Five Billion Dead, and One Person Has Six Minutes

Lex Fridman opens with the darkest possible question: how many people would a nuclear war between the United States and Russia kill? Annie Jacobsen, whose book *Nuclear War: A Scenario* traces an exchange from first launch to nuclear winter, does not hedge. She comes back with a very dark answer and a very big number: five billion people. To write the book she interviewed retired secretaries of defense, a nuclear submarine force commander, a Stratcom commander, a FEMA director, and nuclear weapons engineers. They belonged to different agencies, different decades, and different parties. The common denominator, she says, was that nuclear war is insane.

She deliberately leaves out the politics, the posturing, and the national-security maneuvers that might lead to a launch. She wants readers to see, in appalling detail, what happens once the buttons are pressed. In the first 72 minutes, deaths run from millions to tens of millions to hundreds of millions. Then nuclear winter arrives, and the billions die of starvation. Hundreds of thousands of people in nuclear command and control sit ready behind a veil most civilians never think about. Richard Garwin, who drew the plans for the 1952 Ivy Mike thermonuclear shot, told her it takes only one nihilistic madman with an arsenal to start it.

The war, she says, begins in space. A constellation of early-warning satellites called SBIRS sits in geosynchronous orbit, about a tenth of the way to the moon, and sees the hot rocket exhaust of an ICBM in a fraction of a second. That detection triggers a U.S. policy named launch on warning: the United States will not wait to absorb a nuclear attack. It will fire back while the incoming warheads are still in the air. A ground radar confirmation is required as well, but the whole sequence is only a couple of minutes. Then the president is told the clock has started.

Stacked on top of that policy is sole presidential authority. In a democracy, Jacobsen notes, you might think you cannot just start a war. You can start a nuclear war if you are commander in chief; in fact, you are the only one who can. The president asks permission of no one — not the secretary of defense, not the chairman of the Joint Chiefs, not Congress. Los Alamos declassified the origin story for her book: the war would unfold so fast that only one person can be in charge. Reagan, in his memoirs, called the six-minute window irrational and asked how anyone could unleash Armageddon on a blip on a radar scope. Former defense secretary Leon Panetta told her that as White House chief of staff he watched President Clinton avoid the nuclear issue; only when Panetta himself became secretary of defense did he feel the weight, because he would be the person the president turned to.

Presidents arrive in office fluent in deterrence: huge arsenals exist so that nuclear war never happens. The philosophy collapses, Fridman and Jacobsen agree, the moment the first missile leaves the rail. There is no Battle of New York. Cold War planners called it push-button warfare, and that is still what it is. The United States has 1,770 nuclear weapons deployed, meaning they can launch in as little as 60 seconds, or up to a couple of minutes; some bomber weapons might take an hour. Russia has 1,674 deployed, with comparable systems. Among nine nuclear-armed nations there are about 12,500 warheads in all.

Those figures are tracked every year. At the Federation of American Scientists, Hans Kristensen runs the Nuclear Notebook: treaty parties report numbers, and intelligence services try to sort what is disclosed from what is hidden. Jacobsen’s point is speed. The arsenals can go in seconds and minutes. Mutual assured destruction says no one would launch because everyone would die. Every top-tier national security adviser she interviewed said it could happen. None said it never would.

Strategic weapons are the big systems. America and Russia both keep a triad: silo-based ICBMs that cross a continent in roughly 30 minutes; nuclear-capable B-52s and B-2s, which take hours and can be recalled; and submarines. Land-based ICBMs cannot be recalled or redirected once they leave. Reagan, at a 1983 press conference, misstated that submarine-launched ballistic missiles could be recalled. They cannot. The man in charge of the arsenal, she notes, did not know that.

Tactical nuclear weapons are smaller warheads designed for the battlefield, which is what Russia has been threatening. Every arms-control treaty, Jacobsen says, sits on a single line: no nuclear use. Cross it and there is no off-ramp. The jargon is “escalate to deescalate.” Translated, she says, they launch one, we launch 80, and pound the other side until it stops. U.N. Secretary-General António Guterres’s line is the one she keeps quoting: we are one misunderstanding, one miscalculation away from nuclear Armageddon.

