November 25, 2021
What kinds of catastrophic risks could drastically impact global food supply or large-scale electricity supply? What kinds of strategies could help mitigate or recover from such outcomes? How can we plan for and incentivize cooperation in catastrophic scenarios? How can catastrophic and existential risks be communicated more effectively to the average person? What factors cause people to cooperate or not in disaster scenarios? Where should we be spending resources right now to prepare for catastrophe? Why does it seem that governments are largely uninterested in these questions?
Dr. David Denkenberger (also known as 3D) received his master's from Princeton in Mechanical and Aerospace Engineering and his Ph.D. from the University of Colorado at Boulder in Building Systems (dissertation on his patented heat exchanger). He is an assistant professor at University of Alaska Fairbanks in Mechanical Engineering. He cofounded and directs the Alliance to Feed the Earth in Disasters (ALLFED) and donates half his income to it. He received the National Science Foundation Graduate Research Fellowship, is a Penn State distinguished alumnus, and is a registered professional engineer. He has 73 peer reviewed publications and is the third most prolific author in the existential and global catastrophic risk field. His work has been featured in more than 25 countries in over 200 articles, including articles in Science.
JOSH: Hello, and welcome to Clearer Thinking with Spencer Greenberg, the podcast about ideas that matter. I'm Josh Castle, the producer of the podcast, and I'm so glad you've joined us today. In this episode, Spencer speaks with David Denkenberger about catastrophic climate change events, abrupt infrastructure loss, and scaling resilient food production in the aftermath of these circumstances.
SPENCER: David, welcome.
DAVID: Thanks for having me, Spencer.
SPENCER: So a lot of people underestimate, I think, the chance that civilization could have a really major setback. There's a giant war or extreme climate change, or solar flare, or all these different sorts of worldwide events that could put civilization into a sort of uncertain state where it's not that clear that we can survive. And I know you've done just a ton of thinking on these topics about how do we make sure that if such a terrible thing happens, we can keep civilization going, we can kind of reduce the damage that occurs. So I want to dig into that topic with you today.
DAVID: Sounds great.
SPENCER: Now my understanding is that you kind of divide these catastrophes into two categories. You want to just talk about those two categories for a moment.
DAVID: So one of the ways that crops can be killed directly is through climate change. And this could be over a century timescale, or it could be more abrupt. There could be a breakdown of the overturning of the ocean and abrupt cooling over a decade in Europe. Europe actually has lost 10 degrees Celsius or 18 degrees Fahrenheit in just one decade in the past, and that could potentially happen again.
SPENCER: When did that occur?
DAVID: That was about 11,000 years ago. Another risk related to climate change is the extreme weather on multiple continents. This could be floods and droughts. This could reduce global food production by about 10%. There was actually a UK government study that estimated this had a chance of about 80% of occurring this century. Then, there are other climate changes that could be caused by a volcanic eruption. There was an eruption in 1815 that caused the year without a summer in 1816 that actually caused famine in parts of Europe. There could be an asteroid impact; though that's less likely. There could also be a nuclear war, and that would cause the burning of cities, and smoke would go into the upper atmosphere and could reduce sunlight for a decade or more.
SPENCER: Is that the so-called nuclear winter scenario?
DAVID: That's right.
SPENCER: Got it. And so these ones you're talking about all have in common is that they could rationally affect the food supply and basically put just the very concrete like day-to-day survival of humanity at risk, is that right?
DAVID: That's right. And then there are other catastrophes as well, such as if there were a crop disease that could kill many different types of crops. I'm particularly worried about the grass family, which is not just grass that animals eat, but is also wheat, maize or corn, rice, sorghum, barley; even sugarcane is all part of that family. And it makes up something like 70% of our total food supply.
SPENCER: So interesting. Are there other kinds of challenges to crops or the food supply that you want to talk about?
DAVID: There could be a super crop pest, something that eats the crops. In fact, in the last couple of years, there has been a very severe locust outbreak in parts of Africa and also spreading into the Middle East, but potentially something could be even worse than what has happened historically. There could also be a weed that outcompetes crops and is somehow resistant to other methods of control.
SPENCER: Got that. So now that we've stressed these out to our listeners, let's dig into some of the possible strategies or solutions to help make it more likely that if one of these really bad things were to occur, we could actually keep civilization going and keep feeding people. And then we'll come back later to the second type of calamity, one that has to do with electrical grid. Sound good?
DAVID: Sounds great.
SPENCER: All right. So huge catastrophe occurs, like how do we actually keep people fed?
DAVID: Well, there are some pretty conventional responses here that we could do. We could focus more on increasing crop output with whatever means we have, we could try to waste less food, feed less food to animals, turn less food into biofuels. And these are all good, but they might not be sufficient depending on how severe the catastrophe is. So some of these are what we call the around 10% food production shortfall. It looks like we could handle it technically by reducing waste, etc., but the food price could increase so high that hundreds of millions of people might not be able to afford it; and that could potentially cause cascading consequences.
SPENCER: Right. So let's talk about some of the consequences. Obviously, there's a lot of starvation. You know, there's a lot of direct harm, but I imagine it also would destabilize society in many ways.
DAVID: There's a lot of uncertainty, what could happen. But yes, there certainly could be a massive migration crisis, as tensions are high and resources are limited. I think conflict is more likely. So it really could cascade into something like nuclear war.
