Who's got the ball on carbon removal?

Carbon removal has gone from a niche climate concept to one of the world's most important challenges. Reid and Aria sit down with Nan Ransohoff, Head of Public Goods at Stripe and a leader behind Frontier, the advanced market commitment helping build the market place for carbon removal. Nan explains why cutting emissions alone won't be enough to meet climate goals, what it will take to scale carbon removal from thousands to trillions of tons, and why governments—not just companies—will ultimately need to create and fund the markets that make it possible. They discuss the most promising carbon removal technologies, the role AI could play in accelerating climate solutions, and what it means to be a "general manager" for challenges that affect all of humanity. For more info on the podcast and transcripts of all the episodes, visit https://www.possible.fm/podcast/

• 0:00Speaker 00

  You're looking at this report that is like the canonical climate change report, the IPCC report, and you're like, everybody's looking at these numbers. Everybody who cares about climate is looking at these numbers. And then you go out and talk to everybody in this space and it takes you three weeks and you're like, what is going on here? It is on us to sort of like figure out what that pipeline looks like. What is the PM to GM pipeline? How do we create these talent onramps and for great people, help them find a problem that they are really excited about and can wrap their arms around and want to work on for long periods of time. And we have to do that at a pretty large scale. This is a sort of wartime mobilization effort.

• 0:32Speaker 01

  The climate conversation has long focused rightly on reducing emissions. But there's another piece of the math that is becoming unavoidable. We also have to remove vast amounts of carbon already in the atmosphere and oceans and store it durably, not offsets, not vibes, real tons, measured and paid for.

• 0:56Speaker 02

  Carbon removal is a public good, which means everyone benefits, but no single buyer naturally steps forward. So the question becomes, who builds the market? Who takes responsibility for the whole field? And what does new philanthropy look like when the work is not just writing checks, but creating the conditions for an entire industry to exist?

• 1:17Speaker 01

  Every new field needs a first customer, someone who buys it before it's cheap, before it's obvious, before the rest of the market believes it exists. That was true for vaccines, for solar, and it's true now for carbon removal.

• 1:35Speaker 02

  Nan Ransohoff has spent the last several years trying to create exactly that market. As head of climate at Stripe, she leads Frontier, the advanced market commitment that has contracted nearly $700 million to pull carbon out of the atmosphere. But Frontier is more than just a buyer. It's a blueprint for how to shape a market around a public good that no single institution would otherwise fund.

• 1:59Speaker 01

  Through all this, we also have a management story. Nan has argued that the world's most important problems are orphaned.

• 2:09Speaker 01

  Plenty of people work on pieces, but no one feels responsible for the whole. Her answer: general managers for civilization-scale problems who define the outcome, find the bottlenecks, and pull whatever levers the moment requires.

• 2:26Speaker 02

  Carbon removal is a climate problem, but it's also a market design problem, a management problem, and a philanthropy problem. Nan is one of the people trying to own all three. Welcome to Possible.

• 2:38Speaker 01

  You co-run a scone-review website called Sconeheads with your sister. This feels important. So what's, does a perfect scone look like? And is there a climate angle that I'm missing?

• 2:55Speaker 00

  This is the best branding that Sconeheads has ever had. What makes a perfect scone? Well, there are really many things, but I'd say one is density. You know, you want, you want some heft to your scone. You're, it's not cake here, so that's kind of an important quality. Moisture, you know, you don't want it to be too dry. And then flavor profile, which I think is very personal, but you know, I think savory scones are kind of under, underdone. There is no climate angle. It was like a, it was a, what we thought at the time was a funny, a funny pun. You know, Sconeheads, that's pretty funny. Anyway, like years later, we're now maintaining a website and like vaguely reviewing scones occasionally, but has been like a fun side hustle with my sister and a friend.

• 3:37Speaker 02

  It's inspirational. But let's get to climate. So you grew up in North Carolina. You got interested in climate, I think, studying abroad in India. You spent time at Opower, Nest, Uber, Neuro. We all care about climate. When did climate go from, I care about this as an individual, as a hobby, as a passion, to I have to work on this for my job?

• 3:57Speaker 00

  There were sort of like two, two versions of that. The first was after undergrad, I worked in consulting for a couple of years. And while I learned a lot, realized that sort of it doesn't pass what I now call my 2AM test. Like, am I excited to work on this until 2:00 in the morning? And I thought climate would be that, and that is what got me into Opower and Nest. I got into carbon removal essentially in I did climate for a couple of years. I kind of got a little away from it. And I was essentially in North Carolina, helping my dad through a bone marrow transplant, a lot of downtime at the hospital. And I was like, I'm going to reread, I'm going to read the 2018 IPCC report to re-educate myself. For fun, like we all do. As one does. And in that report, there's of course, all of the emissions reduction

• 4:40Speaker 00

  that we need to do. And there's this big thing about carbon removal. And, you know, I was reading this and sort of looking around like, wow, this is, these numbers are quite large. And who is working on this? And just because I was curious, I started to kind of like poke around different folks who were working on it. And it took me three weeks to talk to like, everybody that I could in the field. And that was spooky to me. Right? Like ostensibly, we need to do tens of billions of tons a year, a trillion tons over the next 100, 150 years. Like, how is that going to happen? That is when I panicked and decided that I wanted to work on carbon removal full time.

• 5:15Speaker 01

  Let's start digging into this. And, you know, probably a number of our listeners are new to this. So what counts as durable carbon removal? And why is emissions reduction alone not enough?

• 5:30Speaker 00

  The spirit of durable carbon removal is essentially around counteracting a ton of emissions. So like, a ton of CO2 in the atmosphere is effectively permanent. And so you want to take it out permanently as well. I think people can debate the exact numbers, but that is the spirit of what you're trying to get at. I would say in the thousand to tens of thousands of years is sort of like a good proxy for durability. And if we sort of collectively had done a better job with emissions reduction decades ago, we wouldn't have to do carbon removal now. And that's the world we all would have liked to be in. But it is not the world we're in. So we have to radically reduce emissions, and we have to suck CO2 that is already in the atmosphere out and store that somewhere permanently. We're going to have to do a huge amount of both. And roughly, I would

• 6:13Speaker 00

  say, you know, we think about it like 90% ish should go to the reduction side. And then 10% roughly is on the removal side. But like, these are very large numbers at the scales that we're talking.

• 6:24Speaker 02

  So Frontier has contracted nearly $700 million worth of carbon removal. And portfolio companies are building towards 1.4 million tons of annual removal capacity.

• 6:38Speaker 02

  Can you tell us why those numbers are a drop in the bucket? Like, what do we actually have to do? And then also just tell us what Frontier is.

• 6:45Speaker 00

  Carbon removal, durably sucking CO2 out of the atmosphere and storing it somewhere. Over the next 100 to 150 years, to stay within warming targets, we're talking a teraton. So a trillion tons of carbon removal in 100, 150 years. That's a lot.

• 6:59Speaker 02

  Do we have to do some of it more now, though? Or no, it doesn't matter? Evenly over the next 100 years?

• 7:04Speaker 00

  It's not going to be, like, there's no way that we could ramp up to, like, 10 billion tons a year right now because

• 7:10Speaker 02

  We're not there.

• 7:11Speaker 00

  We're not there. And the starting gun for carbon removal was functionally the 2018 IPCC report. Like, through that lens, carbon removal has made a huge amount of progress over a very short period of time. And when you compare it to the teraton that we need to do, we are still closer to zero than any other number. The point of Frontier, and the why we started Frontier, was essentially to try and help build a market for carbon removal to catalyze the technologies that are needed. Carbon removal is a public good, right? When you are sucking CO2 out of the atmosphere and storing it somewhere, that does not directly benefit any end buyer. It benefits all of us. But as a result, there have been no natural customers for it.

• 7:45Speaker 02

  Totally.

• 7:45Speaker 00

  And without that, if you are an entrepreneur, why would you start a company if you have nobody to sell your product to? And if you are an investor, why would you invest in a company that is not going to generate revenue? So the idea with Frontier was to, it's called an advanced market commitment, and the idea is to send a really loud signal to entrepreneurs and investors that there is going to be at least a seed of the market for the thing that you want to build. And $1 billion compared to the cumulative $30 million that had been spent on revenue for customer carbon removal companies at the time is a lot of money compared to the hundreds of billions of dollars per year that we're going to need to spend on carbon removal. $1 billion is a drop in the bucket, but like, to the point of every technology has to start somewhere.

