How often do you think about all the ways the world could end?

As the host of The Gray Area, I find myself engaged in this macabre exercise more than most. We’ve done episodes on runaway AI and climate change and extinction panics. One of the few topics we haven’t covered, however, is nuclear war. Which is surprising because this scenario is near the top of basically every list of existential threats — and now feels newly salient with recent news involving North Korea, Iran, and China.

Annie Jacobsen is a reporter and the author of a new book called Nuclear War: A Scenario. I read a lot of books for the show and this one stuck with me longer than any I can recall. It’s a book that clearly wants to startle the reader, and it succeeds.

Jacobsen walks you through all the ways a nuclear catastrophe might unfold, and she gives a play-by-play breakdown of the terrifying choreography that would ensue in the minutes immediately after a nuclear missile launch.

So I invited Jacobsen on The Gray Area to talk about what a nuclear exchange would really look like and how perilously close we are to that reality. As always, there’s much more in the full podcast, so listen and follow The Gray Area on Apple Podcasts, Spotify, Pandora, or wherever you find podcasts. New episodes drop every Monday.

This conversation has been edited for length and clarity.

I suspect the image most of us still have of nuclear bombs is the image of Hiroshima and Nagasaki, but that was a long time ago. How much more powerful are the thermonuclear weapons we’re talking about today?

To give you an idea of a thermonuclear weapon, I went to one of the ultimate sources, a 93-year-old nuclear weapons engineer named Richard Garwin, probably the most famous nuclear weapons engineer, physicist, presidential adviser, still alive. Garwin drew the plans for the very first thermonuclear weapon. Its code name was Ivy Mike; it’s on the cover of my book. It was 10.4 megatons. 

So consider that the Hiroshima bomb that you referenced was 15 kilotons and then think about 10.4 megatons. It’s about 1,000 Hiroshima-sized bombs detonating at the same time from the same center point. Garwin explained it to me in the simplest of terms when he asked me to visualize this fact: A thermonuclear weapon uses an atomic bomb as its fuse inside of the weapon. That’s how powerful it is.

Paint the picture for me, as you do in the opening pages of the book, where you imagine a nuke is dropped on Washington, DC. What happens next?

So with a 1-megaton bomb on Washington, DC, what happens in the very first millisecond is that this thermonuclear flash expands into a ball of fire that is one mile of pure fire. It’s 19 football fields of fire. 

Then the fireball’s edges compress into what is called a steeply fronted blast wave — as dense wall of air pushing out, mowing down everything in its path three miles out, in every direction, because it is accompanied by several-hundred-mile-an-hour winds. 

It’s like Washington, DC, just got hit by an asteroid and the accompanying wave. When you think about this initial 9-mile diameter ring, imagine every single engineered structure — buildings, bridges, etc. — collapsing.

There’s also a thermonuclear flash that sets everything on fire and melts lead, steel, and titanium. Streets nine miles out transform into molten asphalt lava. The details are so horrific; it’s important to keep in mind these are not from my imagination. These are sourced from Defense Department documents because the Atomic Energy Commission and the Defense Department have been keeping track of what nuclear bombs do to people and to things ever since the Hiroshima and Nagasaki bombings of 1945.

When all that happens, we’re in what you call “Day Zero,” and then the nuclear winter begins. What does that look like?

One of the big premises of the book was to take readers from nuclear launch to nuclear winter and the nuclear launch up to Day Zero takes place over this horrifying 72-minute period. As STRATCOM Commander General [C. Robert] Kehler said to me in an interview when we were talking about a nuclear exchange between Russia and the United States: “Yes, Annie, the world could end in the next couple of hours.” 

So nuclear winter begins in essence after the bombs stop falling and there is a process of mega-fires. The area around every nuclear detonation is going to ultimately result in what is known now as a mega-fire. You’re talking about 100 to 300 square miles of fire per bomb where everything in that area is burning until it doesn’t exist anymore. This is because, of course, there are no first responders anymore. There are no fire trucks, there’s no way to put anything out. 

With all of these explosions, 330 billion pounds of soot gets lofted into the troposphere. That is enough soot to block out 70 percent of the sun, creating a dramatic temperature plunge up to 40 degrees Fahrenheit, certainly in the mid-latitudes. 

Those areas, for example, from Iowa to Ukraine, that whole band of the mid-latitudes, the bodies of water in those areas become frozen over in sheets of ice. With that temperature drop, you have the death of agriculture and that is why nuclear winter after nuclear war will result in what is now estimated to be 5 billion dead.

And if I remember correctly, those models also estimated that in places like Iowa and Ukraine temperatures basically wouldn’t go above freezing for something like six years at least. Is that right?