SPENCER: Okay, so maybe if there's a small food shortage like 2%, or something, maybe we can use sort of these more commonsensical strategies, reducing our feeding animals, things like that, and maybe we can get through, but what if it's a much larger food shortage?
DAVID: So certainly, if the sun is mostly blocked, like supervolcanic eruption, or the asteroid that killed the dinosaurs, or full-scale nuclear war - involving many nuclear warheads - then there would be basically collapse of the agricultural system. So we need to have some other solutions. And people have proposed, well, we should just store up food. And that would work technically, but if you want to store up food for a billion people over five or 10 years, it's just 10s of trillions of dollars. And it would take a long time. So it wouldn't protect you from anything that happened in the next few years. And if you want to do it fast, you would increase the food price now, so you would make current malnutrition worse.
SPENCER: So what do you suggest, then?
DAVID: Well, the history was when I started looking into this problem in 2011, that I was reading this paper called Fungi and Sustainability. And they observed that after the dinosaur-killing asteroid, there were lots of dead trees, and there were tons of mushrooms growing on the trees. And they said, “Well, maybe when humans go extinct, the world will be ruled by mushrooms again.” And I said, “Wait a minute, why don't we eat the mushrooms and not go extinct?”
SPENCER: It seems like a pretty good idea. So what makes mushrooms appealing?
DAVID: Well, they can grow without sunlight. And they also can have a billion spores, which is their equivalent of seeds. And that means they can be scaled up very quickly. But even though I started with mushrooms, they are a relatively expensive source of calories. So we're actually focusing on other so-called resilient foods that we think could be more economical and actually be afforded by nearly everyone.
SPENCER: So what are some of these options?
DAVID: One of the options is converting fiber or cellulose into sugar. This has already been done in the case of what's called second-generation biofuels; where instead of taking corn, you take the corn stock, and you break that cellulose into sugar. Then what they do is they turn the sugar into ethanol, which can substitute for gasoline or petrol. But if what we really needed was food, we might be able to interrupt that process at sugar and eat the food. There are actually a couple of companies that are looking at doing this specifically for human edible food, but they're not looking at how to scale it up fast in a catastrophe. So we're interested in trying this out, and we've identified a repurposing opportunity of paper factories. They actually have about 80% of the equipment required to turn wood into sugar. So we would like to try out repurposing an actual factory to see if it can be done.
SPENCER: That's really a cool idea. Where could this material actually come from after disaster that would then be converted into sugar?
DAVID: Well, we would have a limited supply of leaves, and what's called agricultural residues, the leftover from producing food. But it turns out, there's a tremendous amount of wood in the form of dead trees after a catastrophe. So that could actually last us many decades.
SPENCER: So essentially, you go harvest that kind of dead for us, turn them into food.
DAVID: That's right. Another opportunity is actually turning natural gas or methane into food. There's actually a type of microorganism that eats natural gas. That microorganism is rich in protein. There are actually a few companies that are commercializing this technology to produce fish food.
SPENCER: Assuming we're still able to get natural gas, we could just keep getting it out of the ground and then use that as a food source.
DAVID: That's right. And in any of these catastrophes, there would likely be significant destruction of infrastructure, but it's unlikely to be global. So we would be able to use the industry in the unaffected countries to be able to produce these foods.
SPENCER: So how are you imagining this would go? Is the idea that governments would fund having these kinds of manufacturing plants or whatever, we would need already to go or some other way can actually be prepared when disaster strikes?
DAVID: The ideal here would be to have the factories already in place, maybe they're producing animal feed, and then we could rapidly switch to human food. However, that is a bigger ask. So what we're focusing on now is what could we do with a lot less money, say hundreds of millions of dollars, to do the research and do the piloting, to show this can be done perhaps at a little smaller scale, and then also do the preparedness plans, so that governments, corporations, international organizations know what we should do when a catastrophe hits.
SPENCER: You say hundreds of millions of dollars, and obviously, from the point of view of a nonprofit, that's a ton of money. From the point of view of a government, that's not that much money, especially like the US government. But I assume your organization doesn't have hundreds of millions of dollars to spend. Is that is the idea that you would basically get governments to invest in?
DAVID: Right, we're trying to do at the Alliance to Feed the Earth in Disasters or ALLFED, we’re trying to do the research. And we're going to be doing geographic information systems or GIS analyses to map out where these resources are to produce the resilient foods and then produce a very rough plan for governments or companies. But then, they can use their much more detailed information to make a better plan.
SPENCER: It seems to me that the government would have to be involved in some way because let's say you did have these plants that were operating and producing pet food or whatever, presumably, you don't just want them to be at the whims of the market forces, right? You don't want them to go bankrupt, or be like, “Oh, actually, we're gonna produce this other kind of fish food.” Right?
DAVID: The other option here is even if we don't have these factories already built that could be repurposed to switch to food production, we might be able to quickly construct these factories. So we've looked at how much faster we could construct them than normally, say, with round-the-clock construction.
SPENCER: It seems like in a disaster scenario, it's sort of hard to count on having the right skills available, having the right materials available.It seems like it'd be much better to just have them already in place now.
DAVID: It certainly would be better, it's just then it would be a much larger investment sum.
SPENCER: Right, whereas if you could just pull out a plan and say, “Okay, we have our plan, we're going to build these factories in the event this happens.” How costly is it actually to feed people if you use these highly efficient methods - like if your goal is not people enjoying their food, but literally just survival - what kind of cost per meal or cost per day are we talking about?