• 8:30Speaker 00

  We're starting from zero, we're moving quickly, and we are not out of the woods yet.

• 8:33Speaker 01

  So let's look at some of the technology of this, which is across the different pathways. There's direct air capture, maybe enhanced weathering, ocean alkalinity, mineralization. Has anything particularly gotten more promising in the last year and maybe anything less?

• 8:51Speaker 00

  We think that one area that is very underexplored relative to its potential are rocks. So most of the world's carbon is actually in the lithosphere. It's in rocks, it just takes a really long time to get there. And if you can find a lot of reactive rock, grind it up, expose it to air, give it some water, it essentially acts like a sponge and sucks CO2 out of the air. And it mineralizes, so it will stay there permanently. So imagine like, just to give you a sense of scale here, if you were to take all of Mount Everest, we're not doing anything with Mount Everest, but for a sense of scale, took all of Mount Everest, which is about three teratons of earth, and you turn that into, you know, not all of that is going to be reactive rock, etc.

• 9:32Speaker 00

  So you took that and turned it into 50-meter-tall piles the size of Honduras. That could probably get you all of the carbon removal that you need to do, period. Which is to say, that is obviously a huge amount. It's not like you're, it's not like the whole, it's a, it's a sort of size that you can wrap-

• 9:50Speaker 02

  Honduras isn't that big, we got it.

• 9:52Speaker 00

  Size you can wrap your arms around.

• 9:54Speaker 01

  Sort of.

• 9:55Speaker 00

  Sort of, yeah. It's, well, yeah, to me, I'm like, it's largely doable. And it's also-

• 10:01Speaker 01

  What I just did was the little in there.

• 10:02Speaker 00

  Oh, yes.

• 10:05Speaker 00

  Incredible. The nice thing about rocks is it can be, potentially be very cheap. We have a huge mining industry, we know how to move earth at scale, we know how to grind things up at scale, and the technology here is quote-unquote not, it's not very technology-forward. Like, the earth's been doing this for, you know, for, for, for all of time. So that's an area that I think, we think is, has a huge promise and has been very underexplored by entrepreneurs and scientists in the space.

• 10:34Speaker 02

  What's keeping us from doing more? It sounds like we have the tech, but we just need to drive down the cost?

• 10:38Speaker 00

  Part of the way that you get- We call this superficial mineralization to be really cheap is by giving it more time. So if you can trade time for cost and because the market today is largely voluntary buyers who are quite time sensitive, they want a ton in the air to be able to meet their accounting requirements.

• 10:57Speaker 02

  Right.

• 10:57Speaker 00

  You don't get a lot of export, it's not as attractive to that end market. But if in the end, we're talking about countries that are going to be the ones that are the market makers at scale, potentially this could be, they would absolutely trade time and cost for one another. And so I think there are, for that structural reason, this has been underexplored and yet it could be, you know, a ter ton of removal for less than $80 a ton. It just, there's this like weird sort of structural disadvantage to it today, but that structural disadvantage should not exist.

• 11:28Speaker 02

  Might go away.

• 11:29Speaker 00

  Yeah, in the future.

• 11:30Speaker 01

  But I'm sure you've thought about this because I know you think a lot in markets and so on, so it's a question to scratch at the depth of it. Eighty, obviously, a hugely better number than all the other numbers.

• 11:41Speaker 00

  Yep.

• 11:41Speaker 01

  So therefore worthy of focus. Why don't we set up like specifically new accounting rules that allow it to be treated as, like to have essentially a different version of advanced market commitment, like you can account for it now with this kind of commitment of what you're doing that goes over time. I mean, like, we do things like this before, like this asset can depreciate 100% year one, etc., that kind of thing.

• 12:07Speaker 00

  Yep.

• 12:07Speaker 01

  What would it take to do that and would that be a sufficient unlock for this?

• 12:11Speaker 00

  The challenge is that like the voluntary market today isn't that big anyways. So I think that that type of question is very relevant for countries and countries should be thinking about that question and they should be thinking about questions like, how do we make it such that if I buy a ton of carbon removal in Japan that is out of country for me, that definitely counts for it. So there's this whole thing called Article Six that needs to get sort of renegotiated. And I think that the accounting rules around both geography, both space and time are things that need to get included in that.

• 12:40Speaker 02

  So you mentioned, you know, companies are doing this and they're on shorter time horizons, they need to get the carbon removed this year in order to sort of meet their accounting. What do you say to people who say like, that's great, but companies are just going to emit more now because they can carbon capture later? Or like this is actually a moral hazard, you doing this and it'll just lead to greater emissions?

• 12:59Speaker 00

  Yeah, the moral hazard thing is has sort of been like the the the thing folks have been worried about with carbon removal since the beginning. And I think that the unfortunate reality is that the science tells us we have to do both now. We have to do emissions reduction and we have to do removal. So that's unfortunate, but that is the world that we're in, so we got to learn to walk and chew gum at the same time. I think in practice, that hasn't, you know, carbon removal isn't cheap enough right now for that moral hazard to actually be a thing. If we get it to be cheap to dump in the future and that opens up, that would be a good problem to have. We're not there yet.

• 13:29Speaker 01

  Probably some of our listeners will be surprised at how long it took us to get to an AI question.

• 13:36Speaker 01

  But we will now also steer into there, which is.

• 13:40Speaker 01

  So what are you, what you're thinking about how AI could most help frontier? And are there any areas where it could distract, weaken, create additional risk, et cetera?

• 13:55Speaker 00

  If AI creates the economic growth that we think it could, that increases the likelihood that we are going to be able to buy the carbon removal that we're going to need to buy. So as we talked about, carbon removal is a public good. It is going to at a hundred dollars a ton, you know, these numbers add up very fast. This is going to be hundreds of billions of dollars per year. And the likelihood that the world is going to be able to afford that goes up as AI makes the world richer. I think that that is like genuinely the single best thing that could come from from AI growth. I think the risk is that in the short term, AI requires a lot of energy. And what we build for to power, you know, power these data centers can create that energy with emissions or not with emissions, depending on what we choose to do.

• 14:40Speaker 00

  I basically got very nervous about this a couple of years ago and then did the math and convinced myself that it's probably not that big of like, I was like, should we stop working on carbon removal and go work on, you know, data centers or something like that. And long story short, I think that there will be more emissions in the short term, but long term, we will figure this out. And clean energy is going to be sort of cheaper and faster over the long term. In the short term, we will probably build more natural gas, at least in part. And but I think that that's in the short term.

• 15:11Speaker 00

  It's a sort of what's the time horizon that you're looking at? And I think that is overcomable.

• 15:15Speaker 02

  Very related. Like, the minute you mention AI, well, the minute I mention AI in New York, someone says, like, what about the energy? What about the environment? Like, you don't use AI. You're I thought you cared about. That's just like such a top critique. And so, similar to what you just said, AI companies are some of the biggest new creators of emissions. And yet, I think AI companies are also some of the biggest folks who are working on carbon removal. Is that just good PR? Do they genuinely care? Like, what do you say to people who sort of bring up that critique?

• 15:47Speaker 00

  I think that, like, the also the interesting thing about the hyperscalers is that many of them have great margins, and they also have climate commitments. And like, Google, for example, has done a lot, spent a lot of money trying to bring down the cost of clean energy, and that's fabulous. They should do more of that. So I guess I see it as a yes and, like, it is both of the things, both of the things have to be true. And if we can use the climate orientation and margins from these hyperscalers to help accelerate the rest of these clean technologies down the cost curve, that's a real win. And I think in the long run, my, maybe I'm being Pollyanna-ish, but I do think that will happen. And in the meantime, doing some carbon removal is, I think, is also a smart thing to do.

• 16:31Speaker 01

  I think you're 1,000% right on carbon removal, also, of course, in reducing emissions. And one of the things that I, you know, I answer the, let's call it the New York critique

• 16:39Speaker 00

  Uh huh.

• 16:40Speaker 01

  On is essentially that all of the hyperscalers, all of them, Google, Microsoft, Amazon, et cetera, are doing enormous amount of funding of green energy.

