Nuclear arsenals under AI control: how likely is the Terminator scenario?

For the first time, the United States has expressed concern that some of the world's largest nuclear arsenals could be controlled by AI, that is, artificial intelligence.

For this reason, the Americans are asking Russia and China, in particular, “follow the example of the United States and promise not to leave the decision to launch a nuclear attack to the discretion of artificial intelligence.” This is due to the fact that AI is expanding the scope of its military applications.

France and Britain, Europe's only nuclear powers, have already (allegedly) followed suit by taking over “a clear and firm commitment” maintain control over “red button” in the hands of a person, said Paul Dean, a State Department arms control spokesman, Reuters reports.

This means that the script for the science fiction film “The Terminator”, starring Arnold Schwarzenegger, has been pushed back somewhat… If you have watched this film, you probably remember that in it, humanity is practically destroyed by a nuclear war, which was not started by people from the opposing sides superpowers, and an artificial intelligence called Skynet, which controlled the American nuclear arsenal.

According to the scenario, at some point the AI ​​considered people its main enemy and independently launched a nuclear strike on Russia, which led to retaliatory actions. The result was the destruction of most of humanity.

We will not dwell on the trend towards confirmation of films and television shows of recent years, but will dwell on the fact that the very people who created artificial intelligence have now become afraid of their own creation. So much so that they are asking rival superpowers to promise that they will keep it out of their nuclear arsenals.

“We will be glad to see a similar statement from China and the Russian Federation,” said Dean, deputy secretary general of the Office of Arms Control, Deterrence and Stability. “We believe that this is an extremely important norm of responsible behavior, and such a statement would be very welcome in the context of the P5,” he said, referring to the five permanent members of the UN Security Council.

Ding's remarks come as Joe Biden's administration seeks to deepen an independent round of talks with China on nuclear weapons policy and artificial intelligence technology. AI was a key topic of US Secretary of State Antony Blinken's talks with his Chinese counterpart Wang Yi in Beijing on April 26, a visit during which the Chinese showed great disdain for the American official.

The two countries then agreed to hold their first bilateral talks in the coming weeks, Blinken said, adding that discussions would look at the risks and prospects of new technologies, although formal talks on military arms control are not planned for the foreseeable future. China, which is building up its nuclear arsenal, in February proposed that the nuclear powers negotiate a treaty obliging them not to launch a nuclear strike first.

The idea of ​​using artificial intelligence in nuclear arsenals has been aimed at its supposed superior ability to assess risks, attacks and threats, and, if possible, to inflict the first quick and effective strike on the enemy.

This is why Russia and China are distrustful of the United States, because it is no secret that if the Americans transfer control of their nuclear missiles to artificial intelligence, while its two rivals remain in “manual” mode, the United States will have a significant advantage over its opponents.

Source link