DAVID: It's surprisingly low-cost. And so it's in the ballpark of one or two or $3 a day, which is surprising. But here, we're talking about, it's still more expensive than bulk grain is now. That's why these things are not being done now. But yeah, we're focusing on the most promising sources. And I would add a few more - one is, even in a severe super volcanic eruption scenario, the sun would not be completely blocked. So we're looking into growing seaweed, which has already been done at a large scale for things like sushi, but seaweed can grow 10% per day. So it could be scaled up very quickly. It's a less capital-intensive process, in that you mainly need to produce a lot of rope to attach the seaweed to and some buoys; you can just use empty bottles. Then you attach the ropes to the bottom of the ocean, where it's pretty shallow, and you can produce a lot of food.
SPENCER: So the seaweed is using the light from the sun. It just doesn't need that much, does it?
DAVID: That's right, it grows well in low-light levels. That's why it grows many times 10 meters or 30 feet below the surface of the ocean, anyway. So even if 50% of the sun is blocked, they could still grow fast.
SPENCER: Got it. Any other kind of promising directions that you haven't mentioned yet?
DAVID: There are a few others. One is relocating crops to the tropics in one of these scenarios. There was actually a lesson here from when I moved to Alaska. I was investigating - is there a climate now that is similar to the climate that would be near the equator in one of these catastrophes like nuclear winter? The closest I found was actually the Aleutian Islands, which are out in the Pacific off of Alaska. It's very cloudy. It's very cool. It actually is so cool in the summer that trees can not even grow, and there's not a lot of rain, but potatoes can grow there.
SPENCER: It just seems like a case study in that kind of condition.
DAVID: Right. And we've identified a few other crops that we think would be low-cost and handle these conditions: sugar beets, wheat, barley, and canola oil. Another potential solution here is greenhouses in the tropics. This would require extruding or making plastic sheets, much more than we are currently producing, but we've estimated that we could scale up that production pretty quickly and cover a significant amount of area with greenhouses, and that would allow us to grow some other crops that we could not grow outdoors - things like corn, rice, and soybeans.
SPENCER: Is the main purpose of the greenhouse to keep the temperature controlled, or what is the reason that allows you to grow anything?
DAVID: Temperature control is the main issue. You can also do more advanced things like adding carbon dioxide to make the plants grow faster, but what we were looking at is very low-tech greenhouses that could be scaled up quickly.
SPENCER: Right, just like kind of clear plastic sheeting that build them out of.
DAVID: That's right. So another option is taking electricity and using it to split water into hydrogen and oxygen, and then microbes can actually eat the hydrogen. Those microbes are high in protein, and there are actually around a dozen companies that are looking to commercialize that technology now for human food. We actually had a few NASA mini grants looking at food sources that could be used in space. That was one of them and I think that's a significant overlap with space foods and foods on Earth.
SPENCER: Going meta for a second. It's interesting how so much of human existence throughout time has been about finding enough food and getting food security. Now we're in such abundance that now we're talking about, how do you create so much food so cheaply that you can feed the entire world, essentially, in a kind of post-apocalyptic nightmare scenario? It's just kind of remarkable how much we've solved that aspect of nature.
DAVID: It certainly is true that most of people's effort historically has been getting food, but that's actually true for many people now in less developed countries that they spend the majority of their income on food.
SPENCER: Oh, interesting, so more than 50% in a number of poor countries still goes to food.
SPENCER: So another question I have for you. If we think about one of these global catastrophes are occurring, it's obviously really, really hard to predict what would actually happen, but I'm just interested in your kind of best guess, based on the thinking you've done. If we really had this massive global food shortage - it was much harder to make food - what do you think would happen to civilization?
DAVID: I'm quite worried as we are now because if people think that the sun is blocked, and most people are going to starve - I did mention we do have some food storage, but it might only be good for three or six months - and the sun could be blocked for five or 10 years, we're in really bad trouble. If countries and people realize that, I'm worried that there's going to be hoarding on an individual level, but also on a national or international level, that it could be in the country's interest to steal food from a neighboring country. And so I'm very worried about what might happen with our current state. However, I think that if we could get the message out, that we can actually feed everyone. If we cooperated, then our chances of cooperating go up significantly.
SPENCER: Yeah, that's a really good point, it feels like what we definitely don't want to happen is a zero-sum dynamic to evolve, because it's bad enough to have zero-sum dynamics when civilizations are thriving, but then the disaster scenario, you don't want just everyone try to grab everything for themselves, and then ultimately, you know, fighting over the scraps.
DAVID: That's right. And I do think that it's not zero-sum in the case of resilient foods - that say, the higher latitude countries might be more industrialized, they could produce some of the methane, single-cell protein, and the sugar, but that doesn't make a complete diet. So we need to have trade with lower latitude countries that might have less industry but can grow some things outdoors and grow some seaweed. Also, the higher latitude countries are the ones that have the seeds that can grow in this future climate in the tropics. So like I said, we're working on a regional model, and we want to work out what the incentives are, but I am somewhat optimistic that it can actually be in the selfish interest of an individual country to trade and increase the size of the pie.
SPENCER: What is needed to make a complete diet? I mean, do you just go through the kind of recommended daily dose of vitamins and minerals, and you kind of add them all up? Or how do you think about that?
