In “Preventing Nuclear War,” an essay published in the Bulletin of the Atomic Scientists, in 1981, the Harvard law professor Roger Fisher imagines a President in the White House, discussing nuclear war. Fisher was involved in the Camp David Accords, served as an adviser to both sides during the Iran hostage crisis, and helped to arrange the 1985 Gorbachev-Reagan summit. Much of his professional life was centered on managing pressured, consequential situations. During the Cold War—and arguably still today—the most consequential situation was the possibility of nuclear war. Fisher, who was a pilot during the Second World War, makes what he describes as a “quite simple” suggestion to reduce the chances of launching a nuclear attack: “Put that needed code number in a little capsule and then implant that capsule right next to the heart of a volunteer.” Like the rotation of military personnel who today trade off carrying the “nuclear football”—the briefcase that contains the nuclear launch codes—the person with the implanted capsule would be near the President constantly. In Fisher’s scenario, a Navy officer named George does the job. “The volunteer would carry with him a big, heavy butcher knife,” Fisher writes. “If ever the President wanted to fire nuclear weapons, the only way he could do so would be for him first, with his own hands, to kill one human being. The President says, ‘George, I’m sorry, but tens of millions must die.’ ” That there would be blood on the White House carpet is essential. “He has to look at someone and realize what death is—what an innocent death is.” The Fisher Protocol, as it was termed, makes vivid the reality that nuclear war, so often spoken of in the bloodless language of tactics and strategy, is unimaginably horrific.

That was more than forty years ago. The Fisher Protocol has not been implemented, not even in spirit. The Department of Defense, which is spending $1.5 trillion to replace or modernize our nuclear weapons and infrastructure, describes the logic of nuclear deterrence—have enough nuclear weapons to make others not launch nuclear weapons—as “sound.” According to the D.O.D., the presence of “multiple nuclear competitors” is a reason to develop even more nuclear capabilities. China has more than doubled its nuclear arsenal in the past five years. Russia, which has spent decades on modernizing its nuclear arsenal, has put into space what the U.S. military assesses as an anti-satellite weapon with nuclear potential—sometimes called a “Sput-nuke”—and in November, 2024, Vladimir Putin formally lowered the threshold for using nuclear weapons. The only remaining nuclear nonproliferation treaty between the U.S. and Russia, the New START treaty, is set to expire in February.

As the U.S. weakens its international alliances, non-nuclear countries with nuclear neighbors—South Korea, Germany, and Saudi Arabia, for example—may decide that acquiring such weapons is in their national-security interest. Some experts believe that the recent bombing of Iran’s nuclear-weapons development sites will only strengthen the country’s nuclear ambitions. No one has bombed North Korea, after all.

In July, the Bulletin of the Atomic Scientists marked the eightieth anniversary of the Trinity test by helping to organize the Nobel Laureate Assembly for the Prevention of Nuclear War, a gathering of nuclear experts and Nobelists held at the University of Chicago. The assembly put together a declaration that includes pragmatic suggestions for reducing the likelihood of ending civilization, among them that nuclear nations should engage in dialogue; develop back channels of communication; maintain human decision-making and oversight over nuclear control and command; and reaffirm commitment to the Outer Space Treaty, which bans nuclear weapons in space. Like the Trinity test, the Bulletin dates to 1945, and a similar meeting was held that year, also at the University of Chicago. Zachary Rudolph, the deputy director of the university’s Existential Risk Lab, told me that the opening remarks of the 1945 gathering, delivered by the sociologist Edward Shils, were “eerily prophetic.” “Much about Hiroshima and Nagasaki was not even declassified yet,” Rudolph said. “It was, like, one month after it happened. But what he anticipated then—mutually assured destruction, deterrence, arms races, proliferation—is still what we talk about now.”

Shortly after the conference, I spoke with one of the attendees, Herb Lin, a scholar of technology and national security at Stanford. Lin is seventy-three. He has a doctorate in physics from M.I.T., earned under Philip Morrison—a onetime student of J. Robert Oppenheimer’s—who worked on the Manhattan Project. Morrison rode out to the Trinity test site in a sedan with the plutonium core of the bomb in a case next to him; he later became a prominent activist for nuclear disarmament and nonproliferation. Morrison’s career echoes the arc of many Manhattan Project scientists, who went from believing that only a nuclear weapon could bring peace to believing that only disarmament could bring peace. In 1979, the year Lin finished his Ph.D., a large anti-nuclear protest followed the partial meltdown of one of the reactors at Three Mile Island, and, a few years later, a million people walked from Central Park to the United Nations complex calling for an end to nuclear weapons. Harry Belafonte, Bruce Springsteen, and Joan Baez were among the marchers for what was, back then, a widely held concern. Though the capacity to destroy life as we know it several times over remains, the issue is less present in the public mind now; it was scarcely mentioned in the run-up to the 2024 election.

I asked Lin why he believes that is the case. “I think the biggest difference between then and now is that we’ve normalized nuclear competition,” he said. A 2017 study led by the political scientist Scott Sagan, of Stanford, and Benjamin Valentino, a professor of government at Dartmouth, asked Americans if they would support the use of an atomic weapon that would kill two million Iranians if twenty thousand American troops were at risk—and the majority said yes. “You will hear people talk about nuclear war saying, ‘In this option, only two million people die,’ ” Lin said. “Think about that. I’m not saying thirty million isn’t worse than two million, but ‘only two million,’ when what follows is ‘die,’ is a completely insane statement.”