Officials would not give Jacobsen a precise flight time. She found it in Herb York’s papers at the Geisel Library in San Diego. York, the Pentagon’s first chief scientist when ARPA was becoming DARPA, wanted the number in seconds and hired the JASON scientists to produce it. Boost phase lasts five minutes: the rocket leaves the pad with fire underneath, which is what the satellites see. After that the exhaust disappears and space can no longer track it that way. Midcourse is 20 minutes, a warhead flying ballistic at 500 to 700 miles up. Terminal phase is the last 100 seconds, reentry and detonation. Total, from a Soviet launch pad to the U.S. East Coast: 26 minutes and 40 seconds.

Her scenario opens with what Washington fears most, a bolt-out-of-the-blue from a nihilistic madman: North Korea sending a one-megaton warhead at the capital. MIT professor emeritus Ted Postol did the math for Pyongyang to the East Coast and got 33 minutes. About half an hour, either way. Once readers can see the three phases, she says, the rest of the decision logic becomes thinkable — almost like a terrible poem. Nuclear war is simple. People have made it complicated. John Rubel, in her book, calls the system one of mass genocide. After 75 years with two nuclear superpowers, her sources agreed to talk because they are near the end of their lives, and they say the world is closer to this than ever. The only bright side she will claim is that the answer lies in communication.

Submarines crush the half-hour window. A nuclear-armed, nuclear-powered boat is, in the phrase she was given, as dangerous to civilization as an asteroid: unstoppable and unlocatable. Admiral Michael Conner, former chief of the nuclear submarine forces, told her it is easier to find a grapefruit-sized object in space than a submarine under the sea. They move constantly — American, Russian, Chinese, maybe North Korean. A Defense Department budget map she found showed them coming within a couple of hundred miles of the U.S. East Coast. From that range, launch to strike is under ten minutes. The maps are drawn after the fact from underwater surveillance. In real time, you cannot find the boat.

The submarine launches from about 150 feet down. The missile is pushed from the tube, the boosters ignite outside the water, and then it is in boost phase. Trillions of dollars, she says, have been spent so that nuclear war never happens. She calls the boats the handmaidens of the apocalypse.

The United States has 400 underground silos. The missiles are Minuteman, named for Revolutionary War militia, and the Washington joke is that they are not called that for nothing: they can launch in one minute. The president orders, they leave 60 seconds later, and they fly for a little more than 30 minutes. Submarines take about 14 or 15 minutes from command to actual launch, depending on location and depth. Anyone can look up the silo coordinates; they were on maps before they were on Google, because the United States is a democracy. Russia and North Korea also use road-mobile launchers — giant ICBMs on multi-axle trucks that move so they cannot be targeted. America does not, Jacobsen says, because the public would not accept an ICBM driving through town. That is how the Pentagon justifies the submarines as second-strike insurance.

The nuclear football is a leather satchel carried by a military aide who follows the president, and the vice president, 24 hours a day. Former Secret Service director Lew Merletti told her about a trip to Syria: Clinton and Assad’s men got into an elevator, Assad’s security tried to keep the aide out, and it became a standoff. There was no world in which the president would be separated from the bag. People talk about UFO secrets as if they were the summit of classification. She laughs that off. What has never leaked are the presidential emergency action directives inside the satchel. What is known comes from an aide named Buzz Patterson, who described the strike options as a piece of laminated plastic, like a Denny’s menu. The president picks targets and weapon systems off that list.

Three command bunkers run the war. The National Military Command Center under the Pentagon is the beating heart. Cheyenne Mountain is the brains. The bunker under Strategic Command in Nebraska is the muscle: the Stratcom commander takes the president’s launch order and directs some 150,000 people. In that bunker, beside the nuclear strike advisor, sits a weather officer. If the president asks why this target and not that one, the weather officer has to say, very fast, how many people will die in minutes, weeks, months, and years from fallout, which depends on the wind. The human detail next to the mechanization, she says, is grotesque. The president then takes a card from his wallet, colloquially the biscuit. The codes match an item in the football, an officer under the Pentagon answers in a call-and-response, and the order goes to Stratcom. Then it is done.

