An unarmed Minuteman III intercontinental ballistic missile launches during an operational test from Vandenberg Air Force Base, California.REUTERS/Ian Dudley/U.S. Air Force photo/Handout via Reuters

It was once my job to help protect the United States from a nuclear attack. From 2009 to 2012, I was stationed at Tinker Air Force Base in Oklahoma City, serving as a flight officer aboard the E-6B Mercury. The plane can act as an airborne command post in case of a nuclear attack, and its crew members can relay orders to launch nuclear weapons. Our mission was nuclear deterrence—preventing a state from attacking America with its nuclear weapons through the implicit threat of America’s own nuclear-weapon capability.

These days, this mission is especially pressing. Tensions between the United States and North Korea have escalated, in part due to the adversarial relationship between Donald Trump and Kim Jong Un, who, in a New Year’s Day speech, emphasized his country’s ability and willingness to strike the United States, and noted that “the nuclear button is always on the desk of [his] office.” Several days later, President Trump responded with an aggressive tweet, saying that he, too, had a nuclear button, and that his was “much bigger and more powerful” than Kim’s. Despite concerns over how the two leaders engage with one another, President Trump is nonetheless endorsing nuclear deterrence as a peacekeeping strategy.

All this escalation has no doubt sharpened the focus of the current E-6B Mercury crews, whose work forms the connecting thread that makes nuclear deterrence possible. But having served alongside them, I know that the work can feel oddly abstract—like doing nothing. Deterrence means having a nuclear capability that prevents other countries from attacking you, and it also means not striking first. You’re preparing for a threat that may never come; conversely, to the other side, you are that very threat. The dissonance is equal parts absurd and terrifying.

In the air or on the ground, we were always prepared for a surprise attack. In each of the continental United States’s three “areas of operation”—geographic regions where a particular combat command has authority—there’s at least one E-6B Mercury, with its 15-man crew, on alert at all times. Those of us at Tinker would take rotating shifts of an average length of 10 hours in the air.

A large part of the job was practicing for the day when we’d have to tell those manning the systems of America’s nuclear-weapons triad—whether they were in missile silos on the ground, in strategic bombers in the air, or in submarines at sea—that it was time to launch a strike. Under normal circumstances, such an order would originate with the president (or whoever was next in line in the chain of nuclear authority), and then be disseminated through the National Military Command Center (NMCC) and on to the triad.

The point of the E-6B Mercury was to stay aloft as backup to maintain a communications link in case the ground-based communications were disabled in an attack. In that case, we would need to be ready to correctly decode and transmit a genuine order to strike. So we stayed in the air and practiced decoding and transmitting training messages.

We relied on two sets of hard plastic-sealed codebooks: one for training exercises and one for a real attack. Each contained codes that allowed us to decipher what is known as an Emergency Action Message (EAM), a string of alphanumeric characters relayed by the NMCC (or, if the center had been destroyed in an enemy strike, from an alternate site). What separated a real EAM from a practice one was its characters, which indicated which codebook to refer to.

In the event of an actual nuclear strike against the United States, the E-6B Mercury would get a real EAM, and quickly identify that it was not an exercise. We would relay it to the triad. If it contained instructions to launch a nuke, the triad would do so without further verification. A real EAM is “enough” for the E-6B to relay it, and it’s “enough” for the triad to launch the nukes.

We performed daily exercises to test our responses to EAMs. Once we received an exercise EAM on our circuits from the NMCC, I had a set amount of time to determine what type of EAM it was—to determine which country or region it was directing us to strike—hand-transcribe the message, then relay it to the nuclear triad. (The recipients got these practice messages as part of their own readiness and training—the characters of the message would indicate they were not genuine.) This was much harder than it sounds: It’s easy to repeat a series of alphanumeric characters, but just as easy to make a mistake. One slip-up and I would be grounded from flying until I re-earned my job qualifications in the flight simulator. Another mistake, and my flying career would be done. There was a limit to how high the stakes were, however: In an exercise scenario, no mistake could lead to the accidental launching of a nuke.

During long flights we remained at the ready for a surprise attack. To stay busy, we studied, engaged in training exercises, and monitored our communications circuits. My crew and I would also informally game out hypothetical nuclear-attack scenarios. Would we relay an actual EAM—one that directed us to order a nuclear strike—if we received one? Most people were quick to answer that they would. I needed time to digest a question that felt like a Cold War scenario, one that wasn’t geared to our current reality.