• 16:52Speaker 00

  Yes.

• 16:53Speaker 01

  Right? And part of what we should do is saying, look, you must be advanced, advanced market commitments, whether it's geothermal and all the rest, you know, wind, do all of that. Because then, by the way, that could, by building that tech and building that infrastructure, not only is it getting a bunch more clean power, but it funds the tech infrastructure for creating all the additional electricity, including for the thing that people don't really track, which is like HVAC systems and so forth, which is a huge contribution here. Like, everyone goes, well, AI is in the news. And you're, yeah, no, it's your hot summers, which, by the way, do come from climate change and all the rest, that then cause a huge amount of additional electricity, then, by the way, more emissions, et cetera. And that's, that's the problem we need to be most focused on.

• 17:35Speaker 00

  I think that's totally right. It's like all these positive externalities of driving down the cost of these technologies for the rest of the world that are above and beyond just the direct benefits of the clean energy from your own data center. I think that's exactly right.

• 17:46Speaker 01

  You know, AI, even as you build data centers, moves at software speed. Carbon removal, it's more geology speed, maybe a little faster than that, but not a lot.

• 17:55Speaker 00

  You're trying to make it go faster.

• 17:56Speaker 01

  Yes. What are the impedances in putting them together? What are the things that could benefit from the AI side? What are the things that break? Like, how do you, like, thinking about these different timeframes? Because, you know, part of the thing that's, obviously to Aria's earlier question, is like, thinking about this as, okay, we have, we have a compounding problem, which we need to temporarily get ahead of. And so, speed matters. Anyway, that was kind of the, the angle of this, because, you know, part of this, like, thinking about AI speed, I'm trying to figure out how to apply it usefully.

• 18:29Speaker 00

  I don't have a super satisfying answer to that question. I'm curious what you, what you would say. I think one of the, like, I think, in some sense, hard tech, like, building real things in the world just takes time. And one of the things that I worry about, maybe this is a squishier answer to the question than where you wanted to take it, is like, AI is training us all to expect everything to move at AI pace. And not everything does move at that pace, but that doesn't mean it is not worthwhile. You know, those things, there are things that take a long time to do, and you have to continue, you know, that, that, that progress compounds over years. It doesn't mean that it's not worthwhile. And I, I think that the juxtaposition between the two, because they are, and it's not just carbon removal, it's like, everything in the real world, like, those are worthwhile

• 19:13Speaker 00

  endeavors. And I think that, and my hope is that we don't abandon those, because, you know, AI sucks up everything and sort of all of our, sort of, collective attention span, as important and great as it is.

• 19:26Speaker 01

  No, I thousand percent agree. I mean, part of, you know, I frequently talk about there's a world of, of bits, world of atoms. Bits moves fast, easy, cheap, atoms is much harder. But of course, part of the problem is, we live in a world of atoms. And atoms affects everything. So you really want to be doing it. And it's part of the reason, like, I, you know, co-founded this AI company for trying to cure cancer, because it's biobits, it's kind of in between. You know, so you could, and I think one of the things you generally want to do is think about, one wants to think, including you, but it's, it's the How do we benefit the attributes of bits to help with the world of atoms?

• 20:04Speaker 00

  Yeah.

• 20:04Speaker 01

  And so like, for example, if I was thinking about the AI side, be like, well, you know, can, what are all the different ways that AI might take shots on this goal? Right? It's kind of the, it's super intelligence and, and, and it's going to invent fusion. Maybe. And by the way, great if so, but there's many things short of that, which are, well, how can you analyze the grid and make the grid more efficient? Can you make smart appliances in terms of, of knowing like your, your washing machine or your HVAC runs much more intelligently and doesn't just run the whole time? Like when you, you don't have to leave your air con on it, it, it's tied to your phone.

• 20:48Speaker 01

  And when it says, oh, you're 30 minutes out and you're coming home, I turn the air conditioning on, you know, this kind of stuff is ways of doing it. It might be ways of, of working with AI to say, what are bolder science things that might make a difference? And so it's kind of like always looking for the interface points where the bits can help us accelerate atoms. And I a thousand percent agree with you. You can't forget, we live in a world of atoms and this is what life is.

• 21:15Speaker 00

  Yeah, I like that one a lot.

• 21:16Speaker 02

  When you're thinking about this 100 year timeframe, it feels like a lot has to go right for us to get there. And, you know, we talked about, right, AI data centers are creating emissions in the short term, but AI can help with the electricity being, the grid being better, things being made more cheaply, more efficiency, more efficiently. It can help entrepreneurs, get technology to drive the cost down. What path, like, how do we get to the right path on this stuff? Is it just better technology? Is it more philanthropy? Is it more voluntary commitments from companies? Like, what is necessary for us to take the right path as it relates to AI and carbon removal?

• 21:52Speaker 00

  The thing that has plagued carbon removal to date and that is going to continue to plague carbon removal at larger and larger scales as it scales is who's going to buy it? What does the market look like? And so I think that the the way that Frontier approached this was to guarantee a market for a billion dollars worth of carbon removal. Like, how do you get to $2 billion? I guess is my question. Totally. And I I think that like eventually if the scale is needed, that demand is going to have to come from governments. And governments can put that in place in lots of different ways. But government's job, part of their job is to solve collective action problems and to sort of help manifest public goods. And so the sort of they could decide they want to buy this themselves and procure it directly. They could decide that they want to put a they want to tax the negative externality that is a ton of emissions and ask emitters to pay.

• 22:36Speaker 00

  There's lots of different policy mechanisms to use. But if you're talking hundreds of billions of dollars a year, you know, global GDP is $100 trillion. Like, it's not going to come from voluntary commitments. It has to be driven by governments. And so a lot of our work right now is focused on how do we figure out what the baton pass between the voluntary market and policy-driven demand looks like? I think that's really hard. It's really important. And policy takes years to become operational before it can actually start paying out to companies and giving them the revenue that they need. So we got to get going now.

• 23:09Speaker 00

  And, you know, some of that has been historically in the US. We're spending a lot of time in Europe and other parts of the world now as well. But, you know, this is going to be a kind of collective effort from lots of different countries. I'll say one other thing, which is, you know, half a percent or to a percentage of GDP, of global GDP on carbon removal is a lot. There is some precedent for this, though, because a lot of the OECD countries came together and were like, we're going to do this for global health. And they were, you know, the commitment was to commit 0.7% of GNI to global health. And many of them exceed that commitment. Some don't. But, you know, that that has been a very, there is some precedent for doing this kind of of public goods work at scale. And to the point of AI, if the world becomes richer, that becomes a lot easier to justify.

• 23:56Speaker 01

  So let's talk about the method a little bit. I mean, I think one of the things that's great about what you're doing with Frontier is creating a, you know, kind of a proof point on, you know, how, you know, demand side, you know, kind of commitments can bootstrap a market. And obviously it needs to scale. It's kind of like it was initially. How do you look at this as a generalizable method? And are there areas where you say it works better here, works less well here?

• 24:23Speaker 00

  So Frontier, the thing, the mechanism that Frontier is, is called an advanced market commitment. And the job of an advanced market commitment is essentially to guarantee future demand for a product if it can be successfully developed. And we borrowed this initially from vaccines. There were, the first AMC was from the pneumococcal vaccine in the mid 2000s. It was invented by a group of economists and funded by Gates and a number of other countries. And this was a sort of the first application outside of global health. There are, it's a very specific kind of market shaping mechanism and it works well, better for some problems than others. But the kind of problems that AMCs work for are ones where you have a product that you want to exist that would benefit the world in some way. And the reason that it doesn't exist is because there is some demand uncertainty, either

• 25:08Speaker 00

  because the market isn't big enough, or because the price, the willingness to pay for a product is not big enough. And an AMC is a way to sort of help fix that problem. The other thing that an AMC is good for, as opposed to like a prize, so a prize is another way one could solve that problem. An AMC is good for, and prizes actually are good for this too, but an AMC is good for the types of problems where like you don't know who the best inventor is. Like there is some information asymmetry of who could actually invent the thing. And an AMC is kind of a tech agnostic demand signal to pull forward the best ideas. I think there are lots of applications outside of this. Green steel and cement are another one. The US government should do that. The US government buys nearly 50% of cement in the United States.

