend of I wake up every day and the compute gets better, faster, cheaper. And then we can build on it. It's a massive tailwind. And so you see other technologies which look very much like that. And the two most obvious ones are solar and batteries.

Molly Wood

because we can.

Mike Schroepfer 

You know, solar is now the cheapest way to add electrons to the grid. You know, in the US in 24, 90 % of new energy on utility scale energy was solar. And it's not because of mandates or this or that. It's because it's like people do the math or like it's cheaper. And then lithium-ion batteries are even more staying. I'm back in the day when cell phones first came out. You had this like briefcase style thing and it lasted four hours. And now we run around with these supercomputers. They last all the time. You know, that lithium-ion battery came out in 91, came out in my lifetime.

You know, now in 2025, it's literally 97 % cheaper. It got like 50 % cheaper in the last year, which is just crazy. And so then you say, OK, despite all the hating on electric vehicles, there's some basic math that people are missing, which is like an electric vehicle powertrain is about three to four times more efficient than a gas powertrain. A good gas engine takes about 20%, 25 % of the energy and turns into the wheels moving.

A good electric powertrain is like 95 % efficient. And so I can screw up a lot of things on the way to making batteries and making electricity and all the rest of it before I still am better than you in efficiency, which fundamentally means cost. Like, how much does it cost me to drive this thing down the road? And so we're on this cost curve with batteries, where in many places in the world it's already overwhelmingly cheaper. And it's just like so obvious in next five or 10 years.

Molly Wood

Mm-hmm.

Mike Schroepfer

that you're going to show up at a dealership, like, this car is cheaper off the lot, and it's cheaper to fuel. What's the problem? besides the fact it's quieter, faster, maintenance is cheaper, a whole bunch of other doesn't literally throw poison out the back where my kid is, those are all good things too. But at the end of the day, the thing that stops everything is, huh, it's cheaper and fast. It's just better in all aspects. that's sort of the, huh, foundational technology.

Molly Wood

Right, and not for nothing, it's hella faster.

Mike Schroepfer

basically start asking some basic questions like, huh, solar gets really cheap, batteries get really cheap. Like what else gets disrupted? The fun part of the cell phone revolution wasn't the cell phone, it was all the apps that came after it. Like you couldn't imagine Uber and DoorDash before you had a cell phone, but afterwards it's like, yeah, okay, so everyone has a cell phone with GPS and then I build all of these things. We're in that part of the clean energy transition, which is like, everyone's gonna be like, wow, it's like really cheap to put new electrons on the grid. We have these really good batteries now.

There's a couple of other technologies that are like three or five years back in the curve from there. know, electrolyzers were there. I think fusion and advanced vision are there. And so they show up and they just like all of a sudden upend a bunch of things that we were assuming along the way. That is A, it's kind of fun. B, that's tremendous business opportunity. It's like textbook business school disruption. Like in our face, it's like a slow and like.

It's happening. It's so obvious. It's kind of like how AI was obvious to me in 2012 when we built an AI research lab at Meta. It's like, huh, this AI thing is going to be big. So that's the one sort of constant is just kind of being able to spot these megatrends that I feel like are fairly obvious and then have massive implications.

Molly Wood 

Yep. We're going to come back to meta and AI in a minute. then tell me how that led to gigascale. So tell us about gigascale and the work you're doing there.

Mike Schroepfer

Yeah, mean, so was sort of a piecewise, you know, I always think about what's the problem and then how do I get myself there. So again, I started with philanthropy and then was like, people heard I was stuff in climate, so a bunch of entrepreneurs reached out. I like sort of did some angel investing, maybe 20, 18, 15, 20 angel deals. I started taking a lot of free calls for people on nights and weekends, like, let me, like, how do I hire VP of engineering? How do I select a site for my factory? Like all these things. And I basically just said, like, wait a second, I don't have enough hours in the day. My time is limited. So I should probably spend my time.

with the entrepreneurs working on the problems that have the most likelihood of impact. And I'm like, well, the whole process for evaluating which companies we think are going to be more successful or not is basically a venture firm. So it's like, that's like, that's what they do. So it's like, I guess I'm starting a venture firm.