DAVID: Yes, and we have a couple of papers looking at that. If you don't have a monetary constraint and you can get a mixture of these different resilient foods, you should be fine. On a low income, that's what we're still working on to make that work out. You also have to worry about getting too much of something. So in the case of seaweed, there's a lot of iodine in it. We might be able to boil it and somehow reduce the iodine, but there's still a lot. So if you look at the maximum limit of iodine, you would say you couldn't actually eat that much seaweed. But interestingly, there was this case of some Peace Corps workers that thought they were taking a regular vitamin supplement, but they were actually taking iodine tablets that were supposed to be purifying their water. So they actually had 50 times as much iodine intake as the recommended upper limit. And yet, they just had minor problems that were reversible. So I think in a catastrophe, we would be able to push the limits a little.
SPENCER: Right, imagine the limits are set pretty conservatively, because we don't want anyone to die. And there's probably natural variation person to person, so they want to kind of have a buffer of safety when they set these hopefully.
SPENCER: So you mentioned cooperation and catastrophe. What are some other things we could do to make the likelihood of such cooperation higher, in addition to these gains of trade and funding, the idea that okay, there actually could be enough in a catastrophe?
DAVID: Well, one question is how to get that message out to a lot of people. Ideally, they know ahead of time. But if this happens soon, most people have not heard about this. So we want to work with social media companies or search engines to say, “Well, how could we get this message out really fast if a catastrophe happened?”
SPENCER: Oh, interesting. Like an emergency system where you know, broadcasts it through Twitter and Facebook or whatever saying, okay, there are enough food supplies for everyone. You don't need to panic, that kind of thing?
SPENCER: Of course, I guess we got to get the food supplies first. How far away are we from actually having enough food is being prepared? I mean, you're writing papers on this, you're doing a lot of great research, but from your perspective, is this something that we're close to here is the thing that we still need a huge amount of work to get to?
DAVID: I think that we've done a lot of the necessary prioritization work, and that we've found what are likely the most promising resilient foods, but we do need to actually demonstrate that it can be done and fast. So that's really the next step.
SPENCER: So like proof of concept, saying, “Okay, we know how to do this, we can build a factory that has this output.”
SPENCER: What's the reaction that you get from most people when you talk about this? I mean, obviously, if you talk to effective altruists, you're gonna get one opinion. But I'm wondering more like, let's say, nine effective altruists, who maybe have not thought a lot about these topics.
DAVID: Well, one reaction is, we don't adequately feed people now. You're crazy to suggest that we could feed everyone in a catastrophe. Certainly, that's true. We do not feed everyone now, but it's not because of not producing enough food. We produce plenty enough food, but it's a matter of whether people can afford it. Then certainly, in certain circumstances, conflict greatly exacerbates the situation. So I think realistically the goal of feeding everyone, no matter what is aspirational. But I am encouraged that it looks like we could, again, if we cooperate, really scale up these foods pretty quickly and make them affordable for nearly everyone.
SPENCER: Do you have trouble getting people to take seriously the idea that it could be one of these catastrophes? Or do people say, “Oh, yeah, that seems like it could happen.”
DAVID: Well outside of effective altruism, it has been hard for people to imagine these big catastrophes, even though we can point to them happening in the past. I'm hopeful that with COVID, people will now realize that catastrophes are perhaps more likely than they thought, and we'll start to take it seriously. But generally, when talking to people outside of effective altruism, we talk about the less sci-fi and the 10% food shortfalls. It certainly is a challenge to think about, well, if you only care about this next election cycle, the chance is relatively low, that one will happen. So you do need to have a longer-term perspective.
SPENCER: That's interesting, though. So you're saying that by focusing on the sort of less severe calamities, people are more willing to engage with it or it seems like it’s impossible to them. I mean, obviously, the more severe ones have a lower probability, but they obviously are also so much worse - is the idea that people are not adjusting intuitively for the fact that there's so much worse that they're still worth worrying about a lot.
DAVID: I think it's mainly the probability issue that if it's higher probability, then they take it more seriously. Interestingly, there was a Lloyd's of London study that said that the reinsurance industry only covers calamities that have at least a half a percent probability of happening each year, or happening every 200 years, roughly. So they can say, an asteroid is just so unlikely that we're not covering it. But if we can show that this multiple bed basket failure or the extreme weather on multiple continents has maybe more like a 1% chance every year, well then, reinsurance really should be covering that. We should be making investments in making our system more resilient to it.
SPENCER: Could reinsurance companies even cover such a disaster? I mean, at some point is so so big that nobody can really insure against it right, like conditioned on it happening, you know, they can't do anything about it at that point.
DAVID: That's possible. But I think that the governments are the insurer of last resort. And I think they do have the budgets to ensure that.
SPENCER: I see. So that seems like it would only work in a mid-level disaster and like a disaster so bad that literally everyone has to live off these food solutions that you mentioned. It sounds like at that point, just all of society is different. And a lot of the basic structure we have now is broken down. Is that right?
DAVID: Potentially. Though I did argue in the book Feeding Everyone No matter What that if we could scale up these resilient foods, then perhaps most people would be doing their same jobs as before. Of course, food is going to cost a lot more, we're going to be re-employing people who are currently doing outdoor agriculture and the high latitudes to something else, we're going to be having them build factories instead of building more roads, but I think there is hope that we could have a largely functioning civilization.
SPENCER: Just with very bad restaurants.
