The US has created and soon plans to unveil a new doctrine for modernising its armed forces which will include, among other things, the use of artificial intelligence (AI) and quantum computing. The military predicts, that roughly a third of the world’s most modern armed forces will be robotic within the next 10-15 years. The head of the US Joint Chiefs of Staff, Mark Milley, said during a press conference – “We’re not doing this to wage war, we’re doing it to deter war”.
According to him, the main task of artificial intelligence is not to take away the work of military analysts, but to make warfare less deadly. Currently, the US Department of Defense is already experimenting with bots with artificial intelligence, that can fly a modified F-16 fighter jet. And in the near future, robotics will play an increasing role in drones, naval vessels and ground vehicles. General Milley also spoke about future unmanned air forces, a navy “without sailors” and crewles
Decision-making time will also be reduced, to a few minutes instead of hours or days. And here scientists are already warning against relying too heavily on artificial intelligence for command and control of weapons, as it operates at a much higher speed than humans. “Humans should never believe, that machines think,” the experts insist. For example, a report, published by the Arms Control Association, said artificial intelligence and other new technologies such as hypersonic missiles could lead to a “blurring of the distinction between conventional and nuclear attack”.
Mark Milley stressed, that the fundamental point for the US is that people, not computers, must make life or death decisions in time of war. The country’s policy is to ensure, that people remain in the decision-making cycle.
Another concern for opponents of AI is that the technology could allow fraudsters or even terrorists to obtain classified information on building dirty bombs or other deadly devices. Much of the data today is held by private companies, which could be vulnerable to espionage driven by the same artificial intelligence. Declassification could also affect the detection of nuclear weapon sites, a risk that is too high.
While experts and government members debate the appropriateness of using artificial intelligence in military decision-making, the technology has already demonstrated its usefulness, particularly as a tool for predicting a Russian invasion of Ukraine. Artificial intelligence-based models have given a “particularly dire” and clear prediction, that Russia would start a war. This was back, when the possibility of a Russian invasion was still being discussed in Washington and other Western capitals as theoretical.
As for the Ukrainian counter-offensive, on which computer calculations have also been made, Mark Milley said, he was not surprised, that it was slower. Asked whether he was worried that the Ukrainian counter-offensive was not showing sufficient progress, Millet said: “A war on paper and a real war is different. In a real war, real people die”.
Artificial intelligence now predicts, for example, the geography of likely missile attacks on Ukrainian cities. But intelligence and other tools traditionally used by the military continue to play a major role.
And Lieutenant General Jack Shanahan, former director of the Pentagon’s Joint Center for Artificial Intelligence, assured, that the US Department of Defense, while interested in AI capabilities, would not be handing over command and control to machines. The main task for AI, he said, remains to improve the efficiency of technology to save lives and help prevent war.
At the same time, the question is whether other countries will abide by the same rules as Washington. “I don’t believe the US will go down the path of resolving things without human control. But I’m not sure, that anyone else can fail to do so. I think, the biggest concern would be giving this machine or facility too much authority,” Shanahan said.