Molly Wood 

Give us some examples of companies in the portfolio. Like what are the kinds of things that you have funded that people are aware of?

Mike Schroepfer

Great. I'll give you three or four just to cover the spectrum. So let's start actually with consumer, because I think it's easy for everyone listening to this to think about. And I think consumers is a really hard market to attack, but once you attack it, it's really exciting. So I'll talk about two consumer companies, and then we'll talk about some more enterprisey back office stuff. So one is Mill. I think you might have had Matt Rogers on your podcast recently. you know what a Mill is, but for your audiences, this is how I think about it, is

Molly Wood 

Also, I love consumer. That's where I grew up.

Molly Wood

I did, and I have a mill.

Mike Schroepfer

It's a home composting device, so looks like a trash can, but you throw all of your food waste into it and it grinds it up and dries it out, so it kind of turns into coffee grounds. You say composting, what's the big deal? 90-something percent of homes in the US don't have composting, but the pitch to consumers isn't about composting. It's literally, just, whenever I talk, it's like, do you enjoy taking out your trash? No. Would you like to do it less? Yes. Would you enjoy it if it didn't smell? Yes. Cool. I have a thing for you. Like, put this thing in your kitchen, throw all your food waste in it,

Molly Wood 

Mm-hmm.

Mike Schroepfer

Once a month, you get like a little shoebox size thing of grounds that you can either ship back to them, and by the way, we'll turn it in chicken feed, or you can put it in your garden, whatever you want to do. But fundamentally, it like makes your kitchen less of a hassle. And then if you happen to care about the environment, it's cool. And it's a beautiful device. It's written by the founders of Nest. So it's like, in order to do this, you have to be the best of best. So this team, from a product design perspective, is the best of the best. It's quiet. My favorite, I'm sure Matt told you the story thing about them, is they...

Molly Wood 

Right, also it looks super pretty.

Mike Schroepfer

You know, they put one of these in their main conference room and they challenge all their employees to bring the smelliest possible things they can to work. Think like, you know, kimchi, rotten eggs, and you put them in there and it's like, if this thing isn't smell free, we're going to suffer. And that's what you want out of a startup. It's just like, they're just all in on like, we're going to fulfill our promise. So this device is awesome. And it's one of these things that what sold me is their, their MPS is so high and you meet people have it. It's like, they become the annoying people at the party or like.

Molly Wood 

gonna this.

Mike Schroepfer

my god, do you have a mill? Can I type out my bill? And that's all they're going to talk about. They're going to sell three mills every time they see a friend of theirs. That's why consumers are awesome, because once you get it, people are going to tell all their friends, and then you're off to the races. I'll give you another simpler one. Selena de Pauca Wall is just a legendary founder. She started Evite back in the day. She worked at Glaxo. She was at SurveyMonkey. I've been working with her for a while on some climate ideas, and she has basically started a company called HomeBoost. Their pitch is real simple. Most homes are not.

energy efficient, costs most consumers two things. One is if you live in a cold climate, your home is kind of drafty, you know, and two is you're wasting money on your energy bills. So we ship you a little box with a camera and some AI stuff in it. You run around your house and we'll tell you, here's the cheapest, most ROI positive, like interventions you can do to improve the comfort and efficiency of your home. Usually it's things like insulation or replacing an appliance. But the goal is after I spend like an hour or two,

Like within six to 12 months, I've got payback from a financial standpoint, and my house is more comfortable. So again, you're going to do this, you're going to tell your friends, my god, I'm saving all this money. Just super easy, simple, consumer. So those are two AI technology there on the first one. Really good product design on the mill perspective and ability to develop consumer products. And then I'll go totally different direction if that's OK. So let's go into energy. Let's talk about, I talked about solar, I talked about

Molly Wood 

Yeah, please.