SPENCER: Okay, should we switch over to talking about the second type of calamity related to kind of electricity and industry?
DAVID: Sure. So here, the catastrophes include an extreme solar storm that can induce currents in electrical grids and destroy transformers. There could also be a detonation of a nuclear weapon at a high altitude, causing an electromagnetic pulse or EMP. And that would destroy not only transformers but also lots of things that are plugged in like electronics. And there's a concern of this happening once over one country or a region, but it could potentially be multiple ones around the world, especially if there were a nuclear conflict. Then there could also be an extreme cyber attack, perhaps enabled by narrow artificial intelligence. Cyber attacks have already disrupted the grid locally, but this could be a much larger scale. And finally, there could be an extreme pandemic, where people are so afraid to show up to work in critical industries that we just can't maintain our electrical and industrial systems.
SPENCER: How do you think about the differences between these kinds of scenarios? And food-related ones? Like what are the main distinctions you draw?
DAVID: Well, if there is an abrupt loss of electricity in industry, then there are even more urgent issues than food. One person in ALLFED, who used to be in the UK Royal Air Force said, “Well, you'll die in three minutes without air, three hours without shelter, three days without water, and three weeks without food.” So, fortunately, we would still have air. But we do need to continue to heat buildings to have adequate shelter. And we need to get water very quickly. So we're looking at how people could transport water without industrial systems, cars, etc, at least for the first few weeks until we figure out something else. But then eventually, the food system is going to become really important because we have that limited food storage. And the reason we can produce so much food now is because we have industrial fertilizers, pesticides, tractors, improve crop varieties, etc.
SPENCER: So when we imagine these kinds of scenarios that affect electricity, how long term should we expect them to be? Because you know what some people might think. Well, it's just a matter of getting them back online, you know, maybe a few weeks. My senses, though: things are not that easy.
DAVID: Well, people have pointed out that to replace one of these large high voltage transformers, it can take six months or even a year. And so people are worried about long-term outages. But what I'm particularly concerned about is if it were over a large region, or even globally. Well, how do you even make the replacement parts? And there's quite a bit of work on, can we prevent these catastrophes, which is great, but what we found very neglected is well, what if the catastrophe happened? How would we actually meet human needs until we get it back online? But to try to answer your question - how long might it stay offline - there is concern that if we lost industrial civilization, that we might not even be able to restart. If it takes a long time, and people forget things and now, we've used up most of the easily accessible fossil fuels, and we kind of have to start over, there's no guarantee that we could recover from it.
SPENCER: I wonder if some people think it's implausible, because we still have books of information and all these kinds of things. Maybe you could paint a picture there a little more detail of what would happen that would prevent us from really booting these backups. Now imagine people can see how it might take a while, but like a decade, we're not back to having industry and the electrical grid and all that.
DAVID: Well, I'm quite optimistic, especially if we can meet basic human needs. And we can keep cooperation, that we will be able to have surplus labor, such that we can fix the damage and restart these things. But I'm very concerned that if we haven't planned ahead of time - how we're going to meet basic human needs - and then we're basically in an overshoot scenario. We have more people than we can feed with existing systems, that again, you can get into these zero or even negative-sum games, where there could be massive amounts of conflict and spiraling downward.
SPENCER: In the relatively minor disasters I've been in, you know, like hurricanes hitting, it is kind of remarkable how quickly people switch into this different mode. And I remember going to a store and people were screaming at each other to buy batteries and flashlights and hoarding stuff. And these were like, really minor credibly minor by the center you're talking about. It seems to me that there are a lot of people that are going to act basically kindly to other humans. But in a scenario where you feel like your survival is on the line, a lot of that goes out the window very quickly.
DAVID: There's definitely debate about this, there have been studies indicating in these more localized disasters that sometimes people are surprisingly cooperative. But I do think that it can be different, when you know that eventually, help is going to come versus a really terrible scenario where no help is going to come.
SPENCER: Right. And I've seen that too. Like I've seen people directing traffic in the street when all the traffic lights went out in New York City. And that was really cool. People are just volunteering to do it. I've seen people bringing water to the people direct control ministries just to contribute, you know, that's really nice. I think a lot of it has to come down to safety. Like, if you feel like you, yourself, and your family are literally threatened. I think that's when people get into this other mode where it becomes very, okay, we're gonna grab anything we can to stay alive. Whereas if someone feels like they themselves are not threatened, then they can go out and help the community and kind of look to other people's needs.
DAVID: I think that's right. And on a country level, people have observed that if a country feels sufficient in a particular food, then they will be willing to export it. But if they feel they don't have enough for themselves, then they're very hesitant to export.
SPENCER: Yeah, that makes sense. So what are some strategies we should be thinking about in terms of helping make sure that we can deal with these huge electricity disasters?
DAVID: Well, one big one is having some way to have a communication system. And fortunately, there's this technology called shortwave or sometimes called ham radios, where $20,000 system can actually communicate across an ocean. And so we're interested in having a global backup radio communication system that would have its own power supply, you know, maybe solar panels, and would also make sure to be protected from any electromagnetic pulses. So typically, it would be unplugged. Normally, in terms of food, we completed an effective thesis, looking at what is the agricultural yield if we lost those industrial inputs. And in order to make this work, we would have to scale up hand and animal farming tools very quickly, which we need to work out. We also need to work out the transportation of the food and some things we're looking at is whether we could basically go back to animal power for plowing fields and transporting goods.