After giving the order, the Stratcom commander is supposed to run out of the building and onto the doomsday plane. The plane’s destination is a circle over the United States, so nuclear weapons can still be launched from the air after ground systems are destroyed. The contingency has been in place since the 1950s. On 9/11, President Bush was on that aircraft. A Stratcom deputy, speaking to a closed group, put the paradox in one sentence Jacobsen cannot shake: yes, deterrence will hold, but if it fails, everything unravels.

At a dinner party, a person who Jacobsen thought should have known better told her none of this could happen, because of America’s powerful interceptor system. He was wrong. The United States has 44 ground-based interceptors, period: 40 in Alaska, four at Vandenberg. Their test success rate is around 50 percent. The program is on strategic pause, meaning the missiles are there, but the work of making them more effective has stopped, because they cannot be made more effective. Set those 44 against Russia’s 1,674 deployed warheads, and the arithmetic is the argument.

The interceptors have to leave around minute nine of her scenario, after boost phase ends and ground radar confirms an incoming ICBM. Inside the rocket’s nose cone is an exoatmospheric kill vehicle with no explosive. It is meant to hit the warhead by force, one object at about Mach 20 meeting another in space. That interception is even possible, she says, is remarkable. That it works only half the time is the operational fact. There is no shoot-and-look: not enough time to miss, then fire again. In the book she has the system fire four, which matches what she was told — one to four — because you still have to worry about the next missile. All four miss. Using 10 percent of the interceptor force on one warhead and missing is, she says, entirely plausible.

People imagine something like Iron Dome. The United States does have THAAD on land and Aegis at sea, and those systems are good at shooting down some rockets, one at a time. They cannot handle a mass raid, and they are built for tactical targets. THAAD is all deployed overseas; Aegis is all at sea. After 9/11, Congress considered putting Aegis and even THAAD along the West Coast against North Korea. It has not done so. One nuclear weapon that gets through is full-out nuclear warfare. More warheads are not the solution, she says, and neither are more interceptors.

The chain above assumes detection is right. There have been at least six known close calls. Former secretary of defense William Perry told Jacobsen about 1979, when he was still deputy director of research and engineering and had the night watch. He was told Russia had launched, ICBMs and submarine missiles both, and he was about to notify the president that the six-minute window had begun. A VHS training tape of a nuclear-war scenario had been put into a Pentagon system by mistake. The false attack showed up in the bunker under the Pentagon and the one under Stratcom, because they are linked. Perry described the pause in his spirit, mind, and heart. It is the only firsthand story of that second that she has ever heard. People in the system used to treat talking about it as taboo. Perry spent a career building weapons and later decided the enterprise was madness, thinking of grandchildren inheriting arsenals whose origin stories did not contain much wisdom.

Most Pentagon nuclear games remain classified; one recently “declassified” war game, Proud Prophet, is almost entirely redacted. What the declassification did was let a few participants talk. In 1983, with 60,000 nuclear weapons in the world — the peak was 70,000 — Reagan ordered the game, and everyone who mattered played. No matter how the war started, with NATO or without, it ended in Armageddon. In the 40 years since, Jacobsen says, the policy response has been to lean harder on deterrence, which is just a statement: deterrence will hold. Proud Prophet is what happens if it does not. Former Stratcom commander General Hayden, saber-rattling at North Korea, said if they launch one, we launch one. The actual doctrine, she says, is closer to they launch one, we launch 80.