Mike Schroepfer

batteries which work great for grid storage. Solar only works about a quarter of the time. It's either dark or cloudy, depending on the season. You've got this magic reaction in the center of our solar system called fusion in the center of our sun. It's the most energy-dense reaction in the world. Humans have made fusion happen on planet Earth multiple times. We know it can be done. The question is, can you harness it into a power plant? And if you can do so, you have this like, it's like the.

The end of the tech tree in any video game you've ever played is like, oh, then you have Fusion, and then everything else gets unlocked. Because it's like, give me 40 acres. I'm going to drop down this thing. And all of a sudden, we're just making electrons. And all I need is a pickup truck once a year. That's like my refueling. It's like, I could drive the fuel to the thing for the year to fuel basically a massive AI data center. And so it is just the ultimate humanity unlock. But the exciting thing is there's this joke that Fusion has been 10 years away every 10 years.

Molly Wood 

Mm-hmm.

Mike Schroepfer

But that's actually totally wrong. And again, this is true in a lot of trends. If you look at, there's this like, there's this particular criteria for energy output. And if you look at sort of a log-log chart of time and progress along temperature and pressure and a bunch of other things, we have been making faster progress in many cases than Moore's law. We've seen the national labs with NIF achieve what's called Q greater than one, more energy out than energy in. And so we have several startups out there who are credibly making approaches, Commonwealth Fusion,

in Boston has a magnetic confusion, a tokamak system that like probably will turn on in about a year or two. We've got another company, Exomer, that's working on the same technology as the National Ignition Facility. It's basically shoot a target with a laser and make it fuse. They just happen to be building a much bigger laser and the power output is sort of exponential as you increase the power of the laser. So it's like the obvious thing to go after. But if any one of these things work, you just like all of a sudden have this magic power unlock.

that allows you to do things. And then I'll give you just one last example and it's sort of again in the grungy bits, but another area of technology is this concept of an electrolyzer where I like take two plates and a membrane and I apply a current and like something happens chemically. And you can use it for a lot of different things, but we've got one company, Dioxcycle, that basically has figured out how to do this and make ethylene using carbon monoxide as the input. Ethylene is like this thing probably no one's ever heard of, but it's the most made organic chemical on earth, molecule on earth.

It's a you massive market and it's into everything shoes your your yogurt cup like just pipes in your walls like you name it like ethylene is the primary component and because of advances in this technology They can show up with a system that is by our math cheaper than using fossil fuels to make this stuff So you you start with clean inputs you put in clean electricity you end up with an output and It's cheaper

And the fundamental scale is relatively small. It's a small device, so you can make lots of them. So all over the and I can give you 12 more examples, but we should move on to other things of like consumer, enterprise, but they're all fundamentally like taking a technological thing, whether it be fusion, electrolyzers, AI, or just doing great product design in Mill's case and applying it to a product that the consumer is begging to get and telling their friends about, or an enterprise is like, thank you, you're saving me money.

Molly Wood Voice-Over: Time for a quick break. When we come back, we’ll talk about getting people to adopt climate tech solutions we’ll talk AI and clean energy demand and we’ll talk a little bit of politics.

Molly Wood Voice-Over: Welcome back to Everybody in the Pool. We’re talking with Mike Schroepfer the founder of Gigascale Capital and former chief technology officer at Facebook.

Molly 

I don't I don't think there is a company in history that nailed adoption better than Facebook and Meta. And we both know, I think, as people have been around technology for a long time that you can invent something really great. And if nobody uses it, it doesn't matter. It can, in fact, be the better product. But if nobody adopts it, it doesn't matter. So talk about what you see as you you mentioned this with mail, right? It has to be beautiful and make your life better. These seem like simple things, but what are the criteria that you apply when you talk to founders and companies? And especially, I would say actually at the industrial scale that say this is adoptable.