I was reading Carl Sagan, who was one of the original five authors of the nuclear winter theory, who wrote about how after these bodies of water that get frozen over for years, after they thaw out and expose all the dead people, you then have to deal with the pathogens and the plague. Nikita Khrushchev, the Soviet premier during the Kennedy administration, once said to Kennedy when the two of them talked about this, that “after a nuclear war, the survivors would envy the dead.”

After all the reporting you did, are you confident that there are enough checks and guardrails in place to ensure that we’ll avoid a nuclear exchange if it’s at all possible?

Let me answer that question with a quote from the current secretary-general of the United Nations, António Guterres, who said, “The world is one misunderstanding, one miscalculation away, from nuclear annihilation.”

What does that really mean?

What it means is exactly what he said: that we could just have a mishap. We could have a mishap caused by a misinterpretation. A miscalculation would be one nuclear-armed nation thinking another nuclear-armed nation was doing something that maybe it wasn’t doing. 

This gets us into some of the crazy policies that exist on the books, things like “launch on warning” whereby once the United States learns that it is being attacked by an ICBM or a sub-launched ballistic missile, the president then has six minutes to decide how he should respond, with nuclear weapons. That’s what Guterres is talking about when he talks about a miscalculation. 

How much room is there for human agency in these command and control protocols? You always hear people say in presidential elections, “Do we really trust that guy with the nukes?” But is that the right way to think about this? 

You’re raising an existential question that everyone should be raising. 

We’ve been living in what some call a 79-year experiment. Yes, you could say, “Deterrence has held all these years.” Never mind the fact that there used to be two nuclear-armed nations, and there are now nine; never mind the fact that you have new technology factors coming into the mix. 

Never mind the fact that nuclear saber-rattling has suddenly become acceptable among world leaders. This is astonishing. If you look at history, this was never part of the rhetoric, particularly out of the mouth of a US president, as happened with the former President Trump.

When I began reporting this book, the fundamental question that I was trying to answer was not, “Is deterrence great?” but rather what if deterrence fails? The Defense Department predicates its nuclear arsenal on this idea that deterrence will hold. That is the fundamental assumption. It’s written everywhere. “Deterrence will hold.” 

Well, I also found a discussion with the deputy general of STRATCOM talking to his colleagues, not in a classified setting but in a somewhat rarified setting. What he said was this: “If deterrence fails, it all unravels.” 

I think it was former CIA Director Michael Hayden who told you explicitly that this process is designed for speed and decisiveness. It is not designed to debate the decision. On some level, I get that. But the automaticity of the whole process, given the stakes, is more than a little terrifying.

You better believe it is. And Hayden actually told that to members of Congress. And by the way, I believe that with the rhetoric from the former president, Donald Trump, all that talk about “fire and fury” with North Korea, it worried Congress to such a degree that they issued a number of reports that drilled down on a couple concepts that the public was not clear on. 

One of them had to do with what’s called sole presidential authority. So when Trump was saying, “I have a bigger button,” and that kind of rhetoric, Congress released a couple reports making clear that the president of the United States does have sole presidential authority. That means he needs to ask permission of no one to launch a nuclear war — not the secretary of defense, not the chairman of the Joint Chiefs of Staff, and not Congress.

You write something near the end of the book imagining that the secretary of defense, who’s the acting president in this hypothetical situation, what if this person has a crisis of conscience and wonders, “Is there really any point in firing these bombs and wiping out the other half of humanity?” 

And it’s pretty clear that there really isn’t any room for that because the whole logic of deterrence is predicated on the absolute promise that the process is fixed and automatic. That’s what makes it a deterrent. But then again, it imprisons the actors in this process so that they don’t really have any control over it.

Let me add something because Dr. Glen McDuff of the Los Alamos Laboratory, who is both a nuclear weapons engineer who worked on the Star Wars program during the Reagan administration and has served as the historian at the classified library at the lab. I asked him, “Do you think anyone would defy orders?” And he said, “Annie, you have a better chance at winning Powerball.”

Is there some near-future where in order to further reinforce the automaticity of this process, we just have AI controlling the whole thing from start to finish? 

I can’t imagine a worse nightmare scenario than bringing AI, or more machine-learning technology, into the mix. There’s an incredible amount of machine learning that is built into the system. For example, the satellite detects the launch and then that data is processed in space. About one-tenth of the way to the moon is where a geosync satellite sits and that data is processed and streamed down to the nuclear command and control bunkers in the United States. This is happening in seconds. But to the idea of putting an “AI” into the mix on the human decision-making level or identifying level, that seems like a recipe for disaster and is a reason why so many of the systems within the triad are still analog, not digital. In other words, they continue to be similar systems to when they were invented decades ago so that they can’t be hacked.