I first met Edward Vance in a office on West 45th Street that smelled like stale coffee and old paper. He was sitting behind a desk that was so covered in blueprints I couldn't see the wood beneath them. He looked up when I entered, and the first thing I noticed about him was his hands—long-fingered, stained with ink, trembling slightly, the way a musician's hands tremble before a performance.

"Mr. Delaney," he said. His voice was quiet, precise, the voice of a man who had spent his life talking to drawings instead of people. "Please sit."

I sat. I was thirty-five then, three years out of the Navy, two years into journalism, and still walking with a slight limp from a bullet that had missed my spine by half an inch in the Phu Cat province. I had been assigned to write a profile piece on Edward Vance for The New Yorker—the "greatest living naval architect," the title called him. I was skeptical. Most "greatest" people I had met were just the ones who had survived longest or lied best.

But Edward Vance was neither. He was something else entirely.

Over the next six weeks, I interviewed him fourteen times. Sometimes in his office, sometimes at his apartment in Morningside Heights, sometimes at a bar near Columbia University where he would drink black coffee and talk about hydrodynamics the way other men talked about women. He was brilliant—there was no other word for it. His understanding of naval architecture went beyond engineering into something that bordered on art. He could look at a hull form and tell you, without calculation, how it would behave in a Force Six sea off the Cape.

"Most engineers design ships," he told me once, stirring his coffee with a plastic spoon. "I design consequences."

At the time, I thought this was pretentious. It took me months to understand that it was the most honest thing he had ever said.

Vance was forty-one when I met him, born in Brooklyn to Irish immigrants who had come to America in the years after the famine and ended up working the docks. He had grown up watching ships being built and repaired, and he had developed, as a child, an almost supernatural ability to look at a vessel and understand how it worked. By sixteen, he was working as a draftsman at the Brooklyn Navy Yard. By twenty-two, he had a degree from MIT in ocean engineering. By twenty-eight, he had designed his first submarine.

"The Nautilus class," he said, describing it to me. "Attack submarine. Designed for the Pacific. Silent running, deep diving, long endurance. The Navy loved it."

"How many were built?"

"Twenty-three. All of them served in Vietnam."

I had been in Vietnam. I knew what submarines like the Nautilus did there. I knew about sonar pings in the dark water, about depth charges dropping like stones, about the things that floated to the surface in the morning and the things that didn't.

"And how did that make you feel?" I asked.

He looked at me for a long moment. His eyes were gray, and they were the eyes of a man who had seen exactly what his creations had done and had decided, consciously or not, to see no more.

"It made me feel proud," he said. "And then it made me feel nothing at all. And that, Mr. Delaney, is the most dangerous feeling in the world."

The Nautilus class submarines were workhorses. They patrolled the coast of Vietnam, they hunted supply ships along the Ho Chi Minh trail of the sea, they dropped depth charges on fishing boats and cargo vessels and, on at least one occasion that I would later confirm from declassified Navy records, a civilian ferry carrying refugees from Da Nang to the Con Dao islands. Three hundred and twelve people died on that ferry. The Navy called it a "tragic misidentification." Vance called it a consequence.

After Vietnam, Vance didn't stop designing. If anything, he designed faster. Each new submarine was quieter, deeper-diving, more lethal than the last. The Sealion class. The Thunder class. The Leviathan class. Each one a small step forward in the art of killing people from beneath the waves.

"Why?" I asked him, one night, in the bar near Columbia. "Why keep making weapons that kill more efficiently?"

He swirled his coffee and looked at me with those gray eyes. "Because if my designs make the enemy afraid to attack, then no one attacks. If my designs make the other side think twice before launching a first strike, then we all live a little longer. It's simple deterrence theory."

"Simple," I repeated. "You're sitting here telling me that three hundred and twelve dead refugees are a simple calculation."

"I'm not telling you that anyone died," he said carefully. "I'm telling you that every weapon ever built has killed someone. The question is whether the weapon that exists now kills fewer people overall than the weapon that would exist if I hadn't designed it. And the answer, Mr. Delaney, is yes."

It was a logical answer. It was also, I suspected, a lie he told himself every night so he could sleep.

The turning point came in the autumn of 1963. Vance had been working on a new design for eighteen months—classified, top secret, funded by a budget that I couldn't even begin to understand. He wouldn't show me any of the drawings. He wouldn't even tell me the name of the class. But he would tell me this: "It can operate silently at depths that were thought impossible. It can carry nuclear weapons with a accuracy that makes my hands shake. And it can stay submerged for ninety days."

"Ninety days," I said.

"Ninety days. Which means, in the event of a conflict, it could position itself in any harbor in the world and be ready to fire within minutes of the order being given."

"That's a first-strike weapon."

"It's a deterrent."

"You can't have it both ways, Edward. A weapon that can destroy an enemy fleet in the first hour of a war isn't a deterrent. It's an invitation."

He didn't answer. He just looked at me with those gray eyes, and for the first time, I saw something in them that I hadn't seen before. Fear. Not of me, not of the Navy, not of his own creations. Fear of himself. Fear of the man who sat across from him in a bar on West 45th Street, drinking coffee and designing the end of the world.

After that interview, I stopped going to the bar. Vance stopped returning my calls. And then, in early November, he stopped coming to work altogether.

His colleague at the Navy Yard, a Dr. Sarah Chen, agreed to talk to me. She was a professor at MIT, Chinese-American, sharp-eyed and even sharper-tongued. We met in her office, which was filled with models of ships and submarines and equations scrawled on the whiteboard behind her desk.

"Edward's going through a difficult time," she said, without preamble.

"That's an understatement."

She looked at me over the rim of her glasses. "You're the journalist."

"I am."

"Then you should know that Edward is one of the most brilliant minds I have ever encountered, and also one of the most broken. He designs these machines because he believes in deterrence. He believes that if he can make the weapons good enough, no one will use them. And every time he designs a new submarine, he proves himself wrong."

"Because they get used."

"Because they get used. And because even if they don't get used, the fact that they exist makes their use more likely. Every submarine he designs lowers the threshold for nuclear war, just slightly. And he knows this. He knows it better than anyone. And he keeps designing them anyway."

"Because of deterrence."

"Because he's trapped. He's trapped in his own logic, in his own brilliance, in the terrible clarity of a mind that can see every consequence and feels powerless to stop himself from creating them." She paused. "Edward isn't a monster, Mr. Delaney. He's just too smart to see the exit."

I wrote the article. It was published in December 1963, three weeks before Kennedy was shot, three weeks before the world changed in a way that no one could have predicted. The article was titled "The Man Who Designs the End of the World," and it caused a stir. Not a huge stir—Vance was classified, his work was classified, most readers didn't understand what they were reading. But it was enough. Enough that Vance stopped coming to work for a month. Enough that Dr. Chen transferred to a university in California. Enough that I started drinking more than I already did.

I never saw Edward Vance again after that winter. I heard rumors, as you do. That he had moved to Europe. That he had been recruited by a private defense contractor. That he had retired to a house in Maine and spent his days staring at the ocean. I never verified any of these rumors. Some stories are better left as stories.

Sometimes, when I'm sitting in a bar in Greenwich Village and the whiskey is warm and the jazz is bad and the rain is hitting the window in that particular New York way that makes everything feel both endless and ephemeral, I think about Edward Vance. I think about the man who could look at a hull form and tell you how it would behave in a Force Six sea. I think about the man who designed the end of the world one submarine at a time, believing all the while that he was preventing it.

And I think about the simple, terrible question that he never answered and I never asked: if a man is too smart to be innocent and too honest to be a liar, what kind of person is he?

I don't know the answer. I suspect he didn't either.