Mike Schroepfer 

I think there's two really big things here. One is you have to be, like no amount of quality entrepreneur will save you from being in the wrong market. So you have to be in a market, a broad market space where, and I define this as like power generation, chemicals, like you have to be in a space where there's enough demand and it's growing. In the Facebook, in 2008 when I joined Facebook, word on the street was that social networks didn't make money. MySpace had signed this big deal with Google. they had a minimum ad they were supposed to deliver, they couldn't hit their minimums because the ads didn't really work. And like everyone's example is like, know, AOL Instant Messenger and ICQ and email, like these are products that everyone uses, nobody makes money from. So the word on the street at that time was like, it's a product everyone uses, that company's never going to be successful. Like they're going to run out of money. It was literally what a lot of people were telling me. But when I looked at the company and met all the people, I sort of answered some simple questions. It like, what problem fundamentally are they solving?

In 2008, this was like, how do I keep up with my friends and family? And who in the world needs that? And I was like, literally everyone. It's like, have you met a person that's like, don't actually care about anyone else on the planet? OK, there's a couple. But the other 8 billion people care. So complete universal need. And you just look at how people are using it. And it's like, huh, this is really hit on something. And you're taking a bet. We're going to figure out how to turn this into a business because it's such a big need.

And then the second, and this is really important for all of our founders, is like the thing that Meta slash Facebook is just world class at still to this day, is like obsession with like test, iterate, and speed. It's like, I think of this as like speed to answer. How quickly can I get an answer to the question? And it isn't like, let's sit here and try to be smart. It's like, let's try something. Let's like build a prototype. Let's ask a customer. Let's put a product in market and then quickly, quickly learn from that and then move on to the next thing.

And this can be small incremental improvements. And people often malign this sort of move fast sort of attitude of meta and say it doesn't work in hardware. But I build hardware. I built the data centers there, OctaOS Quest. We built the reddit meta RayBans. You've got to apply the same principles. Yes, it takes six to eight weeks to spin a piece of hardware. So if you make a design mistake, it's very expensive. That is absolutely true. you've got to be, this is definitely just measure a couple of times, then cut. But.

Mike Schroepfer

You have to always look at it as like, cool, cool, that's an invariant six to eight weeks spin cycle. Given that, how can I get data as fast as possible? And some of our best entrepreneurs are like, okay, this part's gonna take three months to order. I'm going to Home Depot and figuring out how I can put this thing together with heat tape duct, a fan, a blower, and all the rest of it, and get like 80 % good enough answer to my question this weekend while I'm waiting for this thing to show up. Because the companies that fail is like, wait two months, the thing shows up like, geez, that design doesn't even work. I just lost two months. So to me, it's obsession, obsession with like every day, how do we get an answer to this question faster? And I don't take no, it can't be done, it takes too long, whatever, ever as an answer. It's always just like there's gotta be some way around this problem.

Molly Wood 

Mm-hmm. It's so interesting because in the deep tech space, you tend to encounter a lot of academics. Like people who are just like, I operate on bench scale for years. And it feels like that mental transition can be hard to like, how much matchmaking do you do of a really brilliant scientist with like an entrepreneur who's like, I'm going to Home Depot, you know? Yeah.

Mike Schroepfer

It's so challenging and you're definitely described as outlier or unicorn hunting. This isn't like we're not dealing with averages, we're dealing with outliers fundamentally in this business and it really breaks everyone's brain. I'll say that I'm just going to pivot for one second is like the hardest thing for me is switching from an operator to an investor is like my joke is like as an operator, you have to be right more than you're wrong. If I was like wrong three times out of 10, I'm fired. So I got to be wrong like nine. So you spend a bunch of time trying to be smart and be right.

Molly Wood

Right.

Mike Schroepfer

As an investor, actually not the batting average. It's like the number of home runs. It's like, can't miss out on a huge opportunity because I'm trying to be too perfect or too correct. And so you have to open yourself up for possibilities in these unicorns. I think that there are, bringing this back to your question, there are some people who can make this transition, and they're very rare, but when you find them, they are magic. And I'll give you two specific names.

Bob Mumgard is the CEO of Commonwealth Fusion. If you've ever chatted with him, if... Great. Great. I would challenge you you like didn't get his bio in advance, you'd be like, is that a plasma physicist or a CEO? It's like, he sounds like a CEO now. He's like talking about running the company and customers and this and that. he is, he is all in like I, when I talk to him, like you're an operator. You're like, you've lost, you're no longer a plasma physicist. And like he has made that to come out of PhD. He is all the one that ever like...

