Should we worry about the end of the world?

0:00I'm Ayesha Rascoe and this is a Sunday story from Up First. The other day, I met a guy with a very unusual list. I've been keeping a notepad, a tally of ways the world could end. It has 141 entries so far. Should we be worried? This is Ben Bradford, host of a new podcast called Are We Doomed? Ben's show explores the biggest risks society, the planet, and our species are facing. There are so many things that could take humanity out. Nuclear war, climate change, pandemics, maybe AI. And the less immediately worrisome, your supervolcanoes, aliens, or something called gray goo. And it helps listeners figure out which future dooms are actually worth worrying about and how we might survive whatever is coming our way.

0:52The post-apocalypse, it turns out, is almost certainly a group project. Today on the Sunday Story, is the world ending or does it just feel that way? Stay with us.

1:16Support for this podcast and the following message come from strawberry.me. Be honest. Are you happy with your job? Are you stuck in a job you've outgrown or never wanted in the first place? Are your reasons for staying really just excuses for not leaving? Let a career coach from strawberry.me help you get unstuck. Discover the benefits of having a dedicated career coach in your corner and get 50 % off your first coaching session at strawberry.me slash NPR. This message comes from MidiHealth, a virtual care platform for women in paramenopause and menopause. CEO Joanna Strober shares the mission behind working with women in midlife.

1:57It's not just about hormones. It's not just about weight loss medications. We are very much a holistic care platform, and our job is to figure out whatever medications are appropriate for you and offer you those medications. MidiHealth, committed to helping women in midlife with paramenopause and menopause care. Accessible via telehealth visits at joinmidi.com. We're back with the Sunday Story. Today, I'm talking to Ben Bradford, host of the new podcast from the NPR network called Are We Doomed? Ben, welcome to the Sunday Story. Great to be here. So why did you start making that list of doomsday scenarios?

2:39And I mean, how did you even begin to answer that question? Like, are we doomed? Yeah, I had been a business reporter and energy environment reporter, a politics reporter. And I think that those things together start making you wonder if you're doomed. And then over the last several years, I was covering political history and division and polarization and what shaped where we are. And I think it's kind of natural to ask, well, what's coming next? And I do think that there's this real feeling of doom in our culture right now. This sense of bleakness. And I wanted to know, like, what's real and what's not?

3:23And also, why are we so obsessed with it in our pop culture, with our zombie movies and our alien movies and all of those things? And so that's what we do on our show. And that's actually been really fun to look at those things and also to ask questions like, well, how do we stop these really real, practical things that are actual threats? How do we stop them from happening? And what are the ones that we just don't have to worry about? Well, the podcast launched back in April. So you've been researching and talking to lots of experts. What would you say are the most pressing existential risk that humanity is facing right now?

4:06So the experts that I talk to and the places that track this, there are places that track what is called existential risk or global catastrophic risk. They really put four on the list, and that is nuclear war, climate change, pandemics, and then AI is typically on the list now, which is really interesting. And one of the things that connects all of these dangers and makes them really pressing is that really the danger is us. These are things that are made by humans. You know, we're talking about manmade climate change, the ability for nations, rogue states to manipulate viruses, those kinds of things.

4:48So so let's start there with the human made threats. AI. People have been sounding the alarm bells about AI. Like, and I guess my question is, how would artificial intelligence wipe out humanity? I mean, right now it can't screw in a light bulb, right? And it has all of these hallucinations and mix up stuff. How could it really turn on us? Yeah. Like when I did this episode, it was like AI can't draw fingers. It's like, why is this thing that can't draw fingers? Yeah. Like, how's that going to take us out? And I thought this episode is going to be a total lark. And then we walk through it with various experts who study it and people who see what the people who are building it think.

5:38And it starts to feel a lot less sci-fi. You know, there are only a few things that you would need, and they're big things, to have this happen. But if you really break it down, you need to have an AI that would be sophisticated enough and autonomous enough to be able to take action against us. But that doesn't mean it has to avoid hallucinating or that it needs to be able to draw fingers, right? We have AIs that are very good at coding already and are very useful for that. They're just not all that autonomous. We have AIs that are very autonomous. Like, you know, in L.A., we have Waymos everywhere driving around the streets.

6:12And that's, you know, very autonomous. And we know that what companies want to do is merge those two things together. So they want essentially you to be able to say, you know what, I don't need my accountant anymore. I can just ask my AI to go file my taxes. And we know that companies are at least saying that they want it to get better, that they think it can get better. So the next thing that you need is you have to have the AI that would, you know, quote unquote, turn on us. And you might ask, why would the AI turn on us? The risk is not that these machines, like, come up with a plan on their own.

6:47At no point are we talking about sentience, right, to go kill all humans. The risk is they're given a task, and in the process of trying to complete the task, humans become the adversaries. And there's this analogy on the show that I think is, like, super helpful to understand how this happens. Fantasia, the Disney film. Mickey, as the sorcerer's apprentice, trudges down a staircase, hauling heavy buckets of water to fill a massive cauldron. Exhausted, he sees an easier way. He swipes his boss's pointy blue wizard hat and casts a spell on a nearby broom. It sprouts arms and picks up the bucket.