• 25:51Speaker 00

  They would be a great organizer for an AMC for green, for green cement. But there's lots of other applications, but that's the shape of the problem that I think AMCs are especially good fit for. Can I just ask, like, why is Stripe doing this? Why are they committed? Well, our sort of foray into carbon removal was, it started as a small experiment back in 2019 with initially a blog post that Christian Anderson wrote, who since founded Watershed. But it was to buy a million dollars of carbon removal at any available price. And it kind of ratcheted up from there. The reaction from the market was really positive when we announced that. And we got a lot of notes from Stripe Climate users saying like, hey, I want to do something for a while on climate, but like, I'm busy. Can I just send you some money and you figure out to do with it?

• 26:35Speaker 00

  And that led to Stripe Climate, which is the software product that we built for Stripe users to make it really easy for them to direct a fraction of their revenue to carbon removal. And then that led to Frontier. We essentially were like, how do we add another zero to this? And the advanced market commitment idea was one that we sort of came out of a group brainstorm and came out of that. But that's a long-winded way of saying, this was not some like preordained thing. Right, no, no. It was a little like, we tried something and then new adjacent possibilities opened up and we kept leading into them. Yeah, very cool.

• 27:08Speaker 01

  On advanced market commitments, which I think is an excellent new mechanism. I think the parallels between health and climate that you are kind of work on are exactly right in a lot of different ways, not the least of which, which is, well, our health is dependent on the climate.

• 27:25Speaker 00

  Yeah.

• 27:26Speaker 01

  You know, it is just another, you know, kind of health initiative. Is there any way to get a bunch of forces to align to creating advanced market commitments? Like, sometimes you have a leadership thing, like, for example, Gates Foundation, you know, doing this with health and vaccine. But it's like, could, for example, we get, and you know, it's interesting, because part of the reason I'm asking this question is I think about each of the hyperscalers is kind of solving their green energy thing separately. And yet, actually, in fact, it's probably a lot more positive if they just kind of got, look, we're all going to do this, we're going to compete on our data centers, but we're going to do these advanced market commitments together, because then, by the way, you have a much broader base of entrepreneurs, other kinds of things doing it, because it's adding it all up.

• 28:11Speaker 01

  Is there any coordinate function for getting a bunch of people behind an advanced market commitment?

• 28:17Speaker 00

  So Google and Microsoft and a couple other companies, I believe, did that for some of the technologies that you described. It was geothermal, nuclear, and...

• 28:28Speaker 00

  Maybe one other, but I think some of that is naturally happening.

• 28:32Speaker 01

  Great.

• 28:32Speaker 00

  Coordinating is really hard. That has been one of the hardest parts of frontier. We have amazing partners and it's just coordinating at a company.

• 28:41Speaker 01

  Sure.

• 28:41Speaker 00

  You know, that is just a hard thing. So I'm like, my instinct is only coordinated if you have to. And so if you have huge pots of money, you, like a country, for example.

• 28:51Speaker 01

  Just start it.

• 28:52Speaker 00

  Yeah, they can do it all themselves. So I think that like, and I think that, you know, for example, the OpenAI Foundation has put in a lot of money, has a lot of money now that's sort of endowed to it. And they could probably do a lot of these unilaterally. Absent that, if you do need to get folks to coordinate, you need, you know, it's tricky. And I don't know if there's like a perfect recipe for it. You need people who are willing to run through walls to make that coordination happen. And, you know, Godspeed to them.

• 29:16Speaker 01

  One of the reasons I asked is I was thinking about like, what are the things the government can be doing?

• 29:19Speaker 00

  Yeah.

• 29:19Speaker 01

  And obviously the spend is important, public, you know, the, all the things you mentioned earlier. But on the other hand, that is, one of the reasons it doesn't happen is because it's very difficult politically. Like in a democracy, what people tend to do is, is how do I throw some money in your pocket right now? And just like, for example, you know, building up huge amounts of national debt, you know, we'll, we'll, we'll charge that to the future. And that's part of the problem that a lot of governments, you know, kind of run into. And that's one of the problems of like, no, no, we're going to spend money now for that advanced thing. But like, one of the things they could do is potentially coordinate these, like say, we'll create the incentive such

• 30:00Speaker 01

  that the, that there is a benefit in coordinating for these advanced market commitments. Like, for example.

• 30:07Speaker 00

  You want to incent the coordinators?

• 30:08Speaker 01

  Yes.

• 30:09Speaker 00

  Yeah, yeah, yeah.

• 30:10Speaker 01

  Yeah. You know, just, it's, this discussion is just making me think about that because it's another tool that the governments and policy could be thinking about as, okay, we have, we should of course be getting the spending the money, but maybe we could do this in parallel or with it. Because obviously the more that you get, because this is one of the things I like about the, like, like, you know, there's this idiocy of, oh, let's ban data centers or restrict them. It's like, no, no, no. Let's, let's encourage their construction in a way that gives us a whole bunch of benefits.

• 30:37Speaker 00

  Yeah, yeah, yeah.

• 30:37Speaker 01

  Some benefits in jobs, but also benefits in. Be developing the grid, the clean, the clean energy technology, all the rest. And then you're powering the natural commercial incentive

• 30:50Speaker 00

  Yeah.

• 30:51Speaker 01

  into these things that are good. And that you can do smartly from a government.

• 30:54Speaker 00

  I agree. I don't, I totally agree. I wonder how much of that doesn't, again, to the point of like, if you can avoid coordination, like, try to avoid it. One way that one could, like, you could, you know,

• 31:06Speaker 00

  prioritize permits for clean energy development. So like, these folks, the hyperscalers are more than, even more than price sensitive, they are speed sensitive.

• 31:13Speaker 01

  Yes.

• 31:13Speaker 00

  They're time sensitive. So like, what are incentives you could play into there?

• 31:17Speaker 01

  Yeah.

• 31:18Speaker 00

  That are kind of also free, so to speak, to, to the US government. I think there's a ton of, ton of things to be done, to be done there. The other thing that I think is nice about AMCs for, from the perspective of governments, is that the cash isn't due upfront. You only pay if you get the outcome. Yeah. And from a tax dollar perspective, that's a really nice property of these things. It's very de-risking to sort of sell that internally because you're not on the hook. If it doesn't go well, you're on the hook. And I think that, like, the, the power of Frontier was sending a billion dollar demand signal very early. We even paid out a billion dollars because those companies haven't delivered yet. But like, you know, that's, there's, there's a sort of time arbitrage that you get from AMCs that, that I think is under leveraged.

• 32:02Speaker 01

  Like, another thing they could do is say, hey, we as a government will providing a matching AMC.

• 32:08Speaker 00

  Yeah.

• 32:09Speaker 01

  to, to actors who are doing this within a certain environment, et cetera, et cetera, in terms of doing it. Like, there's a set of other ways of, I'm thinking about like amplifying. And I, by the way, look, as a fellow Silicon Valley person, I totally agree. It's like, no, no. Drive to it. Don't have a committee talking about it. Totally get it. And I, and I, I think that's important. It just what, that it also is like, if you can make a network ecosystem work, like, it's part of the thing that makes Silicon Valley work. That, that, that, some of your thinking, yeah, that's, that some of your thinking is, is, and, and work is kind of causing me to think, ah, I wonder if there's a way of putting that together. Anyway, that's the reason I was asking.

• 32:47Speaker 00

  I think the matching thing is also a, a, a interesting thread to pull on because,

• 32:55Speaker 00

  I think it can accelerate both sides. Like, the public sector likes to pull in capital from the private sector, and it makes it easier for them to sell to their constituents. And so, and the same goes for the private sector. So, there's like some nice complementary benefits of there, of that, that I think are probably underexplored as well.

• 33:13Speaker 01

  So, one of the things I love about the work you're doing is it's part of the thing I think is really important is leadership amongst all tech companies to be saying, what are the ways that we can be leading in these problems that are not just our market developments? Because when you're in a very successful industry, you should be saying, okay, great, that's successful. How do I help society? How do I help the world, etc.? And so, part of what I love about Stripe's involvement in Frontier is I think it gives us some examples about what a tech company's relationship to civilization problems might look like going forward. I'd love for you to talk a little bit more about Stripe and Frontier and the work you're doing on this.

