VSIt is no longer science fiction: a race began around the application of artificial intelligence to nuclear weapons systems – a development that could increase the likelihood of nuclear war. However, with governments around the world working to ensure AI is developed and applied safely, an opportunity exists to mitigate this risk.
But for world leaders to grasp it, they will nevertheless have to admit that the threat is serious… In recent weeks, the G7 agreed to Hiroshima Process Code of Conduct for organizations developing advanced AI systemsin order to “promote safe, secure and reliable AI across the world”.
US President Joe Biden took a presidential decree setting new standards around AI “safety.” The United Kingdom also organized the first ever AI Security Summitwith the aim of ensuring that this technology is developed in a manner “safe and responsible”.
A threat not taken into account
But none of these initiatives measure up to the risks raised by the application of AI to nuclear weapons. The G7 code of conduct and Biden’s executive order only briefly address the need to protect populations from chemical, biological and nuclear threats that could result from AI.
Likewise, although he has explain While a common understanding of the risks posed by AI had emerged at the AI Security Summit, the British Prime Minister, Rishi Sunak, nowhere mentioned the high threat posed by AI applications relative to to nuclear weapons.
No one is unaware of the existential risks posed by the use of nuclear weapons. Their use would cause immeasurable destruction for humanity and the planet. A nuclear war, even a regional one, would lead directly to the deaths of several hundred thousand people, then a significant number of deaths and suffering indirectly.
A disaster averted thanks to man
THE climatic changes which would result alone would threaten several billion people with famine. Nuclear history is full of disasters narrowly avoided. Often, the reason the apocalypse hasn’t happened is because a human being chose to trust their own judgment rather than blindly following the information provided by machines.
You have 70% of this article left to read. The rest is reserved for subscribers.