Molly Wood 

We had him listeners, he was on at the end of last year.

Molly Wood

building an ecosystem for, yeah.

Mike Schroepfer

wow, we don't believe in PhDs coming. It's like, he's the counter example. And then Sarah, who runs this company Docs, like I mentioned, same thing, postdoc from Stanford, technical background. But man, she is just relentless in getting things done. She's often in Europe, so we're often talking early morning, my time on weekends, because she wants to talk about hiring someone, operations, customers, whatever it may be. And so it really does take this person who can convert themselves from

All I care about is getting this new data or discovery and opening myself up to that to, I'm going to dedicate my life to solving this problem. Bob's like, I'm building a fusion power plant. That's what I'm doing. I'm not trying to publish physics papers. And that sort of transition is magic. Sometimes you can get one person to do it. That's the best. Sometimes some people are good in the early stages, and then you find someone who can pick it up and take it to commercial success. But that's fundamentally what we're trying to do.

Molly Wood 

So switching gears a little bit, you're like the perfect person to talk to in this exact moment where, for example, a lot of new climate tech investors are fleeing to AI. And also, we're at a moment where AI is reigniting energy conversations in a really fascinating way and potentially even introducing not only new challenges, but new sustainability solutions. So just like,

Before I get to specific questions, give me your kind of 10,000 foot view of what the hell is happening right now? Shit, shit is crazy, Shrep.

Mike Schroepfer 

It's exciting! It's fun! Definitely not a time to be, like, definitely not boring. And boring is the worst. So, look, I think it's... I don't have a perfect crystal ball for all of these things. But I would say that, like, I'll touch on a couple of different things. I think the tourists have come and gone in climate tech, and, like, I'm fine with that. So I think that, like, it takes long-term...

Molly Wood 

Nope. Did they do damage? I'm just gonna interrupt you. Did they do damage with the valuations or are we gonna recover from that too?

Mike Schroepfer 

I just I've never seen a cycle that didn't do this so dot-com boom mobile phone early web AI like I just I guess there is a way to do it where everyone perfectly Invests and doesn't overdo it on the up and then crash down and come back I like in theory in some economics textbook. There's some way to centrally plan that I just have never seen it in my lifetime so like so like I'm just kind of a little bit more sanguine about like yep We got like we overdo it on the up then we come down and then all the real work happens and I actually

Molly Wood 

Sure, but where's the fun in that? Yup.

Mike Schroepfer 

least enjoy the super hypey up. Don't love the down, I love the bottom. Because then it's like, that's when like, again, all the like, tourists go and like, cool, we're gonna go build some stuff and show up in three years. And everyone's like, Whoa, where the hell did that come from? It's like, yep, good luck catching us now. So that and that, I mean, I built my first company after the dot com crash. And it was that it was awesome. Like we hired so many great people. Nobody thought we could succeed. It was like a great time to build a company. So from that perspective, I'm kind like, it just is what it is.

Molly Wood 

Mm-hmm.

Mike Schroepfer

We're going to build. It's fine. fun of like batteries are still getting cheaper. Solar's still getting cheaper. Like all people want energy. So like it's going to be fine. So there's that. So good news, good news. Like for 20 years in the US, we've had relatively flat energy growth. So if you were like, oh, I'm going to make energy, I'm going to store energy, everyone's like shrug emoji. know, now 3 to 5 % growth a year, 5x by 2050.

Molly Wood 

You look over here, we'll look over here.

Mike Schroepfer

Everyone's freaking out about where to get the electrons, right? And when you're in that growth market that seems a bit more like a tech market, which like, okay, cool, everyone's like racing past each other to get their thing. You've got a lot of room to bring new solutions to market, right? It is really hard to disrupt a static market. is like disruption, healthcare is really hard. Once something gets cemented, it's really hard. So I think it's a massive opportunity because a lot of stuff is getting built and people care how fast and how cheap.

And so if you can show up with like, can get you electrons fast, then you've got a real good business now. So that is awesome. And I think this will continue. 