• 33:54Speaker 00

  You know, when we look at like corporate philanthropy, for example, in the US, corporate spends about $50 billion a year on philanthropic activities. That's a lot of money. Like, that is, I think that that's something to be sort of scoffed at. And one of the things that we sort of observed in starting Frontier is that it's a bit of a weird mechanism, right? Like, AMC's are not, I guess, a traditional thing that corporate philanthropy would sort of spend time on. I think one of the things I learned from it was like, if you take kind of outsized risk, you can also get outsized return. And I, my sort of hope is that corporates can spend more of their philanthropic dollars in this way. Like, it's interesting, because in Silicon Valley,

• 34:38Speaker 00

  we apply kind of power law mentality to all of the work that we do. And yet, that, those same companies often don't always apply that to their philanthropic work. And so, my suspicion is that there are a lot of like really great outcomes that we could deliver if we were more risk tolerant. For some reason, corporate philanthropy has become like quite risk intolerant. And I think as a result, we're leaving a lot of good on the table. So, what can we do there to sort of? To sort of like permiss or encourage folks to take bigger swings. And I think that like higher than sort of unusually high risk appetites could result, not it won't definitely result, but it could result in sort of outsized returns.

• 35:22Speaker 00

  Where are we are, we've done that with Frontier and it's given us a lot of other ideas for sort of other things that we might start kind of in a similar vein.

• 35:30Speaker 02

  I also think, I mean, $50 billion is great and it's always good when companies give money. I am very pro that and you can give so much more. You can give expertise. You can figure out which of your customers can join in. You can figure out what's your tech network so you can get the other hyperscalers involved. Like there's just so much more than money that a tech company can give because of everything that it has. And it just feels like this is using all of your assets as opposed to just one.

• 35:55Speaker 00

  I think Stripe Climate is a really good example of that. Like the software product that we built to make it really easy for Stripe businesses to put money into carbon removal. This would never have worked as a standalone app. Nobody cares enough about climate to like set up their own app to like give a little bit of money to carbon removal. But if you can piggyback on, on someone's, you know, on sort of like other services that they're using and make it really easy, a single click to do this and give them that option, you know, Stripe has amazing distribution. And so that, I don't know if Stripe Climate would have worked outside of this context. But I think that's a really good example of exactly what you're saying is like, what are the other assets that we can sort of bring to bear for our problems that matter to the world?

• 36:33Speaker 01

  One of the things I think you've demonstrated is a market shaping method needs a leader, needs a champion, needs someone running it. You, I think have called that person a general manager for an orphaned problem. Lots of problems are orphaned, but many probably should stay that way. How do you tell which ones are worth a world-class leader?

• 36:57Speaker 00

  I can share what my instinct would be to that question, but I am not the arbiter of which problems get worked on. I think it's probably some combination of what is the benefit that it could give to the world? That's one. Two is, is it tractable? Like, do we think that it is possible to make progress in some set of realistic-esque constraints?

• 37:19Speaker 00

  And then the third is, am I, the person considering working on this, really excited about it? Am I, do I want to devote a lot of time to it? I think those are like three broad criteria. In practice, it's probably more opportunistic than that. Like, these waves come in somewhat unpredictable times. It's, we're not going through some sort of like stack rank list of problems and just picking off the next one. Like, you have to find, it's, to me, it doesn't, in some sense, it's not dissimilar to the private sector. It's like, all of the conditions have to be there for the right person, the right team, the right funding to come together to sort of like manifest these new organizations and general manager combos. But yeah, probably some broad, broadly, probably some combination of tractability, how big the problem is, and interest in the problem.

• 38:04Speaker 01

  And I like your 2:00 AM measure, too. Part of passion is, would you work to 2:00 AM?

• 38:08Speaker 00

  I think that's part of the third. It's like, do I want to add, do I want to do this? Is, and run through walls, because doing anything, starting anything is really hard. You've got to be, got to be ready for that.

• 38:17Speaker 02

  I mean, I think, Reid, you talk about this so often. When you think of companies, or NGOs, or whatever, things that you want to exist in the world, and you just say to yourself, who's the CEO? And even if something's number one, two, three, four on your list, like, it's not going to get done unless you have the CEO. So we might be doing things eight, nine, and 10 because we have the best CEOs for that, because you need someone who's going to run through walls. And so, related to that, we have so many problems. And, I mean, in every situation, you're like, bah, man, I wish there was an ice cream shop in my neighborhood. I wish someone would, I wish someone would start that. And it's like, what about you? And you talk about this bystander problem, like especially for these big problems. You're like, someone must be working on whatever, global poverty, global hunger, climate change, carbon removal.

• 39:01Speaker 02

  Like, do you think actually the bystander effect, like, ah, someone else is doing it, is one of the big barriers here?

• 39:06Speaker 00

  I really do. I think that's a huge, a huge problem. And it's certainly, that, that definitely describes my experience in carbon removal of... You know, you're looking at this report that is like the canonical climate change report, the IPCC report, and you're like, everybody's looking at these numbers. Everybody who cares about climate is looking at these numbers. And then you go out and talk to everybody in this space and it takes you three weeks and you're like, what is going on here? Like, who's got the ball? And I, I think that we would all benefit from a little bit more kind of institutional distrust that like, somebody else has, has this as their number one priority. And if you do, and go find that person, great. But I think we assume that that is the case and like, markets are efficient and everything is efficient more often than it actually is. Absolutely.

• 39:51Speaker 01

  Let's talk about kind of the characteristics of general managers or CEOs for problems. So D A Henderson didn't invent the smallpox vaccine. He changed the strategy, which is frequently what you're looking for, for a founder, a CEO, et cetera. What, what the data loops were, the accountability, the will. Presumably, this is at least part of the kind of leadership that the world needs on these problems and perhaps under values. How would you characterize the kind of leadership attributes? What should we try to promote and evolve? Who, what kinds of people should we be trying to support? How should people identify themselves as this kind of leader? Talk about the leadership for problems.

• 40:35Speaker 00

  I think it probably, like, the specifics of what's needed probably depend on the problem itself. Like, you're, in the case of Henderson, in the case of smallpox, it was strategy, it was execution, it was will. I think in the case of many of the topics that we're grappling with today with AI around risk mitigation and with human flourishing, it, my suspicion is that it's also one of vision. Like, what is the future that we want to build towards? And we are probably short vision. But I think it probably is, is, is person dependent. The shape of person, though, that I refer to when I think about general managers, that I would love to see working on more kind of public goods shaped problems are Silicon Valley founder types.

• 41:15Speaker 00

  It is like, I, I think that especially this new wave of funders is going to be hesitant or like, a little bit wary of funding kind of more traditional NGO folks. And we grew up in Silicon Valley and were familiar with this, the sort of execution and the hustle and the smarts and, and the strategy around, around this kind of work. And I, I think that's sort of what I. The spirit of what I mean when I say a general manager is like, these are founder mentality people. They're just working on public goods.

• 41:49Speaker 02

  Does the founder mentality break down when you think of a GM? You think about sort of one person who holds that accountability over time and they are sort of laser focused on that problem and they're sort of holding all the strings. But when you think about institutions, you think about that committee. You think about governments, foundations, international bodies. You have to have cooperation. You have to have all of that. And they're specifically designed to prevent the one person who's holding all the power. Do you think like the general manager model is like the workaround for the broken institutions we have? Or it's just a critique saying like, the current institutions we have aren't up to the challenge because they need to have more of a GM framework?

• 42:27Speaker 00

  Well, I think that current, like, a current institution also, in theory, has a GM, which is their leader, like the CEO, the executive director, et cetera. So I don't, I definitely don't think those are intrinsically or definitionally at odds.

• 42:40Speaker 02

  Opposed.

• 42:40Speaker 00

  I think the difference is that there are a lot of problems that have nobody who thinks that it's their problem to solve it. That is thing one. And I think there are a lot of institutions that are too narrowly scoped, that have not scoped themselves to deliver a specific outcome, rather than, and they're working on a slice of the piece. And that's okay. We don't need every institution working on the whole thing. But I would love to see more institutions saying like, here's the outcome that we're going to deliver and we're going to pull whatever lever that we need to in order to make that happen.