The fact that chilled her most is geometric. If a madman launched from North Korea and the United States had to fire ICBMs in response, those missiles do not have the range to go another way. They must fly over Russia. Panetta confirmed it. You would be asking an adversary, in the middle of saber-rattling, to believe that 82 warheads overhead are not for Moscow. Russia’s early-warning satellites, a system called Tundra, are badly flawed. Pavel Podvig, a leading Western expert on Russian nuclear forces, told her Tundra can mistake sunlight for flames and clouds for a launch. For years Russian officials claimed they had no launch-on-warning policy, that they would wait to absorb a blow. American experts often called that propaganda. Two years before this conversation, Putin said the position had changed: once they see the trajectory of the missiles, they launch.

Fridman wants to believe that in a real crisis, someone near the top would pick up the phone and cut through politics. Jacobsen hopes so and doubts it. During the war in Ukraine, news reports briefly claimed a Russian missile had hit Poland, a NATO country. It was a mistake, but it ran as breaking news for hours. Thirty-six hours later, Chairman of the Joint Chiefs Mark Milley admitted he had not been able to reach his Russian counterpart in that window. A ground war was on, nuclear weapons were on alert, and the two militaries’ senior officers could not get on the phone.

The people in the room with the president are not brewing coffee and strategizing. He has his secretary of defense and the chairman, maybe the vice chairman, and they have a second job at the same time: continuity of government. While they advise on the counterstrike they are also thinking about how to keep the government functioning once the bombs land, which means getting out of the Pentagon and into bunkers. Jacobsen learned a term from the historical literature: jamming the president. It has never been tested, but doves worry that military advisers will push an aggressive counterattack immediately. Speed, she says, is not conducive to wisdom. The fight in the room is which targets, not whether to fire.

Did warmongers ever believe a nuclear war could be won? The record says yes. After the Soviet Union got the bomb in 1949, declassified early planning documents show that in the 1950s the generals and admirals running nuclear command believed the United States could fight and win despite hundreds of millions of dead. That was the prevailing thought. Only later came the joint Reagan–Gorbachev statement: a nuclear war cannot be won and must never be fought. Before that, people were preparing to win one.

Fridman asks whether cognitive ability belongs in this discussion. Jacobsen does not treat it as ageism. Judgment and cognition, she says, are the two biggest issues in a nuclear launch, and she looks at a field of candidates — one with cognitive problems, one with judgment problems — and asks where the youngish, thoughtful, forward-looking civil servant is. She knows that sounds fantastical. She wishes it were not. The character of the commander in chief matters for any war, she argues, because the wars since World War II have been fought by the young. With nuclear weapons there is no leeway: one person has to make the call while Secret Service tries to move him and Stratcom demands launch orders.

Would a Stratcom commander refuse? She put that to Glenn McDuff, a Los Alamos nuclear weapons engineer who had run the lab’s classified museum. His answer: you have a better chance of winning Powerball. You do not become Stratcom commander unless you follow orders. During COVID, when President Trump’s dealings with North Korea swung from threat to friendship, Stratcom leaders spoke to Congress more publicly than she had ever seen, and the defy-orders question came up. The caveat to McDuff, she says, is that if the commander judged the president’s behavior unreliable, you would be in radical territory. Nuclear command itself, she notes after talking to Obama’s first cyber chief, General Touhill, still runs on analog systems. You cannot hack what has no digital interface. The cyber problem she worries about is what happens to communication after a city is hit, when chaos takes over.

Washington’s great fear, she was told, is a bolt-out-of-the-blue attack on the capital, with the Pentagon as the aim point. That is how her scenario begins. She is not sure which is worse: writing the effects in graphic detail, or the fact that the Defense Department has documented them for decades. The data start in Hiroshima and Nagasaki, then run through more than 200 atmospheric tests. Scientists know, in millimeters, that seven and a half miles out the upholstery on cars will spontaneously combust and pine needles will start more fires. A one-megaton fireball is 19 football fields of fire at 180 million degrees. Nothing remains.

