>>A fighter plane equipped with artificial intelligence could eventually execute tighter turns, take greater risks, and get off better shots than human pilots.

>>A fighter plane equipped with artificial intelligence could eventually execute tighter turns, take greater risks, and get off better shots than human pilots. No. (1) An AI-controlled fighter is still limited by the laws of physics. In 99% of circumstances the pilots ability to "pull G" is not the deciding factor in maneuvering. Pulling 10+ is only possible for a wing, ANY wing, in very specific circumstances of speed an air pressure. Most realworld dogfighting occurs well below human G limitations. (2) Can AI take greater risks? Fighter pilots in combat don't avoid risk out of fear for their own live. That isn't part of their thinking. They avoid risk because the loss of their aircraft will have knock-on impacts to the rest of the operation. Fighter pilots do not protect themselves, they protect a host of other aircraft and people standing behind them. The concept that an AI pilot will press an attack where a human pilot would back off out of fear evidences a lack of understanding. Fighter pilots don't back down for such reasons. Listen to the recordings/transcripts of pilots in dogfights over Vietnam. They are aggressive to a fault, once in an turning fight pressing the attack no matter the likely outcome. (3) What "better shots" will these AI be taking? Missiles are not arrows. They are autonomous vehicles. Whether an AI or a human pilot launches them is of little consequence to a sidewinder. There is a place (cranking) where an AI pilot might be able to better perform the maneuver and better-guide a missile, but that isn't AI rather than a basic autopilot-type assist (holding target at a particular angle). As for gun shots, AI is no needed to calculate a trajectory and fire the cannot at the appropriate moment to most likely intercept the target. That is well within non-AI tech.