Mike Schroepfer 

We're going to use it, if not for training, then for inference for the foreseeable future. And that, yes, a big energy need. But from buyers who are motivated and have the margin and profits to pay premiums to make it come either faster or be more sustainable or to offset the carbon in some other way via voluntary carbon purchases. So this is incredibly good news for anyone who cares about this. Because I think we're building a lot of new stuff, and we have a chance to do it a little bit differently.

So can you build a completely off-grid one gigawatt data center with solar and batteries? Can you pre-purchase from a fusion or next generation fission plant? Can you pre-purchase from enhanced geothermal? All of these things are literally happening. Can you buy carbon offsets from high quality carbon removal, DAC or enhanced rock weathering or others? Yes, all of these things are happening. as a startup entrepreneur, you're like, I need to have customers.

and then I raise money from the investors based on customer interest. And like, we have the customers for these things. And so it's an incredible opportunity.

Molly Wood 

Is this the, when like I'm old enough to have known a guy who was mining Bitcoin in his apartment and then was like, wow, this is really jacking up my electricity bill, I'm gonna stop, right? And then the conversations about is cryptocurrency going to catalyze new investment in renewable energy? And maybe it did, but it also brought some coal plants online. This feels like a moment that is to your point, like you're saying hyperscalers can, likely will be?

and should invest in energy sources that are sustainable or even yet to arrive.

Mike Schroepfer 

They are. Now, I will caveat that you can never perfectly plan this. So will some gas power plant be built or be kept online, et cetera, in order to meet someone's demand? Yes. This is not going to be perfect every single day of the year between now and 2050. But if you just look at the amount of money these companies are putting into clean energy and helping to advance technology, advance nuclear fusion, geothermal, battery storage, et cetera,

Molly Wood 

Yes.

Mike Schroepfer

and carbon removals, it's like dominated by the hyperscalers in terms of that, you know, trying to, they're technology companies, they're like, cool, how do we pull this technology forward? Because if we can make carbon capture, you know, good carbon capture, a hundred bucks a ton, that makes it really easy for us to do that. If I can get clean energy at gigawatt scale wherever I want to build a data center, that means I don't have to have, again, back to like no trade-offs, technology eliminates trade-offs, I don't have to make any trade-offs, I just like build that fusion plant, build my data center, good to go.

And so they're all investing and so, you know, I can't say every day is gonna be perfect and there will be plenty of opportunities to write some article about something not happening that's or something that's not great. But I think them again, we're talking about mega trends. The mega trends are definitely positive. If you told me I could invent two futures, one where like we went back to flat energy demand and we're building no data centers or this one, I'd be like, gosh, no, I do not want that. That's gonna make everyone's job here much harder.

Molly Wood 

or hopeful, but what do you think about the potential of some of these new AI-based technologies? I mean, obviously, HomeBoost is one of them at a small scale, but is that real? Is the materials discovery real?

Mike Schroepfer

Yeah. I think there are a lot of, I think it's a little bit overhyped in the near term. I think there's a lot of opportunities and it's a tool. It's not like a magic wand. So I think it helps in a lot of places, but it's not, I don't believe in the like, I'm just going to sit back and wait and the AI will figure it out in 10 years. That's probably not how it goes. But like one of my favorite pitches is like a whole bunch of people would be like, oh, like there was this physics paper written in the 1970s and it was like,

you know, if we could just simulate this particle thing, but we can't because it requires this amount of flops and that's just like inconceivable. And you're like, actually we kind of do have that kind of flops now. We do have that compute power. So we can actually simulate at the particle level what's going to happen here. I think it was like, imagine the energy transition without computation. It'd be like real challenging. And the thing that we're missing is like, the other thing that gets me exciting in terms of megatrends is sort of like a tools megatrend, which is, you know,

3D printing, additive manufacturing, a whole bunch of other things makes iteration time really quickly. So one of our companies is working on a bunch of SpaceX engineers worked on the Raptor engine plus GE gas turbines. We've got a next generation bioenergy plant that's like carbon negative power. It makes clean energy and it actually sequesters carbon from the biomass coming in. It's kind of, it's amazing. But the reason they make this work is it's super high pressure, very high pressure vessel and they're using super critical CO2. These things are very hard to model with computers.