Then the blast wave, several hundred miles an hour, pushes out like a bulldozer. FEMA manuals describe rubble 30 feet deep as buildings go over six, seven, eight, ten miles out. Mega-fires follow. Third-degree radiation burns are the familiar horror; she asks whether you knew there were fourth-degree burns, the wind ripping skin off faces many miles away. The stem of the mushroom cloud creates a sucking action at 300 miles an hour. People miles out are pulled into it. When you see the cloud, she tells Fridman, that would be people — remnants of people and things — in a 30- or 40-mile-wide mass that blocks the sun. Radiation poisoning comes after that.

Those who survive the blast, shock wave, and firestorm, she writes, learn they are entirely on their own. The fight for food and water begins. Obama’s FEMA director, Craig Fugate, gave her those details. FEMA plans for asteroid strikes and other low-probability, high-consequence events, and in ordinary disasters it does population-protection planning. After a bolt-out-of-the-blue nuclear strike, Fugate told her, there is no population protection. Everyone is dead — metaphorically, and almost literally. At that point you hope you stocked Pedialyte. In the scenario the FEMA director is pulled off the road into a helicopter for Mount Weather. The military bases people imagine as sanctuaries run on diesel. When the fuel stops, the generators stop.

After the first hour comes the rest of the Earth. Brian Toon, one of the original five authors of nuclear-winter theory and a student of Carl Sagan, told her that better climate models now show nuclear winter is worse than the 1980s versions the Pentagon once dismissed as Soviet propaganda. In her scenario the bombs stop falling 72 minutes after first launch. Then the fires. Each weapon, by Defense Department figures, produces a mega-fire of 100 to 300 square miles. A thousand or 1,500 weapons put forests, cities, and high-rises into the air as soot — some 300 billion pounds, Toon says — and the sun goes out. Ice sheets form; bodies of water in places like Iowa freeze for ten years; agriculture fails. After eight to ten years the thaw arrives under a shredded ozone layer, sunlight itself poisonous, pathogens and plague coming with the melt. Small-bodied animals reproduce; large-bodied animals, including people, go extinct. Toon’s comparison is the asteroid 66 million years ago that killed the dinosaurs and 70 percent of species. Nuclear war would likely do the same. The difference: you cannot do anything about an asteroid. You can do something about nuclear war.

Would anyone survive? Carl Sagan, in *The Cold and the Dark*, wrote essays about a return to the worst, most base versions of people: no civil society, no rule of law, fighting over scraps, back to a hunter-gatherer state. Einstein’s line belongs here. Asked what weapons World War III would be fought with, he said he did not know, but World War IV would be fought with sticks and stones. To think about what would be buried, Jacobsen went to Göbekli Tepe in Turkey, the oldest known archaeological site. She interviewed Michael Morsch, who as a graduate student was with Klaus Schmidt when they found it in the early 1990s. Locals knew a wishing tree on a hill. Morsch tripped on a stone that turned out to be the top of a 12,000-year-old sculpted pillar. Hunter-gatherers built the place. No one knows why.

That image lets her ask, as a person on Earth for a finite time, what nuclear war would actually be like afterward. Would someone in the future become an archaeologist and rediscover us? Would civilization rebuild computers? She hopes we never have to find out. What is buried becomes, strangely, optimistic: if you can see the wishing tree, you ask what they wished for, then what you wish for. She thinks everything starts around the dinner table. Put your eyes on a subject and the conversation expands. The long arc of civilization, she says, makes you want to talk more with adversaries. Fridman asks about the Great Filter: maybe other civilizations hit something like nuclear weapons and failed. She is not sure, but Ed Mitchell, the sixth man on the moon, gave her an answer she is willing to borrow. Coming home in 1971 he looked down at a planet in the Vietnam War, then back into the galaxy, and felt a near-life experience: inner life and outer life connected the way a person is connected to the galaxy. That, she says, is the large version of the wishing tree, and it is the kind of thought that de-escalates conflict.

