first published on August 24, 2020 by Funker
An artificial intelligence algorithm “easily” defeated a world class fighter pilot in a simulated dogfight at DARPA’s AI air combat competition on Aug. 20, 2020, according to C4ISRNET.
The AI system, developed by Heron Systems, defeated a carefully selected F-16 pilot in all five rounds at the AlphaDogfight Trials, the finale of the Defense Advanced Research Projects Agency’s Air Combat Evolution program.
It should be noted that the algorithm was only fed “perfect” information, which doesn’t happen in the real world where much of the chaos is mitigated by a well-trained human pilot using initiative and improvisation, “flying by the seat of his pants.”
Additionally, the AI system also made several erroneous maneuvers based on incorrect predictions, and those mistakes were made up for by the algorithm’s incredible targeting accuracy.
Despite the disclaimers, Col. Dan Javorsek, program manager in DARPA’s Strategic Technology Office said, “Artificial intelligence shows a lot of promise.”