• 43:15Speaker 02

  It's sort of like when these institutions tout, like, we've been working on this problem for 40 years, 40 years of experience.

• 43:21Speaker 00

  Is that good?

• 43:22Speaker 02

  What? Well, is it done? Did you fix it? I guess, you know, you don't know the counterfactual, maybe we would have backslid and now we're at even. But, right, we wanna, if you've been working for 40 years, let's see some results.

• 43:33Speaker 00

  Yeah. And there are lots of great institutions out there. I think that, I think that they're The number of orphan problems and I think the number of amazing people that could be working on this shape of thing to me seems far, far lower than it should be.

• 43:49Speaker 01

  So this leads to a natural question, which is if you could assign one world-class general manager to an orphan problem tomorrow, other than carbon removal, because we've already talked about that in length, what problem would you pick?

• 44:03Speaker 00

  I can talk about one that I have gotten nerd-sniped by basically this time last year, which is that of respiratory infections. You know, most otherwise healthy people get sick two to four times a year. We spend 5% of our lives sick with respiratory illnesses. It is very bad for the economy, probably like 500 billion to a trillion dollars per year due to lost productivity. These things are probably really bad for us in the long run. We just like haven't done a great job characterizing that. And yet, and they also periodically threaten civilization through pandemics. Like, I think that like, these are just bad. They are bad. And they are also one of these problems that is so, we just accept it because it is like part of life and like, oh, your kid's sick and my,

• 44:46Speaker 00

  you know, where everyone's, that's fine. It's annoying, but we tolerate it. But why? I think that this is the kind of thing, it is really hard. I don't want to minimize that. There's a lot of viral diversity. They mutate a lot. But the tools have gotten much better in the past couple of years. And I think that this is increasingly a tractable problem. So I have been spending a lot of time sort of thinking about what would GMing respiratory infections look like? If the goal is to make respiratory infections a thing of the past, could we do that? How would we do that? Is it possible and over what timeframe? I think that this is, has definitely been something that's been on my mind.

• 45:22Speaker 02

  How do you start to figure out that it's tractable? What gave you the confidence to be like, you know what? Like, ah, we could do this?

• 45:29Speaker 00

  Okay, that's a great question. So last spring, when I was sort of, I initially got into this topic because I was talking to David Weesler at the Institute for Protein Design. I was thinking about some of the sort of incentive structures that we could do to fix this. But I'm not, this is not my sort of technical area of expertise. So we ended up hosting a symposium at Stripe with sort of 40 people, largely scientists, but also policymakers and folks from pharma companies. And that was, that was a big part of the question. Like, is it possible? Is this technically possible? And so we had, we flew all these people in and they were giving these sort of seven-minute lightning talks. And it was a day and a half of just intense focus on this question. And we left basically being like, This is hard, but it probably is. And it's not definitely, but to the point of

• 46:13Speaker 00

  you don't get big outcomes if you don't take some risk.

• 46:17Speaker 02

  You have enough confidence that this might be.

• 46:19Speaker 00

  Yeah, we have enough confidence this could be a thing and there is some risk, but like you got to swing to to sort of get the get the benefits. And so that gave us enough confidence to ultimately spin up a new frontier shaped initiative that is going to be focused on trying to make respiratory infections a thing of the past.

• 46:37Speaker 02

  That's amazing. I was sick so much this year. That's incredible. Okay. So you just put out a Substack on the new philanthropy that I was obsessed with and saw so many people chattering about. And one of the things you discussed is all seven anthropic co-founders have pledged to donate 80% of their wealth, which will be potentially $37 billion conservatively. We'll see. And that's 10 times what the world's largest research driven grant maker has given away in its entire history. And so we had the last, you know, you talked about sort of three waves and we had the Rockefellers in terms of philanthropy. And then sort of the next wave was the early 2000s tech philanthropy, Gates, Moskowitz, Zuckerberg, I might say Hoffman. Does that offer a useful model for

• 47:22Speaker 02

  what's going to happen next? Or why is this next wave different?

• 47:24Speaker 00

  Okay, just in terms of scale, I would even like broaden it outside of the bucket that you mentioned. There's sort of like these, these, there's the anthropic founders and their anthropic employees. And OpenAI has, you know, put initially 26% is in the OpenAI Foundation. All of those numbers add up to a lot of money. Do some math, it's probably around, you know, 37 to 100 billion per year. That's kind of like what we're looking at in terms of, in terms of scale. Global and US philanthropy is about $600 billion a year. So it's a, it's a lot. It's not doubling that. I think that the reason that this is, part of the reason that this is different is, is a couple of things. One is the shape of the problems that these funders care about. These funders are, you know, rightly worried about the AI transition. And there are sort of two components of

• 48:06Speaker 00

  that. There is how do we navigate all the minefield of risk to get to the other side? And if and when we do, How do we help humanity flourish from that? And so I think there's a problem area, problem set that they're especially interested in. The second is sort of core belief is one around speed. Like if AGI is coming quickly, we need to get to work very fast. And then the third sort of difference here, and I don't think this third difference is actually that different than the last wave, but it's really around talent and what are expectations of the types of people that we, that sort of they would be excited about funding and we'd be excited about funding. I think it's the combination of these three things in my mind that makes this wave potentially hugely impactful.

• 48:51Speaker 00

  So $50 billion a year, if you look at some of the projects that they think this group would be excited about funding, $50 billion a year would fund 333 ARC institutes per year, fully fund their budget. It would fund 5,000 Institute for Progresses. There's a huge, you know, the potential for impact here is huge. And as I was doing the math to write this piece, the thing that stuck out to me was just like, we don't have the organizational capacity to absorb that money. And so what are we going to do to fix that?

• 49:19Speaker 01

  Well, let's talk about one part of that. We'll start with one part of that organizational capacity, which is capital allocation. So what do you think it would take to get to an effective capital allocation kind of layer and manager? What would be the, you know, kind of a program manager, is the venture capitalist, the kind of equivalent of the capital allocation? And what is their capabilities and skill set look like? And then what does, what will they need to do to be able to be good stewards of this particular fund of capital? And then what kinds of things will they need to be doing?

• 50:01Speaker 01

  Because they'll have to probably be developing their markets some. What would that look like as well?

• 50:06Speaker 00

  At the risk of overfitting everything to Silicon Valley, which is, you know, where we're at.

• 50:12Speaker 01

  I do it all the time.

• 50:13Speaker 00

  Yeah. I think that this essentially is a sort of Silicon Valley for public goods is what this could look like. And the components might look quite similar. You have funders and then you have capital allocators, which are kind of like philanthropic VCs. And then they're giving money to, so call it like philanthropic startups. And there are some of these capital allocators today. The OpenAI Foundation is definitionally going to be a huge one. Coefficient Giving is another good example. Renaissance Philanthropy. But there are You know, $50 billion a year, if you're giving away, you know, on average coefficient givings grant last year was $1 million a year, that is, you know, 50,000 grants that you got to distribute in a year. That's just a very large number. And so we probably need many more of these philanthropic VCs, so to speak. And we also need many more of

• 50:57Speaker 00

  these philanthropic startups. So again, I think that from a who do you want working on these, it's probably not dissimilar to the kinds of folks who are VCs today. It's identifying excellent talent, figuring out where is this space going and how do I make sure that there are great people working in each of those. And I think that that is also true on the startup side. It's like, again, looking for these GMs, looking for these founders, helping them get into this space, find their problem, and start an organization. The thing, you know, again, I ended up starting, like I did this math for myself. I was not intending to write a piece. And then I was like, oh, my God, who is going to do this? You essentially, the main takeaway is like, we don't need just a few more people in philanthropy.

• 51:39Speaker 00

  We don't need just a few more people doing capital allocation or working on public goods. We're talking about thousands of organizations that we need, and as an implication, thousands of founders, and probably hundreds of thousands of people working to power those organizations. This is, this is a pipeline sort of workforce reorientation problem. And how exciting, you get to be working on some of the problems that matter most for humanity. I think that that's a really, it's a very exciting time for that reason, in my opinion.