Her book *Phenomena* is the secret history of U.S. government work on extrasensory perception and psychokinesis. She spends her career on machines of war; the military and intelligence services, she found, have never lost interest in the mind. Before science, the supernatural ruled. Leaders went to the Oracle of Delphi to learn what would happen. Most people think the ESP programs started in the 1970s. The files show they began right after World War II, because the Nazis had a large occult program — ESP, psychokinesis, astrology — and Hitler and Himmler were deep in it. After the war the Nazi documents were split: half to the Soviets, half to the United States. A psychic arms race started, with the same logic as the nuclear one: each side wondering what the other had.

From the documents, the CIA believed there was something legitimate about ESP. It was uncontrollable and unreliable, but it existed, and an intelligence agency cares less about why than about how to use it. Then placebo effects crept in. When the military took over and tried to teach people to be psychic, she thinks that was a disastrous idea. The programs have not gone away. The Navy has one built on war-on-terror reports of soldiers who suddenly knew not to walk down a path because of an IED — “spidey sense.” Her deepest takeaway is the Thomas theorem: if people define situations as real, they are real in their consequences. The word placebo itself, she notes, traces to Dr. Henry Beecher and, in her telling, to the CIA. Fridman asks if that origin story troubles her. She has mixed, even conflicting, feelings about the CIA, as she thinks a reporter should. Her first encounter was Area 51’s aerial reconnaissance, built to prevent World War III. She interviewed Hervey Stockman, the first man to fly a U-2 over the Soviet Union. Then she wrote *Surprise, Kill, Vanish* and learned about the kill programs. That is a different set of issues.

Area 51 is, in her description, a top-secret military facility inside a top-secret facility inside the Nevada Test and Training Range. It began as the CIA’s place to build the U-2 away from the public, then the A-12 Oxcart, a CIA Mach 3 stealth reconnaissance aircraft of the early 1960s, before the two-seat SR-71 the movies made famous. She interviewed 75 people who lived and worked there for long stretches, mostly before the 1990s, which is still classified. Colonel Slater, the base commander, had flown U-2 Black Cat missions over China’s Lop Nur nuclear site. One day the crew wanted lobster in the middle of the Nevada desert and had it flown from Massachusetts so fast it did not need ice. Those men are gone. She is glad she got the details.

Were there ever aliens at Area 51? She has interviewed hundreds of people across national security, military, and intelligence programs. Personally, she has no reason to believe aliens have visited Earth. She has no information that causes her to conclude that. Many of the primary players in today’s “aliens among us” narrative appear in *Phenomena*, including astrophysicists who fundamentally believe it. They differ. What colors her thinking is firsthand knowledge of CIA strategic deception. The U-2 flying at 70,000 feet, sunlight on the wings, looked like a UFO to people who did not think airplanes could fly that high. The CIA opened a UFO disinformation office, headed by Todos Odarenko, for that reason. Not every sighting was a U-2. But looking across decades of the same stories, she sees a catchall designed to keep the public’s attention on that, not on this.

She tells people to search for Paul Bennewitz. In the late 1970s and early 1980s he thought he had seen a UFO. Air Force intelligence, working with the CIA and other agencies among the 17, ran a deception campaign that destroyed him and sent him to a mental institution, because he was seeing technology at Kirtland Air Force Base that he was not supposed to know about. Ufologists are sure aliens are here; “skeptics,” she says, tend to be self-righteous. She does not want to be either. She calls herself agnostic. The Thomas theorem applies again: “we are being lied to” becomes real in its consequences, and the spiral loses a center. She leaves a door open: or she is wrong, and there are aliens among us. Her counter is the Chinese balloon. The Defense Department guards airspace so jealously that a balloon was front-page news. She does not believe a handful of officials who say they have lost control of the skies over a craft they cannot explain. Jacques Vallée, the scientist she calls the grandfather of ufology, stands at the opposite pole and told her the most interesting sentence she has on the subject: it is not a military problem, it is an intelligence problem. The closest she can get to that is consciousness. Fridman calls the flying-saucer story a trivialization of what an alien civilization would actually be.

