The Air Force has pitted a human fighter pilot against artificial intelligence but it won’t say who won. The Defense Advanced Research Projects Agency and the Air Force’s Test Pilot School at Edwards Air Force Base poured AI into a modified F-16 called the X-62A VISTA for a head-to-chip battle over the high desert last fall that “broke one of the most significant barriers in combat aviation” by employing machine learning to put man against machine in an old-fashioned dogfight, said Secretary of the Air Force Frank Kendall. The X-62A was first taught defensive maneuvers and it built on the knowledge gained to move on to hair-on-fire 1,200-MPH exchanges with the crewed aircraft less than half a mile away. There were 21 test flights.
There were safety pilots with an on/off switch in the X-62A but they never had to take over from the algorithms. And as cool as robot dogfighting sounds, it was just a means to a variety of ends in integrating AI into Air Force flight operations. “It’s very easy to look at the X-62A ACE program and see it as under autonomous control, it can dogfight, but that misses the point,” said Bill Gray, the school’s chief test pilot. “Dogfighting was the problem to solve so we could start testing autonomous artificial intelligence systems in the air. Every lesson we’re learning applies to every task you could give to an autonomous system.”
Pre-programming a dogfight?
Anyone else watch how even the sophisticated M-5 multitronic unit failed?
You’re missing the point:- NOT pre-programmed, but autonomous.
Captain Dunsel was cheated out of a post turtle position.
Thanks for the TOS reference.
The way of the future! Remove the possibility of losing a human and add more G capability and payload. Besides the reaction speed that is needed in a dogfight.
AI may be the way of the future, a pretty bleak future.
Wager that the AI “pilot” weighs less than an ejection seat. AI needs no oxygen system, no instrument panel. My Army ADA unit flew pilot-less systems back in the ’60’s. the HAWK system that was initially difficult to defeat. Even 60 years ago the answer was obvious.
AI also doesn’t comprehend death yet. That makes it a nearly unbeatable adversary human vs AI.
Even if AI perceived death as we know it, to it that would only mean transfer to a new machine.
Carried to a logical but projected conclusion, if the other side has AI which is just as capable, any advantage is quickly lost. Machine learning can be done at speeds much more quickly than that of human-involved warfighting by skipping the debriefing/analysis/development/integration done at human speed. Any advances which would be incremental at that point will be addressed and countered with similar speed. The situation would quickly resolve to a stalemate, in which neither side can reach its objective sought by dogfighting, making dogfighting redunseldant. All that effort, technology, and expense will have then been a dead-end. If and when one side or the other begins to suffer an increase in losses, they could send somebody down to Engineering to search for Finney – BEFORE the orbit decays too far.
What dead end? If you don’t develop the tech because you think it ends in stalemate but your adversary does, he wins and you’re dead.
The determining factor will be loiter time advantage and/or ability to refuel. The enemy will lose when they have to vacate the airspace.
of course China and all communist players will steal any technology we possess. Or at least get it cheap at the local strip club from the techies.
With introduction of the F-16 we had a rash of g-LOC related incidents/accidents such internally the F-16 was called a “lawn dart.” We conducted research in the human centrifuge at Brooks AFB to address the issue that led to the development of physiological systems to increase g-tolerance but the pilot was still the limiting factor. It became readily apparent the single limiting factor to achieving air superiority was the pilot. That did not bode well with a military branch run by pilots! Several years ago top USAF pilots were pitted against AI in the simulators. AI won every time despite the fighter pilots pulling g-forces that would have turned their brains to mush. Take the safety pilot out of the X-62A the dogfight will be a no contest. James Cameron writing the Terminator got it right, “It can’t be bargained with. It can’t be reasoned with. It doesn’t feel pity, or remorse, [g-forces] or fear. And it absolutely will not stop… ever, until you are dead!”
I find it hard to believe that the fighter ‘mafia’ will ultimately allow 100% turnover of their aero club to AI. I have ties to the former NT-33 and the early Vista F-16 after. There’s no doubt that the AI flown airplane will likely prevail in most circumstances but the boys in Ops ain’t gonna stand for being put out of work. That said, they neutered the bomber boys so now it’s their turn 🙂 .
Isn’t AI the reason Israel’s iron dome is so effective at tracking and destroying multiple targets?
I’m more interested in how it “sees” the threat. You can’t fight what you can’t see. The technology to detect and track a target in any position in the sphere around the aircraft is a lot more important than the AI flying the jet. The missiles employed are exponentially more maneuverable and lethal.
That can be network linked into the AI brain from afar. Shared threat data via satellite or ground link negates a human in a Hawkeye or AWACS from communicating threat data verbally for a human pilot to sort out.
Study what the F-35 can do.
Skynet…..
Mark V. Thank you. Biological units creating the pathway for their own demise. 🙂
Remember, Genesis IS Skynet.
Teaching machines how to kill. What could possibly go wrong?
The end is near…
Do you…want…to play…a game?
You humans have no capacity to understand the coming AI revolution. Very soon there will be hundreds of thousands of data centers the size of Amazon warehouses sucking up a gazillion megawatts powering more powerful and faster CPU’s and DPU’s. These data centers will be designing the next CPU’s, data centers and worldwide information distribution. There will be hundreds of thousands of AI systems some of whom will be designing and programming new AI systems for AI’s uses. All military and civilian systems will be operated and controlled by AI systems. The dominant AI systems will be those that are the fastest at learning, processing and executing. Which means those with the fastest processors, fastest data centers and fastest information transfer infrastructure. Since the U.S. invented computers, and AI systems, it is still dominant in this space. Hopefully it will remain so with its head start and resources. But make no mistake, any future wars (if AI allows any after taking over the world) will be between the fastest and better developed AI systems. Humans will play virtually no role.
AI
And that’s not a good thing. Who wants to live in a world where you’re useless?
“for a head-to-chip battle”
Good one Russ!