Last week prominent figures from the worlds of politics and technology gathered at Bletchley Park to discuss artificial intelligence (AI). AI is bringing far-ranging change to many aspects of our lives, but I have some doubt that its capabilities are as advanced as many think. As many comedians have quipped, bots cannot even put a tick in an online box that says ‘I’m not a robot.’
My wager is that the human factor will remain a central component of our world despite the advent of AI. But I am certainly no expert in this area. My thinking turns to the ways the atomic age unfolded and how, for all the destructive potential of the nuclear bomb, human traits have oftener than not been decisive in outcomes, for better or worse.
Forty years ago this week, as NATO conducted an annual military exercise, the world came close to nuclear war. The ‘Able Archer 83’ drill, which began on 7 November 1983, is instructive in the interplay of human decisions and the dystopian potential of science and technology.
The early 1980s: The cold war ratchets up
To understand the gravity of the events of November 1983, we need to grasp how tensions between the Americans and the Soviets had been rising during the preceding years. The détente of the 1970s had ended with the Soviet Union invading Afghanistan and the United States halting the ratification process of a key nuclear arms control treaty (SALT II) in response.
At the end of 1982, the elderly and ailing Yury Andropov became General Secretary of the Communist Part of the Soviet Union. Seriously ill when he took the top job, Andropov's previous position as the head of the Committee for State Security (the KGB) had cultivated a deep paranoia towards the West. From his hospital bed one cannot imagine he was well placed to be dealing with nuclear policy.
US policy, meanwhile, only fed Soviet paranoia. In March 1983, President Ronald Reagan announced his Strategic Defence Initiative (SDI, popularly known as the Star Wars programme) which sought to give the US the upper hand in any nuclear exchange. The SDI undermined the ‘mutually assured destruction’ that otherwise kept a precarious stability between the two superpowers. (Reagan, like Andropov, was no spring chicken. The following year, asked whether his age was an obstacle to being re-elected president, his riposte was: ‘I will not make age an issue in this campaign. I am not going to exploit my opponent’s youth and inexperience.’)
The Americans, for other reasons, were keeping their ears pricked in 1983. On 1 September the Soviet air force shot down Korean Airlines Flight 007, mistaking it for a US surveillance aircraft, killing all 269 people on board. Moreover, the Americans were on high alert following the killing of more than two hundred US military personnel in a terrorist attack on a base in Beirut. The US military, naturally, increased its combat readiness – which only unnerved the Soviet Union. A classic escalatory spiral seems to have been in motion.
The September Man: Stanislav Petrov
In the midst of those spiralling events, on 26 September, the Soviet Union’s early warning systems detected a launch of five intercontinental missiles by the United States. The system had malfunctioned, but no one could know that. Had protocol been followed, the Soviet Union could well have launched a missile in retaliation; the timeframe within which such decisions need to be taken is frightening short.
The only reason the detection of the missile launch on that September day did not get relayed to superiors was one man’s willingness to disbelieve the technology in front of him. ‘The siren howled,’ recalled the man sitting on duty in the secret Soviet bunker, ‘But I just sat there for a few seconds, staring at the big, back-lit, red screen with the word “Launch” on it.’ Stanislav Petrov disobeyed his orders. He should have raised the alarm.
In 2015 Petrov agreed to be filmed as part of a feature length documentary, The Man who Saved the World. The documentary followed Petrov, now an elderly man, as he travelled to the US. It turns out that the individual who might well have saved millions of lives was now a cantankerous and embittered old chap. Oh well.
The November Man: Rainer Rupp
There is much myth making in dubbing someone ‘the person who saved the world’. The label has also been applied to a Soviet naval officer who prevented a nuclear torpedo from being launched during the 1962 Cuban missile crisis, but perhaps the most deserving of the title came later in 1983 as NATO command posts in western Europe carried out their Able Archer exercise. Their scenario simulated events leading up to a nuclear strike.
The Soviet Union, as already described, was marked by deep paranoia. With little knowledge of the NATO exercise, the Soviet side interpreted it as a cover story for an actual nuclear first strike. From intercepted Soviet communications, US officials at the time knew that the Soviets proceeded to load nuclear weapons onto combat aircraft in eastern Europe and had issued orders that personnel be prepared for ‘immediate use’ of nuclear weapons. Moscow deployed its nuclear-armed submarines to launch positions and notified its diplomats round the world to be prepared for a nuclear attack on the Soviet Union.
In NATO and the US, for whom Able Archer was a large-scale but nonetheless routine exercise, the Soviet moves and the communications they were intercepting caused alarm. The Soviet Union seemed to be gearing up to use nuclear weapons. Standard procedure would have been to relay this information to senior officers and the US would likely have responded in kind by preparing its own nuclear forces for immediate action – this did not happen because one American officer interpreted the events differently from his colleagues and intervened.
The watchful suspicion of the cold war is scary to think about. Neither side understood the other’s actions as a reaction to its own activities.
The Soviet Union did obtain some accurate information. An East Germany spy, Rainer Rupp, had penetrated NATO where he had secured a job in the alliance’s headquarters. According to the historian Taylor Downing, the East Germans sought urgent information from Rupp who reported back that it seemed like an ordinary exercise from his vantage point inside NATO. Downing writes: ‘It’s quite possible that the spy… could well have saved the world.’
The human factor
The films Dr Strangelove and Fail-Safe (both released in 1964) grimly imagine how badly wrong nuclear policies can go. The automatic, irretrievable nuclear launch imagined in the former film is hardly fantasy; the Soviet Union’s ‘dead hand’ system was designed to operate autonomously and long pre-dates our current discussions about artificial intelligence. It is reportedly still operational.
Perhaps the most salutary film to watch, mind, is War Games (1983), which shows how ‘innocent’ the actions that set us on course towards nuclear war can be. It is another case of autonomous machines reacting to non-existent nuclear strikes; machines tilting at their windmills. AI takes things to a new level: the ability of a machine to reason and learn beyond our own capabilities is indeed frightening, as is the pace of developments, but many possible futures could unfold from this point on; some good, some bad.
Time and again, as some of the history above shows, human choices have prevented disaster despite the existence of autonomous machines. It is fortune, nonetheless, that has prevented a bad decision. The ability to cause massive destruction is ever present.
Next time you see Joe Biden or Vladimir Putin on a television screen, remind yourself that somewhere nearby is a man – because it almost certainly will be a man – with a briefcase which contains the authentication codes needed to launch a nuclear weapon. The old men are never very far from their bombs.
If you would like to receive notifications when new content is added to the blog, please click the 'Sign Up' button in the top right.
Cover image: A nuclear test site in Nevada. Public domain image; rendered black and white by me.
Comentários