As the pandemic locked down cities in early 2020, Mike Schroepfer, then the chief technology officer of Meta, found himself with more free time than he’d ever had in his career.
In quiet moments that would have been filled with work travel, social events, or his children’s school activities, he reflected on how well humanity can pull together in the face of an acute crisis—implementing public health measures, mass-producing tests, and turbocharging the development of vaccines.
But the experience also reinforced his view that we are particularly bad at addressing slow-motion catastrophes like climate change, where the risks are grave and growing but mostly looming in the distance.
As he learned more about global warming, Schroepfer came to believe he had a role to play: By leveraging his technical expertise and financial resources, he could accelerate essential research and help society develop the understanding and tools we may need to avoid or prepare for the escalating dangers.
As the threat of climate change consumed more and more of his time, he decided in 2021 to step down from his CTO role and dedicate himself to addressing the challenge through both philanthropic and for-profit efforts. (He remains a senior fellow at Meta.)
I’m willing to take a lot of risks that these things just don’t work and that people make fun of me for wasting my money, and I’m willing to stick it out and keep trying.
Mike Schroepfer
In May 2023, he announced Gigascale Capital, a venture fund backing early-stage climate tech companies, including startups working to commercialize fusion, cut landfill emissions, and reduce methane pollution from cattle. That summer, he also launched Carbon to Sea, a $50 million nonprofit effort to accelerate research on ocean alkalinity enhancement (OAE), a means of drawing down more planet-warming carbon dioxide into the oceans by adding substances like olivine, basalt, or lime.
This year, as MIT Technology Review first reported, he launched Outlier Projects, which is donating grants to research groups working in three areas: removing greenhouse gas from the air, preventing glaciers from collapsing, and exploring the contentious idea of solar geoengineering, a catch-all term for a variety of ways that we might be able to cool the planet by casting more heat back into space.
Last week, Schroepfer sat down with MIT Technology Review in his offices at Gigascale Capital, in downtown Palo Alto, California, to discuss his approach to the problem, why he’s willing to spend money on controversial climate interventions, and what AI and the presidential election could mean for progress on clean energy.
This interview has been edited for length and clarity.
Is there a unifying philosophy across your climate efforts?
The foundation is that when you get a set of people and you get them all pointed in the same direction, and they wake up every morning and say “We’re going to go solve this problem and nothing else matters,” it’s often surprising what they can get done.
I think the other unifying theme, which also unifies my career, is: Technology is the only thing I have seen that removes constraints.
I just saw this again and again and again at Meta, where we would reduce cost, improve efficiency, develop a new technology, and then a thing that was a hard constraint before just got removed.
Through the proper development and deployment of technology, we can remove either-or decisions and move to the world I want to move to, which is a yes-and decision.
How do we bring the standard of living of 8 billion people up to those of the West and have a planet that my children can live on? That’s really the question, and the only answer I can see is technology.
There are a variety of potential approaches to ocean carbon removal—everything from sinking kelp, which doesn’t seem to be working that well, to iron fertilization and other things. So why enhanced ocean alkalinity? Why was that the one where you said, let’s dive deep?
In reading about all the different approaches, it stood out as the most likely, the most scalable, the most cost effective, and the most permanent, yet the least well understood.
And so it was super high impact if it works, but we need to know more.
I had no prior bias to this. I like kelp. I like all these things. I’m not a one-solution sort of person. I want as many things to work as possible.
As an engineer, my reading of technological deployment is that the relatively elegant, simple solutions end up being the ones that scale. And OAE is about as simple as it gets.
Let’s switch gears to a touchy topic: solar geoengineering. Why did you decide that was an important area where you wanted to support research?
We did a broad search for problems that are defined as high impact, high scientific uncertainty. Those are the ones that I think fit what we’re comfortable with and good at. And as we did that search, the two—besides carbon removal—that came out were solar radiation management (SRM) and glacier stabilization.
SRM felt like an orthogonal solution because it is a way to make rapid cooling if we need to—if this becomes a humanitarian crisis.
We’re already losing lives due to heat, but it’s going to get to the point where people aren’t going to tolerate it, and the question is: What do you do at that point?