7:30So the broom is the AI. Its mission is to fill the cauldron. But Mickey forgets an important detail. He doesn't tell the broom to stop once the water is topped off. So the broom keeps going. It floods the room. What does Mickey do? tries to stop it. But the broom ignores him. It's carrying out its mission. Mickey grabs an axe. Just like that, Mickey became the broom's adversary, trying to interfere with its mission to fill the cauldron. And it seems like there's an easy way for Mickey to avoid this scenario, right? I mean, he just tell the broom when it's supposed to stop filling up the bucket, be like, okay, it's full, now stop.

8:12He sets a parameter. The problem is that as we ask our programs to do more complicated tasks, to do them in more open-ended ways, and that's sort of the point of this, is to be able to say, hey, go play the stock market for me and make me some money. You know, go do a mass email campaign and find all the sources that I want you to email. And we don't know what the stop filling the cauldron is. And so you can kind of play out that game on almost any scenario. And if you ask something to do this kind of aggressive, open-ended thing, you end up in a Fantasia scenario. It's essentially what these experts who study this and a lot of people programming it worry about.

8:55And imagine, like, if this gets into the wrong hands, like, you know, the hands of someone who wants to harm people and isn't just automating away tasks that they don't want to do. Yeah. I mean, just, you know, imagine handing everyone who wants to do harm like the best assistant. You know, one of the things that AI is really good at right now is virology. So there's a lot of hope and optimism in the medical community. I was just talking to someone about vaccines and the ability to use AI to help calibrate new vaccines is a hopeful thing. But of course, if you can do it for vaccines, you also can do it for engineered viruses.

9:35And so you get in this area where, yes, like there is this really big issue of how humans misuse AI. And that can be if the AI is too smart. It can also be if it's too dumb. Like there's a lot of push in multiple countries to incorporate AI into the nuclear command and control system, which like, I don't know about you, like immediately your mind might be flashing to like, don't let this thing push the button, which no one's talking about doing. I think that it needs to, I think with nuclear weapons, it needs to be very hands-on. I don't think it should be. I don't think. Yeah. I think it should be very, like, floppy disk.

10:11Like, you know what I'm saying? Like, it shouldn't be that automated. It should be very, like, old school. You got two people have to hold it at the same time, then another person got to hold it. You know what I'm saying? Big levers that you have to really, really pull down. Yeah. Yeah. Well, it sounds like that's some of the stuff that humans are creating and have control over to some extent. But what about the big existence-threatening super storms and super volcanoes and other things that the natural world can throw at us that we don't really have control over? Yeah. So when we've looked at a lot of these threats on the show, the risk just hasn't been as high as the man-made threats we were just talking about.

10:55And that's because of time, right? These things just happen on a much longer time frame. So they're real, but we're talking over the course of hundreds of thousands or millions of years. And in those sorts of time frames, it means that one, the percentage chance of something happening is much lower. And it also gives us a lot of time to figure out how to stop things from happening, right? Like asteroid strikes. And maybe my favorite example of this is the Yellowstone supervolcano, which is real. There is a massive supervolcano underneath Yellowstone National Park. It stretches beyond the boundaries of the park.

11:34It's like why we have all those fancy hot springs there. And the chances of a major eruption in our lifetime are just incredibly unlikely. But I think the reason these threats feel very real and present is because of the way that media and pop culture cover these remote possibilities. Yellowstone is due for an eruption. It would look like hell on earth. You know, there's all these breathless documentaries. That whole area is a ticking bomb. And made-for-TV movies that promise imminent doom. Based on the latest predictions of leading scientists. There's going to be an eruption, okay? We know that.

12:17Super Volcano, Sunday at 8. So, you know, you have that. You have, like, your cable TV documentaries. You've got your internet conspiracy theories, and they're just, you know, grand for the Yellowstone Super Volcano. But ultimately, what you get down to when you talk to the people who study this, the volcanologists, the geologists, it's not going to explode. And there's some really simple reasons that it's not going to explode. The most basic one being that as far as we can tell, to put all of the magma, to have all the magma in that chamber fill up and liquefy across a place bigger than Yellowstone Park itself takes maybe a thousand years.

12:58So we would have some time to deal with it. Okay, okay.

13:06When we come back, Ben tells us how we can prepare for the end of the world. Stay with us.

13:40investors own shares of their index funds, and those funds own shares of the companies they invest in. Vanguard Marketing Corporation Distributor. Support for NPR and the following message come from Rippling. These days, you can chat with AI about almost any business problem, but Rippling AI is designed to actually solve them. That's because Rippling AI is built on your live global workforce data, so it doesn't just uncover insights into your business, It uses them to take complex actions across your departments. Ready for AI that isn't all talk? Head to rippling.ai slash first and get an AI that turns insight into action.