SPENCER: But then in terms of things like having backup transformers, in case they blow out, I guess these transformers are like giant ones that power plants or something like that.
DAVID: That's right. And it certainly would help and they have a certain amount of spares right now, but to do it globally, would be more expensive.
SPENCER: What are some other potentially promising interventions around this?
DAVID: Well, I mentioned how we would lose artificial fertilizers. And so one intervention is to plant more legumes, which are peas, beans, peanuts, that can fix nitrogen from the atmosphere into a fertilizer. But then there are other fertilizers we need such as phosphorus and potassium. So we looked into whether we could take wood out of landfills and burn it to create ash, and use that as a fertilizer. There was actually a time when the US was burning trees and shipping the ash to Europe as fertilizer. In terms of pesticides, there are ways of controlling pests that don't require industrial pesticides. These are already used in organic agriculture and these are things like using one animal that will eat a pest or other techniques.
SPENCER: Okay, so stepping back for a moment, you've talked about a bunch of different potential interventions both on helping protect food supply or helping feed everyone in catastrophes, and then other ones around the electrical grid or dealing with getting the industry back online. If you had to point money at just a couple of these right now, what would you see is really the highest leverage or most promising thing given what you know?
DAVID: Well, it's something we think a lot about prioritizing within ALLFED. And we've written some papers on looking at the cost-effectiveness of these interventions, some from the perspective of the present generation, how many people might die in these catastrophes, how many could be saved if we actually got prepared, but then also from the long-term perspective, if we could prevent a collapse of civilization that we might not recover from. We've identified the broad classes of interventions, of researching and prioritizing, of demonstrating the technologies and of planning and really, we think all of them are important. I would say we're generally finding that these catastrophes that disrupt electricity are maybe a little less bad or a little less likely. So we haven't put as much effort into them. But then, fortunately, they actually don't require quite as much money to get prepared for because we don't need to build factories, we wouldn't be able to build factories in those cases.
SPENCER: That's very convenient. Because otherwise, you'd have to rename your organ to all electricity.
SPENCER: When you think about this cost-effectiveness analysis, are you starting at the level of different catastrophes and trying to estimate the probability of each such catastrophe? And then for each catastrophe, you're saying, well, if that occurred, here's what we would need? Or here's what we'll use? Or do you take a different approach?
DAVID: That's right, though, we generally lump multiple catastrophes together. And then we say, well, how bad would it be with our current state of affairs? And then how much money do we need to get reasonable preparedness? And then how much better would the world be if we do have that preparedness?
SPENCER: What's your final metric in terms of cost-effectiveness? Is it sort of like life saved and expected value or something like this? Or do you use other metrics?
DAVID: One of the metrics is the cost per life saved. Another one, you can think about it in terms of keeping the system running economically. So you can actually think of a benefit-to-cost ratio. From the long-term perspective, though, it's harder to put it in economic terms because the future value could be potentially astronomical. So generally, what we've done is compared to another intervention, and what we've focused on is artificial general intelligence safety work.
SPENCER: So when you're thinking about that, are you thinking about basically preventing civilization collapse? Or what is the way that you connect your work to the long-term future?
DAVID: Well, I think that there may be some possibility that the catastrophes we work on could cause extinction, I don't think it's very likely, but an observation has been made that we have some food storage in industrial countries. So probably, some people would survive on that food storage. But then if we have a collapse of civilization, and we have to go back to hunting and gathering, then maybe we can't figure out how to do that. There's actually this book called The Secret of Our Success, which looks at examples where people have had massive troubles going back to hunter gathering.Then you could say, well, but we have a lot of hunter-gatherers still today, or at least some, but they don't typically have a lot of food storage. So if the sun is blocked for five or 10 years, they would be in trouble. And then people say, well, but they could still fish. But if you have more people than fish, they could actually cause the fish to go extinct. That's a possible mechanism, but I don't think direct extinction is very likely. So we've talked about the potential collapse of civilization and not recovering. But I think there are other routes that these catastrophes could have an impact on the long-term future. One of them is that the trauma of a catastrophe could make worse outcomes more likely, such as global totalitarianism. Another route is that maybe the worst value is because of the trauma of the catastrophe end up in any artificial intelligence that we might create.
SPENCER: So let's unpack those because they're a little, I think, harder to understand. So the first one is the idea - something like if there was a huge global catastrophe leading to kind of widespread famine, it could make people more willing to kind of put up with a long term dictator to gain stability or something like this?
DAVID: That's right.
SPENCER: Got it. And the second one is like a large scale catastrophe could create sort of a more zero-sum or negative-sum world where people are fighting for survival, and therefore, they can kind of shift human values so that if one day humans built artificial intelligence that's super-intelligent that it might encapsulate those more like, cutthroat values.
DAVID: That's right.
SPENCER: Okay, got it. So then, how do you think about comparing this to a cause area, like trying to prevent negative artificial intelligence scenarios?
DAVID: Well, fortunately, some other people have done some modeling on this, and looking at what's the chance that this might go poorly. And then how much effort might it take to increase the chance that things go well, and so they were already looking in terms of, if we spend a million dollars, how much better percent do we have a good outcome? And so we basically use that framework, but then applied it to the catastrophes that ALLFED work on.
SPENCER: Got it, it seems to me like that kind of analysis is going to just have so many assumptions in it that you tried to work out the comments intervals. They are just so massive that it may not be that informative. I'm curious, do you disagree with that?