At the end of *Area 51* she offered an explanation that drew years of criticism: Roswell as a Stalin black-propaganda hoax, a craft with surgically altered, child-size aviators, modeled on the panic of the *War of the Worlds* broadcast, meant to swamp American air defense. On this podcast she names the source for the first time on tape. It is Al O’Donnell, the engineer who armed, wired, and fired 186 of the more than 200 U.S. atmospheric nuclear tests, a Manhattan Project veteran of Operation Crossroads, known as the trigger man. She knew his children and grandchildren. He told it as a tearful confession: he had been involved in an American counterpart program in which people died in human experiments, and he wanted it written so that this kind of thing would never happen again. If you give people too much power, they take advantage of it. His wife of 60 years and his children learned of it when the book came out. The family sat him down and asked him to say it was not true. He said it was. He told her she could name him. If you want to call the source unreliable, she says, you have to explain why he held clearances higher than almost anyone’s. He was the man who fused the bombs.

Diplomacy is the first option, war the second. When both are inadequate, Jacobsen writes in *Surprise, Kill, Vanish*, there is a third: covert action, including assassination. The Cold War record of CIA involvement, she says, is non-negotiable. Eisenhower’s program was even named the Health Alteration Committee. Modern language is targeted killing and executive action; a drone strike is assassination by another name. John Rizzo, the CIA’s longtime lead counsel, talked to her up to the edge of what can legally be known. He had been the fall guy for enhanced interrogation, and he explained without bitterness how the programs are made legal: if the president says they are legal, they are legal. Executive Order 12333 says the United States does not assassinate. It can be overwritten by another order. That is straight from Rizzo.

The military operates under Title 10, with rules for a war theater. The CIA operates under Title 50, which, in her paraphrase, means what the president wants, the president gets. Before she reported it openly, Title 50 was rarely discussed; now operators talk about it on podcasts. The cleanest example is bin Laden. The United States was not at war with Pakistan, so a military operation there did not fit. CIA director Leon Panetta built a capture-or-kill plan that was, in practice, a kill. SEALs, Delta, and Special Activities all trained the same mission at a secret facility in North Carolina; the SEALs got the go. SEALs live under Title 10, so they had to be “sheep-dipped” onto CIA books — a term that comes from Area 51, where Air Force U-2 pilots were dipped into the agency to do things that broke the law. The night of the raid was a CIA mission. The CIA could go into Pakistan and kill someone. The military could not. The uniforms had no markings. The origin of that deniability is Vietnam’s MACV-SOG, whose cross-border jobs were called suicide on the ground: men in what were essentially pajamas, CIA weapons without serial numbers, so that if they were captured the government could say it did not know them.

No one walked her through a complete assassination, because those remain classified. Reconnaissance, she has heard a lot about: you “make book” on the target, photographs run up the chain, so you do not kill the wrong person. Imad Mughniyeh, Hezbollah’s chief of operations, was the world’s most wanted terrorist before bin Laden. For 20 years there was not even a photograph. He resurfaced in Saudi Arabia — a dirty game, she says, given who Saudi’s enemies are. A SEAL tracking him for a CIA–Mossad job panicked, by Billy Waugh’s account, around 2005. They sent Waugh, then in his late seventies. In 24 hours he knew the house. He sat in a Sudanese café across the street, then moved into a dumpster between the café and the apartment, and photographed Mughniyeh coming out with dry-cleaning hangers over his shoulder, at home in the neighborhood. The shots went to the station; Waugh left the country. Mughniyeh died in a car bomb in Damascus. Mossad took the credit. The CIA, she says, often prefers it that way.

