75 years ago today, the first nuclear weapon was detonated. As Robert Oppenheimer recalled of that day, “We knew the world would not be the same.” Although we are no longer (technically) in the Cold War, the threat of a nuclear apocalypse is as real as it has ever been since 1945. And yet I feel stupid writing about it—contrary to the evidence, this scenario feels so improbable, something you would only see in a sci-fi movie or a David Lynch creation1.
And it is true that, at least for now, no single bomb is capable of destroying the whole of humanity. But having a weapon that is so powerful naturally produces a system of strategies and incentives in which the detonation (or even a perceived threat) of a single bomb can easily lead to the launch of a hundred rockets equipped with nuclear warheads. What is so challenging to comprehend is that this is most likely to happen by accident. It is difficult to estimate the probabilities of events that have never happened before, but looking at the history of humanity since the Trinity test it seems that we are really really really lucky to have not experienced a nuclear war yet.
To detect any incoming missiles from the Soviet Union, the United States built a large number of radars in various countries. On October 5, 1960, one such radar in Greenland sent an automated warning to the North American Aerospace Defense Command (NORAD) in the US notifying that a huge number of missiles had been detected, suggesting with 99.9% certainty that the USSR is attacking the US . One can only imagine the amount of panic there must have been at the NORAD.
One thing made no sense at all though—at that time, USSR leader Nikita Khrushchev was in New York! Given the low likelihood of such a Shakespearean scenario, there was no immediate military response. It was later found that the radar in Greenland had actually detected the rising of the Moon, and not a swarm of missiles.
On January 24, 1961, an American bomber B-52 broke up while flying over Goldsboro, a city in North Carolina. It was carrying two Mark 39 thermonuclear weapons, each of which was around 270 times more powerful than the bomb dropped over Hiroshima. The plane crashed and the bombs dropped in a field; fortunately, neither of them actually exploded. However, a document acquired under the Freedom of Information Act in 2013 reveals a horrifying reality. Mark 39 bombs were equipped with six interlocking mechanisms, all of which had to be triggered for the bomb to explode. Inspection of one of the bombs found that “five of the six interlocks had been set off by the fall”. Had the bomb exploded, it could have resulted not only in hundreds of thousands of lives lost, but also in a perception that the US was being attacked by a foreign country.
Having witnessed the Hungarian Uprising with his own eyes2, Yuri Andropov, the head of the KGB and later the leader of the Soviet Union, over the years became increasingly paranoid about possible threats to the Soviet Union. George H. W. Bush’s White House top-secret intelligence review3 declassified in 2015 states that “[The President’s Foreign Intelligence Advisory Board (PFIAB)] believe[s] that the Soviets perceived <…> that the chances of the US launching a nuclear first strike [during the first years of Reagan’s presidency] <…> were growing. [The PFIAB] also believes that the US intelligence community did not at the time <…> attach sufficient weight to the possibility that the war scare was real.” For example, in 1981 Andropov, as the head of the KGB, initiated Operation RYaN (РЯН). Its aim was to collect intelligence on a nuclear attack that the Reagan administration might have been planning (as speculated by the Soviets).
Even if unjustified, the fear was real. The tension was at its highest in 1983 when the Soviets shot down a South Korean airliner carrying 269 people, including a US congressman, on September 1. What happened 25 days later is probably the closest we have been to a nuclear war.
At that time, USSR had a protocol where if a notification was received from a warning system that incoming missiles had been detected, an immediate nuclear counter-attack would be launched against the US. A missile strike had been detected by an early-warning system on September 26, 1983 and a notification was sent to Stanislav Petrov, a duty officer who was working near Moscow. His job was simple—to report any such warnings to his superiors. But he did not. Suspecting a false alarm (though not being completely sure about it), he made a decision that he was not supposed to make. The highest ranking officials in the Soviet Union, including its leader Yuri Andropov, were ready to launch a retaliatory strike. But by making the decision to dismiss the warning himself, Stanislav Petrov may have saved the world. And it did not take long to find out that he had made the right call—there were no missiles after all.
These are just a few of many examples of nuclear close calls. The Cold War ended three decades ago but the nuclear weapons are still here—fourteen thousand of them, actually. Whether it is the US and North Korea, India and Pakistan, or a group of radicals with access to a uranium enrichment facility, a nuclear war can erupt very easily. We want to think that the decision to start such a war would be carefully thought through; the history shows us that this is unlikely. As a matter of fact, some nuclear strike authorization protocols make it so that no conscious decision would need to be made at all. A faulty switch or a software bug could initiate a series of events that, like some mad conductor, would move us toward a war that no one wants.
 Schlosser, E. (2013). Command and control: Nuclear weapons, the Damascus accident, and the illusion of safety. Penguin.