DAVID: There is a lot of uncertainty. In the case of the long-term perspective, we did have orders of magnitude of uncertainty. However, we could at least say that the cost-effectiveness of these interventions overlapped with artificial general intelligence safety. If you're more optimistic, or more optimistic about the ALLFED interventions, you might say, with some confidence that they're actually more cost-effective, at least at the margin. What we found was that, overall, the importance of artificial intelligence was greater, and we think it's best to spend more money on that, but because these resilient foods are so neglected, we found that they're quite cost-effective at the margin now. And we think it does make sense to like I say, spend a few 100 million dollars, even though for artificial intelligence, we'll be spending billions of dollars. From the perspective of the present generation where we're looking at cost per life saved, we compared to global poverty interventions, such as mosquito bed nets for preventing malaria. And what we found was - despite very large uncertainty in the cost per life saved for these resilient foods, you can say with some confidence, you believe our assumptions, which were peer-reviewed - that it would actually be cheaper than these mosquito bed net interventions.
SPENCER: Got it. So you think that it's least plausible that your interventions are on par with classic effective altruism cause areas, and $1 per benefit?
DAVID: Yeah, so as for the cause, area, it is still very small. Fortunately, as I mentioned, there is a fair amount of private industry work that is working on resilient foods, but not for the sake of catastrophes, but instead for sustainability issues, and also increasing animal welfare, which are great causes, as well. But what we're hoping to do is work with them and kind of nudge them in the direction of also providing value and catastrophes.
SPENCER: So just sort of lucky coincidence that the work they're doing also can be ported over with some effort.
DAVID: Right. And some things are conventional already - seaweed and greenhouses, etc.
SPENCER: Why isn't there a lot more money going to this, given that what you're working on has the potential to benefit all of humanity, given that there are lots of governments that are kind of at least theoretically looking out for the benefit of their people, and this is a potential danger that everyone the world faces?
DAVID: We've been trying to figure that out ourselves. We do think that it should appeal, not just to the longtermist, but also the shorttermist.
SPENCER: It could just be a kind of probability neglect thing, right? Like, well, we have so many other things to worry about that are more direct, and here and now this small probability of a bad outcome, you know, that's not something that's gonna happen in my election cycle. And it's probably not gonna happen anyway. Right? I don't have to worry about it.
DAVID: Yeah, one example is talking to organizations such as the World Food Program, and they say, “Well, we have enough issues to work on now.” And so they're kind of overwhelmed with the current situation. And so we need to find the right stakeholders that are looking longer term.
SPENCER: Got it. And where do you find those? I mean, has most of your uptake been in the effective altruism community? Or do you find others interested in that?
DAVID: Most has been in effective altruism. Yes.
SPENCER: What about governments? Like have you had, I don't know to what extent you've pursued this, but have you made any inroads in getting governments interested in backing or anything like this?
DAVID: The UK government has been interested. We actually have more people in ALLFED in the UK than in the US even though I'm in the US. So I mentioned the UK governments study on the 10% food production shortfall, but they are also concerned about large-scale electricity outages. And most of the efforts are on for prevention. But we hope that we can convince them to think about a backup plan.
SPENCER: I don't know if any governments have this, but it seems like government should have an agency that thinks about the long-term future 20 years out 30 to 50 years out, and tries to avoid things like pandemics and things like massive electrical outages and so on. Because it just seems like the structure of society is such that there's so much emphasis on the short term, but there's almost nobody that's really thinking out that kind of time length.
DAVID: I agree, and there have been a few efforts around the world to actually have an agency or something like that, that thinks long term. And one of those efforts is in the UK.
SPENCER: Is that a government agency?
DAVID: Well, there's the All-Party Parliamentary Group for Future Generations. And yeah, this is something where you, maybe you could interview someone who does this.
SPENCER: Okay. So stepping back and thinking about your experience with this, I can just imagine if I was working on this, that it would be hard to be positive, that I would be constantly thinking about all these catastrophes that could befall society and how we're not prepared for them. Does it take a psychological toll thinking about this?
DAVID: It certainly can for some people working on these catastrophes, I personally focus on the positive interventions that we can do, and don't focus too much on the bad catastrophe. If we don't handle it well.
SPENCER: Got it. So it doesn't personally have too much negative impact on your life working on this topic.
SPENCER: How did you get into this originally?
DAVID: So even since high school, I had been thinking a lot about how I can improve the world maximally, you know, whether it was through donating or through my career directly. And I started focusing on global poverty, and actually did research on solar, water pasteurization, so that's using the sun to heat up water to kill the terms for people that don't have water treatment. Actually overlapped with Peter Singer when I was at Princeton and did some work on animal welfare, but then I became convinced of how many future generations we could have, even if we don't colonize the galaxy, or scan our brains into computers, just living on Earth for a billion years, we would have something like 10 million future generations. So there could potentially just be so many more people alive than are alive today. And so, there are many potential ways to influence our eventual outcome, but I think a clear one is these global catastrophes that could really push us off course.
SPENCER: So did you come up with this idea that oh, hey, I wonder if anyone's working on this topic of what do we do in a giant catastrophe? Or how did you stumble on that?