Humans are good in a crisis, but it felt like, hey, we ought to get started now. To really start doing the rigorous work to understand “Does this work? Is it effective? What are the safety concerns?” while we’re not in a crisis moment, so that we’re prepared.
You mentioned glacier restoration as well. Why was that a problem you wanted to contribute to?
Assume we solve every other problem. We remove all the carbon, we electrify everything. We’ve still got a sea-level-rise problem, mostly because of glaciers that are moving.
One of the approaches is to simply pump water out of the bottom of the glacier to remove the lubrication layer that’s causing them to move. We have glaciers with boreholes already in them that are highly instrumented, and they’re already moving. So dropping a pump in there and pumping out water is a very, very, very low-risk activity that starts to answer some basic questions, like: Does this work at all? Would it be feasible? Would it be overwhelmingly impossible because of energy or cost needs?
Whatever approach you take to it, we’re talking about a massive infrastructure project that’s just gonna be incredibly costly. On the other hand, if the Thwaites Glacier (sometimes called the Doomsday Glacier) does slide into the sea, then every city around the world, plus every low-lying nation, has to do these massive infrastructure projects.
Can we pull together as a global society to address this thing in the most efficient way, or are we just going to leave everyone to deal with it on their own?
This is where I think people underweight the power of the prototype or the power of the proof of concept.
We can talk theoretically. I can bring scientists over and they can say, “I’ve got a big spreadsheet which explains to you how expensive this is going to be.”
I don’t know. Maybe they’re right. Maybe they’re not. Instead, let’s get on a plane. And let me show you. It was moving this fast. We did this. It’s now moving this fast. Here’s the pump. We’re pumping water out.
I think a lot of what my role in the world is to do is to get us to there. I’m willing to take a lot of risks that these things just don’t work and that people make fun of me for wasting my money, and I’m willing to stick it out and keep trying.
What I hope I do is put a bunch of proof points on the board, so that when the time comes that we need to start making decisions about these things, we’re not starting from scratch—we’re starting from a running start.
And you think that just having a greater amount of certainty and clarity—in terms of what the risks are, and how viable these solutions are, and what they will cost, and how we do it—can change the dynamics …
I think it does.
… where suddenly you could see nations pulling together in a way where it’s hard to imagine when there’s so much uncertainty?
Yeah. Or it goes the other way, where you decide, “Hey, we’ve had all these crazy ideas, and none of them are going to work, so we got to do something else.”
But as you say, the alternatives are moving lots of people or building big seawalls, and those are going to get pretty overwhelming pretty quickly.
My career has been putting tools in the toolbox. My job was to stock that toolbox such that when we needed it, we were ready to go. And I’m applying that same approach here, which is just like, “Hey, what are the things that I can help push forward in some way so that if we need them, or if we need to understand them, we’re a lot further along than we are today?” Right?
We’ve mostly talked about your philanthropic efforts so far, but you also set up Gigascale Capital, a venture fund. How does your investment strategy and approach differ from that of a traditional tech venture firm? For instance, are you investing over longer time horizons than the standard five to 10 years?
We’re here to prove that if you pick the right climate tech companies with the right founders, that can be an amazing business. They’re disrupting trillion-dollar industries, and so you ought to be able to get good returns on that. And that’s what’s going to be required to get a bunch of people to open up their checkbooks and really spend the trillions of dollars we need a year to solve these problems.
So we look for companies with—we’ve jokingly called it at times the “green discount.”
Those trends are freight trains that are going down the hill and are pretty hard to stop.
Mike Schroepfer
Like, “Hey, this is a better product. [whispers] By the way, it’s better for the environment.” Sort of the little asterisk if you read the fine print at the bottom.
The starting point is, the consumer wants it because it provides a lot of benefits; enterprise wants it because it’s cheaper. That is the selling point of all the products we back. And then it also happens to be a lot lower carbon, or zero carbon, compared to whatever alternative it’s displacing.