It has been eighty years since the first—and last—time nuclear weapons were used in war. The Nuclear Matters Handbook of the D.O.D. notes that the policy of deterrence “has done its job,” but there’s also the matter of dumb luck. “One of the most interesting developments in psychology has been coming to understand that people are systematically irrational,” Lin said. “Yet most theories of nuclear deterrence”—they won’t attack us because they know we can attack them—“are premised on rational actors.” The roll call of people across the decades who have had access to nuclear weapons includes many who are considerably less rational than the irrational rest of us. “And so the whole edifice of deterrence is, from the beginning, in trouble,” Lin said.

How nuclear risk is perceived is an area of special interest for Rose McDermott, a political scientist at Brown who also has a degree in experimental social psychology. “Believing in rational stability is a salve. A way to avoid knowing that we live on the knife’s edge of extinction,” McDermott told me. “It’s avoiding the truth that there are unsuitable leaders in control of weapons that can destroy humanity.” Neither economic sanctions nor the more than a hundred thousand Russian soldiers who have died in the war on Ukraine appear to have had much of an impact on Putin’s decision-making. “We can look at history,” McDermott said. “Stalin, for example—we know there are certain personality types who are more concerned with their own status than with the well-being of the body politic. We know there are leaders driven by narcissism and paranoia, who can make decisions with very few institutional or operational constraints.”

Such leaders are not always beyond our borders. McDermott said, “There shouldn’t be sole authority,” referring to countries, such as ours, that ultimately leave the decision to launch nuclear weapons to one person. In the current protocol, the President can act without congressional or military approval; although the Secretary of Defense is expected to be part of the chain of communication, the secretary does not have a vote. Recently, Trump responded to a tweet by the Russian official Dmitry Medvedev about Russia’s nuclear capabilities by saying that he was sending two nuclear submarines over to “appropriate places.” France also has a sole-authority policy, and it seems reasonable to assume that North Korea does as well; we have limited information on the policies of other nations.

“Another thing we’ve learned is that decision-making is a skill,” McDermott said. She talked about how a series of questions can help people make better decisions. “You actually go through a checklist, rather than do something you absolutely can’t pull back that could destroy an entire country or planet. It would be questions like, ‘What is your confidence that this incoming attack is real?’ ‘What are your values?’ ” She compared this to the protocols that were introduced in hospitals around twenty years ago. “It’s those checklists—like making a mark on which knee to operate on, verifying that you’re operating on the right person.” These ideas can seem overly simple, even silly. “But it turns out that instituting these protocols in hospitals saved more than a hundred thousand lives.”

Some nuclear-risk situations, however, don’t lend themselves to protocols. During the Cold War, the U.S. operated a radar detector in Greenland. Worried about Soviet disruption, the U.S. routinely flew a B-52 with nuclear bombs on board near Greenland, to report any possible attack. “In January, 1968, a crew member of the B-52 had stuffed some extra seat cushions near a heating vent, and they caught on fire,” Sagan, the political scientist, told me. The crew parachuted from the plane, which crashed onto the ice and exploded. Luckily, the nuclear bombs didn’t detonate. “I call it ‘the Problem of the Redundancy Problem,’ ” Sagan said. “That whole logic, of putting a thermonuclear-armed bomber out as part of your warning system—that’s the kind of complexity that I worry about so much.” An increasingly complex system generates its own kind of unforeseeable errors.

Sagan described another major pathway to unintended nuclear catastrophes. The fear of a surprise attack could make a nation consider a preëmptive attack, leading to war between two nations that don’t want a war. For this and other reasons, some countries, including China and India, have made what are known as “no first use” vows.

There’s the problem of someone deliberately starting a nuclear war, and then there are errors and misinterpretations: a radar that mistakes a moonrise for a Soviet missile launch, as happened in Greenland in 1960; a repair officer’s error that leads to an explosion in a silo housing an ICBM, as happened in Damascus, Arkansas, in 1980; nuclear-warhead-armed missiles that are mistakenly loaded onto an Air Force bomber that is supposed to be carrying dummy missiles, as happened in North Dakota in 2007.

What has saved the day in many false-alarm and other near-miss situations has quite often been simple human relationships. During the Cuban missile crisis, for example, U.S. Attorney General Robert Kennedy held back-channel meetings with the Soviet Ambassador Anatoly Dobrynin. “The process of arms control isn’t only about reducing weapons,” McDermott said. “Even if no agreement is reached, the process of meeting and negotiating needs to be reactivated, so that there’s a conversation, and so that relationships are maintained.” The many recent firings at the Pentagon, along with Secretary of State Marco Rubio’s firing of more than a thousand members of the State Department, means less personal continuity in relations between the U.S. and the rest of the world.

In his 1981 essay, Fisher identified other human factors that make the world more perilous, among them the desire to win, a phrase that he puts in quotes and which he characterizes as a concept that makes more sense for a game of football than for one of nuclear football. “We want victory; we want power; we want peace. Exploding nuclear weapons will not help us achieve any one of them,” he writes.

Fisher had friends at the Pentagon to whom he presented his idea about the nuclear codes, the surgically embedded capsule, the butcher knife, the blood stains on the White House carpet. According to Fisher, his friends reacted with tremendous concern, saying, “My God, that’s terrible.” Having to kill someone would affect the President’s judgment, they pointed out. They saw that as a big problem: “He might never push the button.” ♦