• 52:07Speaker 01

  One of the things I thought was an important thread in your piece was that this capital will be coming with expectations about what kinds of founders, what kinds of projects will be good. For example, I could easily see, you know, like, for example, universities saying, oh, great, this is more endowment for us and so forth. And they're unlikely to be understanding where the folks who have put the work into building this kind of thing, what they think about an AI universe, etc., you know, how that applies. So go into a little bit more depth about what should people who think, oh, I'm going to come help solve this capital allocation problem, because the, I wouldn't say the advanced commitments,

• 52:49Speaker 01

  potentially becoming advanced market commitments of we're doing this in the philanthropy will only get unlocked in certain kinds of projects. And I thought that was one of the important threads in your piece.

• 52:59Speaker 00

  I think the expectation for leaders of these organizations is that they are sort of tech caliber people. These are, you know, PMs and otherwise founders who are working on problems that are good for the world. And the counterfactual, if those people don't exist or don't move into public goods and don't start organizations, I think is that the money isn't going to get spent. And so it is on us to sort of like figure out what that pipeline looks like. What is the PM to GM pipeline? How do we create these talent onramps and high status onramps? And for great people, help them find a problem that they are really excited about and can

• 53:42Speaker 00

  wrap their arms around and want to work on for long periods of time. And we have to do that at a pretty large scale. This is a sort of wartime mobilization effort to get great people working on problems that are good for the world.

• 53:53Speaker 02

  It's interesting because when you talk about the money might not be spent, I wish I had the numbers in front of me, but I'm sure there are like, there are more than, let's call it, you said, you know, 37 to 100 billion dollars over 10 years. There is more than that much money already on the sidelines right now that isn't being spent in DAFs, in foundations, in whatever, that is allocated for charitable. And so this isn't actually just a problem of a problem of the new AI money that's coming. Like, we literally have this much money already that's not being allocated for many reasons. And some is because they don't feel like there's a good place to put it, they feel like philanthropy is not meeting the moment. But then there's other critiques people are, don't want to give because, you know, the OpenAI Foundation would get yelled at for giving to the wrong things.

• 54:36Speaker 02

  Like, there's all sorts of reasons why the capital that we have isn't being deployed philanthropically. And certainly, one of the things you talk about is how do we get, how do we get all these amazing PMs? How do we get these, you know, you're essentially a philanthropic product manager right now for these social good problems. How do we make this the career pipeline for the best people? Like, what needs to change in order to push them into these new fields?

• 54:58Speaker 00

  I think there are a bunch of things that need to change, to be honest. I think the first is, we have to make the problems that need to get solved more legible. So I would love to see Coefficient Giving, OpenAI Foundation, other folks who have vision in the space, especially who are thinking about AI, make the problems and the vision for those, for how to solve those problems, more legible, so that somebody can learn enough about a problem to get nerds sent by it. And the second thing is, we want to create high status on ramps for great folks to work on philanthropy. I sometimes am surprised I work in philanthropy. Like, that is not where I expected to be. And I think that, I think the way I sort of rationalize myself through that is like, this is a very different shape of philanthropy than sort of traditional philanthropy.

• 55:42Speaker 00

  But I think a lot of people have, philanthropy has that baggage for a lot of people. And so I think we need to shift that. And so, you know, I was thinking like, we should, who, who, what does YC for? Like a, a, a sort of talent and startup accelerator for public good shaped problems? That would be really cool. I think another thing that we need to fix is pay in philanthropy. I think we've, sort of, it's been gauche to talk about. We haven't really done, made much progress on this. But in theory, if these are really big problems for the world, we should be excited to pay people who fix those problems, a lot of money. So I just, and I think a unique property at this moment is that the funders, where this money is coming from are some of the sort of most coveted

• 56:21Speaker 00

  companies in the world, like Anthropic and AI have, you know, it's all, this hypothetical money is sitting in equity right now. And I think that's never really been the case before. So can we figure out a way to sort of give people incentives and bonuses, et cetera, in the form of Anthropic or OpenAI equity? I think that, that the pay, the pay piece feels like uniquely, uniquely tractable. I mean, I have not looked into it that much, and I'm sure that this is a deeply naive thing to say. But I, I, I want to chase that down. That feels, that feels tractable to me.

• 56:50Speaker 02

  I'll say two things on that, which I think are interesting. I feel like people are like LeBron James, get rich playing basketball. Beyonce, sell some records and get rich. Don't you dare solve childhood hunger and get rich. Like, how dare you? Like, no, no, no, you should be living in poverty wages. So I do think some of it is that.

• 57:08Speaker 01

  Says the former DoSomething CEO.

• 57:09Speaker 02

  Yeah, uh huh, that's right. I also do think it's interesting, because this has happened in so many industries, that right now, the nonprofit sector is mostly women. And it's actually more diverse, certainly, than Silicon Valley. And it's low paid. Actually, 80% of nonprofit CEOs are men. But the rest of them are women and more people of color than Silicon Valley. And so I do. I think it's going to be interesting if all of a sudden we have higher salaries to solve global problems and it's all going to white men. Like, all of a sudden, it's tech people who are more sort of white men, Asian men, et cetera, just, you know, by demographics, and then we give them the higher salaries. Like, I'm also imagining a, you know, some, some people complaining about that. Some people saying like, what the hell? I've been

• 57:52Speaker 02

  working in this forever and I haven't afforded that, or I could have done better if I was able to get that equity or was able to afford this. So I don't think you're naive. I think that is a huge piece of the problem. It's money, and then it's also status. Like you said, you didn't picture yourself being that way because it's higher status to be, you know, sort of a startup at a for-profit than a nonprofit. But I do think there's some other pickups in there. Yeah.

• 58:15Speaker 01

  You, you wrote something that had some parallels to some of kind of talks I've given in Italy, which was Silicon Valley, as you know, kind of styles itself as a new Renaissance. And one of the things that I've wanted it to do is to upgrade the way that it thinks of the full Renaissance. Like, how do you bring art in? How do you bring humanism? Not just the kind of central invention of technology, science, creating new networks and, and roads for commerce and all the rest. And so you wrote that San Francisco might be entering a romantic revival. More beauty, emotion, real-life spaces, maybe less optimization.

• 58:54Speaker 01

  And so I think of this potentially as a humanist frame. What goals should technology ultimately serve? Not just what can it do? And is romanticism pointing out a humanist outcome? Pointing out something different than a humanist outcome? How do you, how, what, what made you right about the romantic revival?

• 59:19Speaker 00

  I've lived in, in San Francisco for over a decade now, and I have felt a vibe shift recently. This was, this piece was not backed by data. It was purely anecdotal on sort of how I have felt this city has been showing up now versus a decade ago. I'm curious what your experience here is. And I can't tell totally what's a reaction to. Maybe it is a reaction to sort of AI is taking off, so what does it mean to be human and wanting to embrace our humanness? Maybe there are sort of strands of nihilism in there of like, you know, and sort of a more hedonist origin, but SF has felt like it's starting to care more about aesthetics.

• 1:00:00Speaker 00

  It's starting to care, talking about religion feels more acceptable now than it did 10 years ago. Starting to like care about how you dress feels acceptable, if not cool, in a way that 10 years ago would have been like you're spending on time on sort of the wrong things. I don't really know what to make of it, but it does feel like a shift to me. What has been your observation in SF?

• 1:00:24Speaker 01

  I don't have as much direct experience with SF. I tend to be more of a peninsula person, but I do.

• 1:00:29Speaker 00

  Bay Area.

• 1:00:29Speaker 01

  Yes, Bay Area, but in Go there. I don't know if I'm seeing it yet. And part of the reason, like, you know, giving these speeches at Perugia and Bologna and other kinds of things, I would like to see more of a explicit cultural focus on kind of humanist virtues. And by the way, it can be how we live our lives. Like, do we go out for walks? Do we, do we have intellectual salons? Do we, you know, I still, maybe I swim a little bit too much in the YouTube short culture of, of like, you know, like too much discussion of peptides, you know, blah, blah, blah, blah. And you're like, okay, right?

• 1:01:09Speaker 01

  And so, like, and it, like, like, it was funny, we, Ari and I were at a dinner last night and one of them was like, I've recently discovered the Brothers Karamazov. And you're like, well, good.

• 1:01:26Speaker 00

  Been around a lot.