Your mentioning the green discount reminds me of Bill Gates’s green premium (the Microsoft cofounder’s thesis that it takes heavy investments in climate tech to reduce their cost premium relative to polluting products over time). There are some products, like green steel and green cement, where the alternatives are more expensive. Does that mean that you’re not investing in those areas, or is it just that you would with the hope that eventually they’ll be able to get those costs down?
Technology takes time to incubate, so no new technology out of the gate is better, faster, cheaper. But in the life cycle of the company, in five to 10 years—I have to believe, at scale, you can be cost competitive or have a cost advantage versus the alternatives. So that means that, yeah, we only invest in things that we think can either be cost competitive or have some other co-benefit that is a decision maker.
This is why I very cleanly separated philanthropic work where it’s like, “I get nothing out of this—we’re gonna send money away and hope public good, papers, knowledge gets created.”
And the venture fund is “Nope, this is the capitalistic endeavor to prove to people that if you smartly choose the right solutions, you can make money and fund the low-carbon economy.” That is the bet we’re making.
Given your recent job leading tech and AI efforts at Meta, I’m curious about your thinking about the potential tension between AI energy consumption that’s very much in the news right now and clean energy and climate goals. What do you think companies will need to do to stay on track with their own climate commitments as data centers’ energy demands rise?
Two thoughts on this.
AI is a foundational technology that can enable a lot of benefits for us moving forward. Part of why I still have an affiliation with Meta is because a lot of the work I do there is on Llama, our open-source model, which is allowing that technology to be used by lots of different people in the industry.
I think foundational technology being open is one of the ways in which humanity moves forward faster and gets more people into prosperity, which is what I care about.
In terms of energy consumption, I start with let’s get AI as fast as we can, because I think it is good.
In my time at Meta, we many, many times had multiple-orders-of-magnitude improvements in efficiency or power use.
So I think the industry right now is trying to build the best thing they can, and that consumes a lot of power and energy. I think if we get to a point where that’s a huge problem and we need to really optimize it from an efficiency standpoint, there are a lot of levers to pull there.
And AI or no AI, if you want to electrify everything and remove all fossil fuels, we just have a tremendous amount of clean energy we need to bring on the grid, right? That problem exists whether you have AI or not. So I think it’s a little bit of an over-highlighted sideshow to the real game, which is: How do we get tens of gigawatts of clean energy onto the grid as fast as possible every year? How do we get more solar, more wind, more storage? Can we bring fusion online?
To me, these are the humanitarian game-changers; it is the sort of unlock for a lot of other things.
I hate to get political here, but in light of these recent Supreme Court decisions about federal agency powers, I am curious what you think a Trump win in November might mean for climate and clean energy progress.
The short answer is, I’m not sure.
Okay, then maybe it’s the same answer to my next question, which is: What do you think it might mean for financial opportunities in the sector, to the degree that Trump has said he would try to roll back Inflation Reduction Act incentives for EVs and other things? Do you think it could weaken the case for private investment into some of these areas?
This goes back to when you asked, What do we believe? What do we invest in?
Basically, it has to start with the business case: My product is better or cheaper. I think that investment case is durable regardless. I think these things like the IRA can accelerate things and make things easier, but if you remove them, I don’t think that eliminates the fundamental advantages some of these technologies have.
The exciting thing about this world is that an electric powertrain on a vehicle is fundamentally much more efficient than a gas power train—like 3 to 4X more efficient. So I should be able to build a product that is very cost advantaged to these petrol-burning things. There’s a bunch of issues with the scale and customer adoption and things like that, but the fundamentals are in my favor.
And I think we see this trend happening in a lot of things. Solar is the cheapest form of energy generation we’ve ever had, and that’s going to continue as we massively increase manufacturing capacity. Batteries have gone down an unbelievable cost curve. And each year, we’re making more batteries than we’ve ever made before.
One of my favorite things is Wright’s Law: this idea that as you double the scale of your production, you generally see a decrease in cost. It varies from product to product, but for batteries, it’s about 20% or so every time we double the production.
If my product gets cheaper by about 5% to 10% a year, at some point I’m gonna win. Those trends are freight trains that are going down the hill and are pretty hard to stop.