DAVID: Well, I created a long list of catastrophes and I looked at how bad they might be, and then how much work has been done on them so far, as a way of prioritizing. And then I just looked at, well, what can we do about it. I'm an engineer, so I think in terms of engineering solutions, and I just found that still, there are fewer papers on existential and global catastrophic risks than there are on dung beetles. But still, even within that field, the majority has been on the preventing of the catastrophes. And of course, that's the best outcome. But I think we need to be realistic that we can't prevent everything. And so we need to have some backup plan for what we would do if one of the catastrophes happened. So that seemed to be the neglected area, and I had some comparative advantage in it.
SPENCER: That's really interesting, because, you know, maybe the story gets kind of cleaned up when you tell it as you know, our pals tend to be nonlinear, but it is interesting to me because you sat down, you analyze these different things that could happen. You noticed the gap that wasn't filled in and you stepped up to try to fill it. I think it's a really good example of the value of having lots of people trying different things and lots have people looking for holes that aren't plugged where a new organization might be able to help improve the world? So I think there's just a really neat example. How long ago was it?
DAVID: This was 2011.
SPENCER: Okay, got it. So you've been at this for a while now.
DAVID: Yeah, it took me a few year. I had a regular full-time job, but to write the book Feeding Everyone No Matter What.
SPENCER: And then when do you found ALLFED?
DAVID: That was actually 2017.
SPENCER: Okay, got it. Cool. And you're based in Alaska? Is that correct?
DDAVID: Yes, I'm in Fairbanks.
SPENCER: So not a lot of people have had experience out there. I'm curious to hear what's been like for you there?
DAVID: Well, it's certainly an adventure often gets to negative 40 degrees Celsius, or negative 40 degrees Fahrenheit. And the outdoor recess for my kids actually goes down to negative 19 degrees Fahrenheit or negative 28 degrees Celsius. And they huddled together like penguins.
SPENCER: The perfect place to consider the end of the world.
DAVID: That's true. There are definitely analogies here, the crops that can grow in Alaska might be able to grow and nuclear winter. And we've also had a pretty atypical case with the pandemic that it was quite delayed, but yet we locked down about the same time so basically had it suppressed in the summer of 2020. And then we also had one of the fastest vaccine rollouts, fastest in the country, which was surprising to me, because 75% of the communities in Alaska actually have no road access. And even the capital, Juneau, Alaska has no road access. So they say there are only three ways to get to Juneau. It's either boat, plane or birth canal.
SPENCER: Wow, that's like blew my mind. You cannot take a road. Did you know? That's correct. Oh, my gosh, what a different world.
DAVID: Yeah. And so despite all these remote communities, they were able to roll out the vaccines really quickly. And there was actually a case where they had a bunch of vaccines, I think it was ten. And they had this tiny, remote community, and they only needed something like five vaccines, and they didn't want to waste them. So instead of flying the vaccines to the community, they flew the people that wanted to be vaccinated to more hub towns so that they could use all the vaccines.
SPENCER: Oh, wow, that's uncharacteristically reasonable for the government coordinating an effort like that.
DAVID: It's certainly an extreme climate; we're not quite at the Arctic Circle, which means on the winter solstice, the sun goes one degree above the horizon for about four hours.
SPENCER: Oh, wow. What is the experience of that like?
DAVID: Well, it's perpetual sunrise or sunset. Fortunately, the sun is just a little below the horizon for quite a while. So we have something like six or seven hours of sunlight.
SPENCER: Wow, that's pretty cool. I think there's like a Douglas Adams book or something where planet is always sunrise. As we wrap up, I just wanted to see if there are any other final things you want to say?
DAVID: Well, I think one thing I'm proud about at ALLFED is that we have a vigorous volunteer program. And so we have people are interested in volunteering, whether it's coming in through the effective thesis process or other ways, we first do a taster task, because we're a food organization, where people will do a task for a week or two, and see if it's a good fit on both sides. And then we have orientation. And we've had very significant contributions from our volunteers, including on the research side. And we make sure to have team meetings on specific-related areas, such that multiple people know what other people are doing. Because of course, sometimes it just turns out that volunteer has to stop volunteering, and then other people are able to pick it up. But yeah, we've had several volunteers be co-authors on papers.
SPENCER: That's great. I feel like it's often difficult for orgs to find stuff that volunteers can do that's actually really beneficial, but also a really great learning experience. So it's like, you know, a really positive-sum for everyone.
DAVID: And several of our volunteers have transitioned to full-time paid work.
SPENCER: I assume the taster task is you send them to Alaska and see if they can grow mushrooms.
DAVID: That's a good idea. I haven't tried that yet.
SPENCER: So David, for those that have been inspired by what we talked about today, and want to support your mission, what's the best way for them to do that? And can you just give your one-minute pitch of like, why do you believe in what you're doing? Why do you think it's impactful?
DAVID: Well, I think that ALLFED is working on an important problem. It's very large scale in terms of saving lives in the present generation and also increasing the chance that we have a good outcome in the long term. I think it's also very neglected. And I think it's tractable in that there are concrete things we can do now to understand these resilient foods and other interventions better and also prove them out actually build the factories and such that we can have a good response in these catastrophes, and if you are interested in supporting us, we are participating in the Giving Tuesday Facebook match. And you can also go on to our website ALLFED dot info.
SPENCER: Cool. David, thank you so much for coming on. This was great. Thanks for having me.
Click here to return to the list of all episodes.
Sign up to receive one helpful idea and one brand-new podcast episode each week!
Subscribe via RSS or through one of the major podcast platforms:
Host / Director