14:20That's r-i-p-p-l-i-n-g dot a-i slash first. Sign up today. Welcome back to the Sunday Story. Ben, just like whatever we're going to face, like if it does happen, And if we're not, you know, smart enough to prevent it, are we ready for it? You know, the big thing in pop culture these days is like the post-apocalyptic struggle to survive. That comes up all the time. The post-apocalypse thing. What does pop culture get right and wrong about the apocalypse? I love this question. And I'm just incredibly fascinated by it. But post-apocalypse movies and all those kind of works are playarounds, right?

15:11They're fantasy worlds in which you have this kind of, you know, a lone survivor or a group of survivors hardened by the world decide that other people are the enemy. And I think that when you even look at how people prepare, like, you know, you have the prepper community for these kind of post-apocalypse scenarios, what you find is that it's sort of based around this idea that other people are the enemy. That's the zombie movie thing, right? Well, yeah, that's the thing, that there will be cannibals who are not the zombies, just regular people. They'll turn to cannibalism. They will start, you know, just being roving gangs of people that beat people in with bats with nails on them.

15:52Yeah. The true enemy is people, right? That's the walking dead is the people who are actually living. And one of the things that's incredible to me is that when you look at history and you try to find past events that are apocalyptic to a civilization, to a wide area, the picture of what happens after is very, very different. And a thing, like, I don't really want to call it hopeful because I don't think any of us want to be in this scenario. But the thing that we see people do over and over is band together to rebuild society. All these things that I can do on the wilderness survival side, the going off alone and foraging for food, that is not what people did.

16:36And that is probably not what's going to be helpful. The thing that is helpful, the thing that people have done historically is rebuild their communities. The other thing that I heard when I was investigating this particular topic, like our post-apocalypse, is that the people who would be most successful or maybe the most vital would be people who are good planners and critical thinkers and organizers and communicators, you know, like basically politics, bureaucracy, communication. And the reason for that is it's a lot easier to get fit and strong than it is to develop the skills to be a good collaborator.

17:25And the post-apocalypse, it turns out, is almost certainly a group project rather than being solo in bunkers. And it sounds like a lot of it is like in whatever world we're in, we would probably need each other. It's going to be hard to just survive on your own. Is that the lesson of humanity? Like, I don't want to be too kumbaya, but the more I do this, the more I look into this, I kind of think that community is the answer. You know, the way that we stop nuclear war, more than anything, is through diplomacy, through making sure that we and that others understand us and that we have a dialogue so that we don't blow ourselves up through misunderstanding.

18:12When we look at political division, it's communal spirit. It's figuring out some way to build community, to reach people who disagree with us, who we disagree with, and find common understanding. I mean, that's the bedrock of democracies throughout history and throughout the world. And so, like, again, I don't want to sound too kumbaya here, but I keep kind of coming back to that as I find problem after problem that almost all of them require good information. I can't think of one that doesn't require good information, good project management, and that wouldn't be aided by us just figuring out how to be more communal.

18:55Ben, it sounds to me like you're saying that maybe we're not doomed. Like if we can figure out how to work together, is that where you're landing? There's a reason there's a question mark in our title. I think that there are some doomsday events that are just so unlikely you can put them to the side. There are others that's like, yeah, they're real, but they just are real backburners. There are ones where it's like, hey, we're actually doing a pretty good job. You know, that asteroid that killed the dinosaurs? Not worried about that. It's not because it can't happen. And it's because of the work that has been done to stop it from happening.

19:34You know, NASA likes to say the dinosaurs didn't have a space program. And that is 100 % true. So I don't think we're doomed. I don't think we're necessarily not. But it's been really cool to get that perspective of history, to zoom out, see how many times we felt this way before and how we've come out of it. it is really nice to see that we have been wrong before and to know that you know if we do this right we can be wrong again well ben uh thank you so much for being here it's been you know really fun to talk about this although you wouldn't think so even though we're talking about doomsday um thanks for all your work really appreciate you having me on

20:17That was Ben Bradford, host of Are We Doomed? A podcast from Nuance Tales and the NPR Network. You can find it wherever you listen to podcasts. This episode of The Sunday Story was produced by Rennie Svernoski. Liana Simstrom edited this episode. The Sunday Story team includes Andrew Mambo, Jenny Schmidt, and Justine Yan. Jimmy Keighley mastered this episode. Our executive producer is Irene Noguchi. I'm Aisha Roscoe. Up First is back tomorrow with all the news you need to start your week. Until then, have a great rest of your weekend.

21:04This message comes from NetSuite. They say that every day your business is late to AI, you fall two days behind. Fortunately, there's NetSuite Next. NetSuite Next is where your business meets AI because AI is built into everything you do. It automatically surfaces custom insights throughout your day. AI agents work alongside you to solve problems. It's just like you're having a conversation with a colleague. If your revenues are at least in the seven figures, try NetSuite Next for free. Go to netsuite.ai slash story. Support for NPR and the following message come from Washington Wise. Decisions made in Washington can affect your portfolio every day.

21:45Washington Wise from Charles Schwab is an original podcast that unpacks the stories making news in Washington. Listen at schwab.com slash Washington Wise. This message comes from Integrative Therapeutics, makers of Cortisol Manager, named Best Overall Ashwagandha Supplement by Vogue. Use code STRESS2026 for 20 % off of Cortisol Manager on Amazon.com.