- An artificial intelligence program developed by Heron Systems went head-to-head against an ongoing Air Force F-16 pilot in a simulated dogfight Thursday.
- Heron AI achieved a flawless victory with five straight wins. The human pilot never scored a single hit, according to multiple reports.
- The “WWII-style” dogfight was part of DARPA’s AlphaDogfight Competition, designed to bolster the agency’s striving to build trust in AI and develop manned unmanned team capabilities.
- Visit the Business Insider website for more stories.
An AI algorithm piloted an F-16 Fighting Falcon in a simulated dogfight against an expected U.S. Air Force pilot on Thursday achieved an error-free victory with five straight wins in a fierce competition, according to multiple reports.
The Defense Advanced Research Projects Agency (DARPA) held the final round of its third and final AlphaDogfight competition Thursday, putting an AI system designed by Heron Systems against a human pilot in a “simulated inside-visual range air camp” situation.
Heron’s AI went head-to-head with a graduate of the Air Force’s weapons instructor course with the callign “Banger,” according to Breaking Defense. An expert commentator, Justin Mock, from DARPA, said that the AI algorithm showed “overwhelming efficiency” during dogfighting. During the fight, the human pilot never scored a single hit.
Breaking Defense characterized the simulated involvement as a “one-on-one combat scenario” in which fighters “shot forward guns in a classic, WWII-style DogII style,” indicating that although artificial intelligence progressed rapidly, more work was needed before ready for the complexities of modern warfare.
Ben Bell, Heron Senior Machine Learning Engineer, told Sandboxx News that the simulated environment put the AI fighter at an advantage over his human enemy. That said, the virtual environment allowed the human pilot to also perform maneuvers of high G-fights that might not be possible for a human pilot in the physical world.
However, the performance is significant for an AI system for which development, according to Breaking Defense, began less than a year ago. Bell said Heron’s AI program learned quickly, gaining roughly “12 years of experience” over the course of 4 billion simulations, Defense One reported.
Heron’s system competed against the human pilot after defeating other AI algorithms in earlier rounds of the competition.
Last year, DARPA declared that although AI can beat humans in games such as chess, there is currently no AI in existence that “can contaminate humans in a high-speed, high-G-dogfight combat jet.” The agency seeks to change that and advance AI for future manned-unmanned teaming.
The AlphaDogfight is focused on moving DARPA’s Air Combat Evolution (ACE) program forward.
The ACE program, according to DARPA, is designed to “provide a capability for a pilot to participate in a broader, more global air command mission, while its aircraft and crewed unmanned systems engage in individual tactics.”
Commentary on the ambitious goals of DARPA’s ACE program, Col. Dan Javorsek, the program manager at DARPA’s Strategic Technology Office, said last year that the agency ‘envisions a future’ in which AI handles split-second maneuvers during in-visual range dogfights, keeping pilots safer and more effective because they carry large numbers of unmanned aerial vehicles. systems orchestrate into a web of overwhelming combat effects. “
Given the Department of Defense’s growing interest in AI, DARPA is not the only one pursuing these types of opportunities.
Speaking at a Mitchell Institute event in early June, Air Force General Jack Shanahan, then head of the Joint Artificial Intelligence Center, said the air force was looking to deploy an AI-powered autonomous drone against a manned fighter jet next year in a dogfight.
It is unclear what the air-to-air battle will look like next year, assuming it happens. It is also unclear exactly what the AI-powered autonomous drone might look like.