And so they're just like, cool, we can just 3D print 16 different versions of these like nozzle and injectors and run them all through a test sequence and be like, OK, this one works better. And they've been doing this for a while, so they have intuitions in how to drive these things. They're not starting from scratch. But they can just run a cycle loop of test and iterate. That's like 10 times faster than you could a decade ago. So I think people miss that there's all of these tooling advancements, which is like when applied properly gets back to that thing which is like turns hardware a bit more into software and a sense of ability to iterate.

Mike Schroepfer 

And that to me is the fundamental sort of promise of AI is like, it's going to turn more problems into software problems, computation problems, rather than six months of building problems. And that is just like, I think of it as a massive accelerant to everything we're doing.

Molly Wood 

The other piece of the landscape that has obviously changed quite significantly is the political landscape in the US, but I would also argue that there are trends in this direction globally. how, it is my opinion that the first Trump administration catalyzed a ton of innovation and private funding in the direction of climate tech. Is it your feeling that that is where we are now as well?

Mike Schroepfer 

I think it's going to take a little while to shake this out. I think there's going to be some puts and some takes, some good and some bad. I think that there are a number of places where the US government has been providing the capital upfront or capital guarantees to help run things down the R &D pipeline. I think that that is less likely to happen. We're just going to have fewer DOE grants. We're going to cut back funding on a whole variety of things. We're going to cut back subsidies. That's all sort of generally bad for the acceleration of climate. On the good is climate's going to require a lot of building stuff.

and a lot of in-source manufacturing in the US. And I think that this administration is going to be very strong on that. If we can build stuff in the US that is better, better, faster, cheaper, everyone's on board with that. And that's been my thesis from the beginning is like, yep, jobs, better for consumer. This is just good, good, good. Doesn't matter what your politics are. And I think that the more you go down that lane, the easier things are going to get. And I think people are going to be real excited about building, building factories, building stuff.

Molly Wood 

I think there have been a number of deep tech companies that I have spoken with who have found an unexpected niche in national security and saying that we're a critical domestic mineral supply, for example. That kind of language, like some of it is going to be storytelling, right? Yeah.

Mike Schroepfer

Yeah, yeah, but I think there's a lot. I I'm excited for America. Like, you know, we're an unrivaled zone of entrepreneurship. And so when you get us pointed in the right direction, you know, for a very long time, you know, from chips to software to smartphones to AI to clean tech revolution, like it's hard to bet against America. And I think that if we can just get aligned with like, yes, that, you know, building new things, doing them in the U.S. is good for everyone.

Then I think there's a lot of people who are all on that. And I think it can be really good and a really good time for us over the next, know, X years. And I think this is why I'm so focused on the economics. Like as an engineer, I spend more time thinking about dollars than anything else because I think that like ultimately that answers all debates. It's like if I'm talking about jobs and I'm talking about lower cost to businesses and consumers, like who's against that? Like nobody.

And so that's where we focus all our time and energy on is when we can show up with that message. And then I think everyone's happy.

Molly Wood 

Mike Schroepfer, thanks so much for the time. appreciate it.

Mike Schroepfer 

Thank you. It super fun.

Molly Wood Voice-Over:

That's it for this episode of Everybody in the Pool. Thank you so much for listening.

Next week, I promise we will take a little break from AI and get back to touching grass as the kids say we’re talking nature-based solutions and how to invest in them.

Email me your thoughts and suggestions on this episode or anything else that’s on your mind to in at everybody in the pool dot com I’ve gotten a lot of questions about waste heat lately I love you nerds, you get it and I’ll compile some answers for you in the newsletter this week!

Find the subscribe link and the latest episodes at everybody in the pool dot com, the website. And if you want to become a subscriber and get an ad-free version of this show, hit the link in the description in your podcast app of choice.

Together we can get this done. See you next week!