Our editor-in-chief looks back at the events of 1983, when a man saved the world from a nuclear war that could have been caused by a machine error.
Andrea Tornieri
“Autonomous weapons systems, including the weaponization of artificial intelligence, are a cause for serious ethical concerns. Autonomous weapons systems do not make them morally responsible subjects. The unique human capacity to make moral judgments and ethical decisions is more than a collection of complex algorithms, and that capacity cannot be reduced to programming a machine. A machine, even if it is 'intelligent', remains a machine. For this reason, it is essential to ensure appropriate, meaningful and consistent human oversight of weapons systems.” This is what Pope Francis wrote in his book. message Towards World Peace Day 2024.
What happened 40 years ago should serve as a paradigm whenever we talk about artificial intelligence applied to warfare, weapons and lethal weapons.
This is the story of a Soviet officer, Stanislav Yevgrafovich Petrov, whose unconventional decisions saved the world from a devastating nuclear war.
On the night of September 26, 1983, he was on the night shift in the Serpukhov-15 bunker, monitoring US missile activity. The Cold War was reaching a critical turning point; US President Ronald Reagan had just spent billions on the military and described the Soviet Union as an “evil empire.” Meanwhile, NATO was conducting military exercises simulating nuclear war scenarios.
In the Kremlin, Yuri Andropov recently said the crisis was escalating to “unprecedented levels,” and on September 1, the Soviet Union shot down a Korean Airlines plane over the Kamchatka Peninsula, killing 269 people.
On the night of September 26, Petrov witnessed the Oco computer system, the supposedly foolproof “brain” in monitoring enemy activity, detect the launch of a missile aimed at the Soviet Union from a base in Montana.
Protocol was that the officer should have immediately notified his superiors, who would then give the go-ahead to launch retaliatory missiles at the U.S. But Petrov hesitated, remembering that any potential attack would likely be massive, so he considered the single missile a false alarm.
He made similar observations about the next four missiles that appeared on his monitor a short time later, and wondered why there was no confirmation from ground radar. Although he knew that intercontinental missiles would take less than 30 minutes to reach their destination, he decided not to raise an alarm, surprising other military personnel present.
In reality, the “electronic brain” was wrong: there was no missile attack: Oko was fooled by the phenomenon of refraction of sunlight in contact with high-altitude clouds.
In other words, human intelligence exceeded machine intelligence: the wise decision not to act was made by a human being who could think beyond the data and protocols.
Nuclear disaster was averted, but the incident went unnoticed until the early 1990s. Petrov, who passed away in September 2017, said about that night in the Serpukhov-15 bunker: “What did I do? Nothing special, I just did my job. I was the right person in the right place at the right time.”
He was a man who could assess the potential errors of supposedly infallible machines and, in the Pope's words, make “moral judgments and ethical decisions” because a machine, no matter how “intelligent” it may be, is still a machine.
Pope Francis has repeatedly stated that war is madness and the defeat of humanity. War is a grave violation of human dignity.
It is even more serious to wage war behind algorithms, relying on artificial intelligence to decide who to target and how to attack them, and comforting one's conscience by knowing that the decisions were made by machines. Let us not forget Stanislav Evgrafovich Petrov.