• 1:01:26Speaker 01

  Yes, it's been around, but, but, but it, like, that could also be looked at as a positive, like encountering more of that. Like, what is the human condition? What should it be? How should we be living our lives individually and together? And more woven through everything, including, of course, your work. And that was the reason why the, the Romantic Revival piece spoke to me a little bit as a hope. But I don't know if I'm actually seeing anything. And I get your data point. I mean, I'm not quite sure how, I guess we could have kind of run a poll or something, but me?

• 1:02:00Speaker 00

  What, what do you think is driving that?

• 1:02:02Speaker 01

  To the extent that it is happening. I do think your point about AI. There's this old book from railway tech revolutions called Megatrends by John Nesbit. And the very first megatrend was high tech, high touch. And as you get to high tech, which is alienating, you tend to get to high touch, which is more humanizing. So you tend to get to like, oh, no, I want something with a wood surface. I want to, I want to, you know, and it's one of the reasons why there is a frequent attraction to Japanese aesthetics within this. You want a record, but you just put on a turntable. Yes, exactly. So AI music on a record, precisely. And so, so I think there's at least some of that.

• 1:02:43Speaker 01

  And I think part of the question kind of comes down to how do we make these things that we're doing, which are trying to make a difference to humanity at scale, but how do we make them human too? I think there's at least threads of that. Now, there's also weird other threads, you know, transhumanism, like, you know, I've lost track of the number of times that people have blithely told me, we're going to be uploading ourselves into robots. And you're like, you have no idea what you're talking about. Right? Like, it's like, what is that? I mean, it's a little bit like people going, well, I understand the Star Trek transponder, like transporter. It's like, I get in here and I disappear and I appear over there. It's like, well, okay, look, there's two framings. One is you're disintegrated and there's a clone.

• 1:03:27Speaker 01

  And the other one is you're transported. You really should have a theory about why it's one and not the other one. You know, and so, and it's kind of a similar kind of thing, but that dialogue, and it's also, I think, so I think that's part of the thing that's hitting us is that the technological environment where, you know, okay, what happens when we get to AGI, et cetera, et cetera? And what would that mean? And what would that mean to be human? And what does that mean for work? And what does that mean for meaning? And, you know, a bunch of other questions. But I think it's also the scale of the tech industry. Because, you know, when. Twenty, thirty years ago, we were all happily the pirate challengers to institutions.

• 1:04:07Speaker 01

  And now, we're much more like, well, none of the things we do affect the entire world.

• 1:04:14Speaker 00

  Right.

• 1:04:14Speaker 01

  And so you also, you know, it's, I think it's a simple line that comes from Voltaire, but with power comes responsibility. It's one of the things, so you've got to like, okay, now start being responsible for this. And by the way, responsible is not, well, I have a theory. It's like, great. Your theory should be in dialogue with other people. Because, by the way, you have a theory. I am the one God person who knows everything. Oh, you have a problem.

• 1:04:38Speaker 00

  Yeah.

• 1:04:38Speaker 01

  Right? So it's like, no, no, be in dialogue with people, kind of find out like, okay, what are other perspectives and what kinds of things should be shaped? And that's the kind of, that was the reason why when like Bologna called me and said, hey, would you do a commencement speech? I think they were surprised when I said yes, because I'd been thinking about this.

• 1:04:53Speaker 00

  Right.

• 1:04:54Speaker 01

  Right? And I was like, no, I think this is important to do. So I'm hopeful. I'm not sure how strong the signals are yet within the valley, but I would love to see it grow. So we have a section called Rapid Fire, which actually you can answer at any depth that you like, because it's really just the questions around that we ask every guest, because it kind of creates a, like a tapestry across the different possible podcasts. So is there a movie, song, or book that fills you with optimism for the future?

• 1:05:27Speaker 00

  There's a collection of essays and talks by Richard Feynman called The Pleasure of Finding Things Out, that I, I'm not really sure how I discovered them in the first place. But whenever I feel like I need a like, mental tune up, or a reset, or like kind of a, yeah, a dose of optimism, I find spending fifteen minutes with one of those essays to do the trick. So they're really fun, you can kind of bounce around a bit. But it's like, they're ostensibly about science, but really they're about a lot of things about life. They're about, you know, curiosity and mischief and, and learning things and, and sort of growing up. And I think that they're, they're a real joy. So highly recommend.

• 1:06:10Speaker 01

  I love giving Feynman as an answer.

• 1:06:13Speaker 00

  Have you read those?

• 1:06:14Speaker 01

  I haven't, but I've read other files.

• 1:06:15Speaker 00

  I'm gonna get, yeah, I'm gonna get you that book.

• 1:06:18Speaker 02

  What is a question that you wish people would ask you more often?

• 1:06:21Speaker 00

  I find that the questions that for me, there's like a category of questions that are, are, I are, tend to be very useful for me. And they are sort of like constraint lifting questions. So, you know, versions of like, what is the 10x version of what you're working on? Or if you had, you know, if you, you're, you're working on something and someone's like, what's the, you know, if you had $10 million to spend on this or 100 million? And not to say they're giving you that money, but I find that having other people ask questions that sort of permits you to think more broadly are sort of surprisingly useful for me. I think Patrick's really good at this, actually. There have been a number of things where, a number of times where he's sort of asked,

• 1:07:05Speaker 00

  he sort of like gently asked these like nudging questions and just the asking of them permits you to think more, more ambitiously or more creatively. So I, I, I love getting those and try when I can to, you know, ask them to other people as well.

• 1:07:19Speaker 01

  Well, somewhat this leads us to our third rapid fire question, which is, so where do you see progress or momentum outside of, say, your areas, which are climate, markets and AI, that inspires you? So where outside of that?

• 1:07:34Speaker 00

  We're seeing a ton of progress right now in biology. I think that that's a really, really exciting area. I think in another world, I would have, I, I wish I'd been a biologist. I think AI is driving a, you know, is of course related to AI is driving huge advances in our understanding and also what we can do with that understanding. And so like, you know, every day on, you know, Twitter or reading, reading different pieces. It's like, we live in an age of miracles. Like what a time to be alive. I think the pace of progress is staggering. And of course, this is going to run up against the real world and we got to do clinical trials and we got to do, you know, we got to turn the, the bits into atoms, so to speak, but. It's very thrilling to watch.

• 1:08:09Speaker 02

  Absolutely. Can you leave us with a final thought on what is possible to achieve in the next 15 years? And, oh, if everything breaks humanity's way, I forgot that piece, and what's our first step to get in that direction?

• 1:08:22Speaker 00

  I think the next, like, the next five, 10, 15 years are going to be very consequential for humanity, right? And, you know, without the risk of sounding too AGI-pilled, there are these two sets of questions that we're going to need to grapple with. One is, how do we navigate the AI transition and manage all of these risks? And two, if we can do that, what does human flourishing look like? The next, you know, 10, 15 years are going to matter a lot in answering those questions. And I think there are, you know, we've talked about many of the things that need to get done, but I think a good first step is taking some of these, like, newer and squishier ideas and trying to sort of more crisply define what are potential visions for them. Like, what, human flourishing is a word that gets thrown around a lot, but what does it really mean? What does it look like? What does it feel like?

• 1:09:07Speaker 00

  How do I imagine it? And get me excited about what that future could look like so that I want to quit my job and go and work on it. I think that we are short vision and, and sort of really thinking about shaping what those visions could look like is a really concrete and possible and tractable first step towards driving all the attention and resources that we need to go and actually make the world look like that.

• 1:09:28Speaker 02

  Awesome. Thank you.

• 1:09:30Speaker 01

  Great. Thank you guys so much.

• 1:09:32Speaker 02

  Pleasure.

• 1:09:32Speaker 01

  This was really fun. So fun. Possible is produced by Palette Media. It's hosted by Ari Finger and me, Reid Hoffman. Our showrunner is Sean Young. Possible is produced by Tenasi Delos, Katie Sanders, Spencer Strasmore, Imo Zhu, Aman Suri, Lexi Kevin, Danny Garrison, Trent Barboza, and Tafadzwa Nima Rundwe.

• 1:09:52Speaker 02

  Special thanks to Surya, Yala Manchili, Sayida Sabieva, Ian Alice, Greg Beato, Parth Patil, and Ben Relis. And a big thanks to Freddie Williams, Florian Maganza, and the team at Greylock Partners.