Waugh is why she talks about willingness. The longest-serving CIA operator in her book, he had never packed for a personal trip. He had worked in 62 countries on agency orders, eight Purple Hearts from Vietnam, after bin Laden in Afghanistan at 72, after Gaddafi at 82. At 89 he showed up at her house with a giant suitcase of dry cleaning on hangers for a jungle trip to the Oscar Eight site. On the taxi ride home from Hanoi he pulled a rolled American flag from the small bag her husband had given him. He had not wanted her to see it going in, because bringing it was not legal. He carried it to honor friends who died there 50 years earlier, then gave it to her. It hangs in her office. That, she says, is worth as much as any mission she can get declassified, because the country is made of people. Asked about Alexei Navalny’s death, she goes to the record: Russia has been assassinating dissidents since the early Cold War. She tells the story of KGB assassin Khokhlov, who knocked on his target’s door, had a crisis of conscience, defected, and was later poisoned with polonium and survived. To her eye, Navalny looks like Russia acting as it has always acted. Compared with CIA programs, she sees in the Russian ones a sadism she has not found on the American side, and in the American files a thread of trying to keep some semblance of democratic ideals — not “fighting fair,” she is careful to say, because killing is not fair.

Five years earlier, Jacobsen would have answered a surveillance question with the NSA’s haystack. The agencies still cannot find the needle, and they keep building the stack. What changed is that civilian companies have done the Defense Department’s biometric job for it. In *First Platoon* she wrote about young soldiers in Afghanistan who unwittingly became part of an effort to capture biometrics on 85 percent of that country’s population — fingerprints, faces, DNA, iris scans, the tools that let you tag, track, and locate. China then emulated the program. Meanwhile people upload identifiable faces and self-written narratives to Instagram, Facebook, and X, and the images enter the databases. The police chief in El Segundo, a defense-contractor dense patch near Los Angeles airport, had facial-recognition software on his phone. He sent her around the corner, glanced down, and by the time she walked back he had her identity, facts, and images without raising the device. During the war on terror, ARGUS-IS, a drone that can read what is on a golf ball from 20,000 feet, was denounced as Big Brother. Pat Bilgin, one of its lead engineers, later told her that what people turn over about themselves already exceeds what the drone can see, because they do it willfully. In a war theater the chain was find, fix, finish. Finish meant a drone strike.

On Hiroshima and Nagasaki, she has interviewed people of every view; most who were there say the bombs ended the war. The most meaningful conversation was with Al O’Donnell, a veteran of Okinawa who had been slated to invade the Japanese home islands at almost certain death. For him the weapons saved his life, his friends, and the war. Could Hitler have built the bomb first? In Albert Speer’s memoirs, researching *Operation Paperclip*, she found Speer recording Hitler’s refusal: no, that is Jewish science. Racial prejudice, in that account, is why Germany did not get the weapon. Fridman asks whether the two bombs also sent a signal that made 70 years of deterrence possible. She does not quite take the gift. What confounds her more is the decision, after the cities were destroyed and the war was over, to build the thermonuclear bomb. Inside the Manhattan Project it was called the evil thing. Atomic weapons destroy cities. Thermonuclear weapons destroy civilizations. Everybody, she says, always wants more power — literally more power. That is the danger.

She opens the book with Churchill: the story of the human race is war; except for brief interludes there has never been peace, and before history, murderous strife was universal. Will there always be war? The optimistic answer is that we could evolve beyond it. Ancestors had not developed consciousness far enough to build the tools we have; the hopeful version is that we evolve past brute force. She wants only a small part, as a storyteller who brings information so people can ask the question with friends and family. Asking why we do not evolve beyond war is already part of the answer. Fridman says her book is a reminder that civilization ends this century. She catches the verb. He means *could* end, not that it ends. As they say, it is time to come back from the brink, and time to have the discussion while we are still talking. AI, cyber, and autonomous weapons, he adds, will only complicate the century.

What gives her hope? She is a fundamentally optimistic person, she says, even when the subject is grim, because she believes in evolution. She has two children, Jet and Finley, and her husband, and she is stirred by legacy: being a good example, passing on what you know, believing in the next generation. The Cold Warriors she interviews say a version of the same thing — look at what the civilization has done, look at where it is going, whether exoplanetary travel or AI. The human appetite for conflict has to be reconfigured, because the peril of new technology is growing faster than the average person can keep up. Fridman thanks her. She thanks him. He closes with Kennedy on secrecy: the word is repugnant in a free and open society.