In recent years, the United States military has worked hard to strengthen its air fleet through drones; But it still relies on human operators to pilot its aircraft. However, it seems that the United States intends to gradually reduce its dependence on human operators. the doctor Will RapperIn an interview with Popular Mechanics magazine, the US Deputy Secretary of Defense recently claimed that artificial intelligence was the first to control a military aircraft and more precisely the US military system.
According to Rapper, in the US military test, artificial intelligence in the flight that took place over California on December 15, 2020, in the role of co-pilot for the Lockheed U-2 spy plane, nicknamed the Dragon Lady. Lady), appeared. ARTUμ version of artificial intelligence μZero is used for games such as chess. In a simulated missile attack on The Bill Air Force Base in California fully controlled the radar AI and determined when to focus on shooting down missiles and when to protect the base. Basically, this is what made AI the leader in operations to defend Bill Air Force Base.
In fact, artificial intelligence did not flourish; But he determined when the human operator should fly. In this maneuver, دستور the AI command was never disobeyed. FedLab U-2 team members spent a month training μZero algorithms so that the AI could steer radar well. In this way, the AI was able to identify enemies and be alert to potential dangers while interacting with the fighter jet pilot. In this maneuver, the connection between the modified version of ARTUμ and other subsystems was severed to minimize the risks of decisions made by this AI.
It should be noted that artificial intelligence is not going to appear on the battlefield soon and take on combat tasks, even on a trial basis. According to official reports, US warplanes using artificial intelligence to engage in aerial combat are not scheduled to fly until July 2021 (July and August 1400). Researchers also say that artificial intelligence-controlled systems such as warfare may be deceived by human genius and are not difficult to deceive. Some time ago, the US Department of Defense began the process of developing ethical guidelines for artificial intelligence to prevent computers from attacking themselves.
To prove the usefulness of artificial intelligence, it is not necessary to take full control of the operation. Dr. Will Rapper says artificial intelligence is used for times when the fighter jet pilot is confused in a complex situation. In this situation, artificial intelligence can help or even take control of some systems. If AI can help pilots focus on the main issues in the mission, the US Air Force could be more successful in military operations.