July 3, 2021

  I’m a warbot. By Kenneth Payne. Oxford University Press; 336 pages; $ 29.95. Hurst; £ 20

  THE UN’S Libyan expert panels rarely grab headlines. However, the report of the graduate president in March caused a turmoil. In a fight around Tripoli last year, the Libyan government said it had “hunted down and remotely controlled” enemies with drones, not just drones. Kargu-2 was programmed to attack “without the need for a data connection between the operator and the munitions”. What that meant was that you could choose your own target.

  Was this a true autonomous weapon, or was it just a clever missile? In June, Turkish manufacturers insisted that drones now require humans to push a button, as opposed to their marketing. This kind of technology is at the heart of Kenneth Payne’s “I, Warbot.” This is artificial intelligence (AI) The conflict changes.

  In a sense, the story is familiar. This includes the intertwined history of computing and war.Recent evolution of the new powerful form of AI It is modeled on neurons in the brain, not on mental logic. It can then see what the weapon is around and attack with superhuman speed and accuracy. King’s College London scholar Payne is particularly bullish on the possibility of a swarm of “specialized robot menagerie” who focus on attacks and quickly melt.

  “The tactical implications are deep,” he predicts. The attack becomes dominant. Defenders must rely on deception to generate clouds of decoy targets, rather than protection of armor, fortresses, etc. Martial arts such as courage and leadership are replaced by technical abilities. Dividing the army into land, air, and sea optimized services may seem more and more strange in the world of machinery that can straddle them.

  But above all, “I, Warbot” reminds us that war is more than a tactic. It’s about choosing fighting battles, how to connect them to successful campaigns, and how to connect military victories to political objectives. In short, war is about strategy. And soldiers and strategies are fundamentally different. Computer programs can already defeat human pilots with simulated dogfights. But can they come up with a bold, swift and visionary attack in which Napoleon Bonaparte knocks out European troops one after another?

  Algorithms are a game that combines skills, chances and psychology, and you can certainly betray your opponents. In 2017, the computer program Libratus saw off four poker stars. AI You can also innovate. In 2016, another program, AlphaGo, beat the world champion of the ancient Chinese board game Go with a move that surprised spectators.

  However, Payne argues that this is a genius imitation, not the real thing. These gizmos represent “exploratory creativity,” that is, essentially a brute force calculation of probabilities. It is fundamentally different from “transformative creativity,” which requires the ability to think about problems in a whole new way, playfulness, imagination, and a sense of meaning. It can all be emotionally dependent and therefore dependent on parts of human biology that are different from computers. “”AI It’s an excellent statistical processor. “; But in essence, it remains a” wonderfully sophisticated abacus. “

  Therefore, a skilled soldier, Warbot, can be a limited general. The problem is that the line between tactics and strategy can be blurred. Battlefield decisions can have geopolitical implications. Consider the case of the Soviet submarine B-59, which was hit by an American depth charge during the 1962 Cuban Missile Crisis. Being aware of the stake, Deputy Commander Vasili Arkhipov refused to approve the launch.

  Did the computer do that? “Warbots are more accurate, balanced, and more likely to be discriminating than humans,” Payne said. The risk is that “the machine is not disturbed by the calm fear that things get out of hand.” ■■

  This article was published in the printed Books & arts section under the heading “Computersays go”.