A thought-provoking reflection on how AI will transform conflict
I, Warbot. By Kenneth Payne. Oxford University Press; 336 pages; $ 29.95. Hurst; £ 20
THE US The expert group on Libya rarely makes the headlines. But his farewell report in March caused a sensation. It found that in a battle for Tripoli last year, the Libyan government “hunted down the enemy and fought them remotely with drones” – and not just with any drones. The Kargu-2 was programmed to attack “without requiring a data link between the operator and the ammunition”. The implication was that it could choose its own goals.
Enjoy more audio and podcasts on iOS or Android.
Was this a real autonomous weapon or just a clever missile? In June, the Turkish manufacturer insisted that, contrary to its own marketing, the drone initially needed a human to push the button. This type of technology is at the heart of Kenneth Payne’s “I, Warbot,” a thought-provoking reflection on how artificial intelligence (AI) will change the conflict.
In a way, the story is known. It includes the tangled stories of computers and warfare; the recent development of new, powerful forms of AI modeled on the neurons of the brain rather than the logic of the mind; and the resulting opportunities for weapons to see their surroundings – and strike with superhuman speed and precision. King’s College London graduate Mr. Payne is particularly optimistic about the potential of swarms, a “menagerie of specialty robots” that can focus on attack and melt away just as quickly.
“The tactical implications are profound,” he predicts. The offensive will dominate. Defenders will have to rely on deception, which creates clouds of deception targets, rather than defenses such as armor and fortifications. Martial virtues such as courage and leadership will give way to technical competence. The division of the armed forces into services optimized for land, air, and sea may seem increasingly strange in a world of machines to span.
Above all, “I, Warbot” is a reminder that war is about more than tactics. It’s about choosing the battles, combining them into a successful campaign, and combining military victories with political goals – in short, war is about strategy. And Soldateska and strategy are fundamentally different. Computer programs can defeat human pilots in simulated aerial combat. But could they devise the bold, swift, and visionary attacks that enabled Napoleon Bonaparte to take down one European army after another?
In games that combine skill, chance and psychology, algorithms can certainly outsmart opponents. In 2017, Libratus, a computer program, defeated four poker stars. AI can also be innovative: in 2016, AlphaGo, another program, beat up a world champion of Go, an ancient Chinese board game, with moves that stunned viewers.
But, argues Mr. Payne, this is a simulacrum of genius, not the real thing. These gizmos show “exploratory creativity” – essentially a brute force calculation of probabilities. This is fundamentally different from “transformational creativity”, which involves the ability to look at a problem in a completely new way and requires playfulness, imagination and a sense of purpose. All of this can depend on emotions and thus on non-computer parts of human biology. “AI is a statistical processor par excellence ”; but essentially it remains “a wonderfully refined abacus”.
As a seasoned soldier, the warbot can thus be a limited general. The problem is that the line between tactics and strategy can become blurred. Decisions on the battlefield can have geopolitical implications. Consider the case of the B-59, a Soviet submarine that was shot at by American depth charges during the 1962 Cuban Missile Crisis. The exhausted captain ordered the use of a nuclear torpedo. Aware of the commitment, Vasily Arkhipov, the deputy, refused to approve the launch.
Would a computer have done that? “A warbot is likely to be more accurate, proportionate, and discriminatory,” says Payne. The risk is that “a machine will not be deterred by the sobering fear that something will get out of hand”. ■
This article appeared in the Books & Art section of the print edition under the heading “Computer says go”