The strategy of nuclear deterrence we use today is a relic of the post-World War II nuclear-arms race between the U.S. and the Soviet Union. During my time on the E-6B Mercury, the Doomsday Clock, a symbol created by The Bulletin of Atomic Scientists that represents the threat of global nuclear war, was between six and five minutes to midnight. These were relatively unthreatening times that didn’t underestimate the global danger of nuclear weapons, but suggested progress. I rarely considered my potential as the messenger of the nuclear apocalypse. But when I did, my answer was, Yes. I would relay the EAM. That’s still my answer now.

Over time, standing watch required a sustained, debilitating vigilance. Some flights required us to do nothing but remain airborne and log mission traffic for half a day. We would man our communication panels, stuffing our faces with candy (“comm candy,” as we called it) and gulping down cup after cup of coffee to stay alert. I checked and rechecked the EAMs we received. I also made sure to properly dispose of the materials we used to encrypt messages. I had to ensure every action I performed was correct, for the country’s sake, for the squadron’s, and for my own.

My crew alternated between day and night flights. I had trouble sleeping. I would wake up during rest periods between flights, suddenly recalling a typo in one of my post-mission reports, or questioning and then confirming that I had secured every item correctly. When I dreamed, it wasn’t about the act of nuclear war, but about making a mistake in a job whose very purpose is to prevent nuclear war.

When we landed back at base, the pilots prepared the cockpit for a possible emergency takeoff. Our ground-based operational control center monitored warning systems for incoming missiles. A set number of personnel were required to remain on alert at what is known as an alert facility, located adjacent to the runway. A certain number of people were allowed to leave the alert facility for days off, but still had to stay on base. Those allowed to leave carried portable pagers in case the nukes were in-bound and they had to speed across the runway to climb aboard the E-6B Mercury airplane that would be preparing to take off before the missiles struck the United States.

Back on the ground after each mission flight, I sweated the caffeine and sugar out by going for long runs, carrying the pager with me. How long I was able to stay on the ground varied. We rarely flew more than two or three days in a row. Whenever I returned to the alert facility after a break, I checked to see that our plane was still there—that the nukes hadn’t launched.

As we accrued flight hours, we stayed at the ready, but, of course, we couldn’t control whether or not nuclear war actually broke out. This constant readiness came with its own set of challenges, ones not necessarily unique to military personnel. Those who lived through the Cold War had trouble escaping the constant fear of a nuclear apocalypse. Studies of children who grew up during that era have found pervasive feelings of anxiety, fear, cynicism, and apathy.

Although I never made an egregious error with encryption or in deciphering and relaying an EAM, it was always possible that I would—just as it was always possible that the United States would be the target of nukes, requiring us to respond. I worried both about the standards I needed to uphold in order to maintain my status in the squadron, and about the standards I needed to uphold to maintain the status of nuclear deterrence. What if next time it wasn’t an exercise? For the most part, I processed that possibility by not processing it. My crew members and I occasionally talked about the prospect of actual nuclear war. Mostly, though, we shot the shit or shared jokes, steering ourselves away from the apocalypse as a means of self-preservation. Just be cool, man.

In my three years aboard the E-6B Mercury, I logged over 1,900 hours, more than any other officer in the entire E-6B Mercury fleet over that same interval. I was exhausted. Today, the Doomsday Clock sits at two and a half minutes to midnight, the closest it’s put us to the apocalypse since 1953. I wonder how the current E-6B crew members feel about the heightened prospect that they might be called on to relay an actual EAM. They, like me before them, are no doubt just trying to focus on their jobs, or distract themselves. But while we talked about nuclear war as a hypothetical, that hypothetical has now evolved into palpable possibility.

In the war of nuclear deterrence, the best I could hope for was for the status quo to maintain—for the nukes not to launch. Dealing with the prospect of a war with global consequences is something I, like many others, wanted to escape. Yet the challenge for me was an intensified version of the one facing the entire country—not letting the fear of the worst possible outcome overtake our consciousness, striking the right balance between vigilance and not letting vigilance take over our lives. In a nuclear-capable world, we all stand the watch.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.