Disrn
An artificial intelligence program piloting an F-16 Fighting Falcon effortlessly defeated a seasoned U.S. Air Force pilot in five simulated dogfights Thursday, according to multiple sources.
The AI algorithm, developed by Heron Systems, went head-to-head against an operational fighter pilot with over 2,000 hours in an F-16 in the final round of the Defense Advanced Research Projects Agency’s AlphaDogfight competition.
The pilot and Heron’s AI fought in five separate WWII-style dogfights in which both combatants could only use the Fighting Falcon’s guns. The human pilot never scored a single hit due to the AI algorithm’s “superhuman aiming ability.” More
Let’s see what happens in a real dogfight when the human pilot knows he dies if he loses.
I suspect a different outcome than we see in this no risk game play.
Yeah, keep developing skynet. It’s just what we need.
This would be great if pilots were limited to gun fights. Guns are the weapon of last resort. After long range, and medium range missles have been fired, This was more a test of computer speed than the pilot’s ability.
“I suspect a different outcome than we see in this no risk game play.”
You would be wrong. This is akin to thinking you can beat a computer at chess. You can’t. After that was firmly established, gamers said a computer could never beat a human at Go. Too many variables, they said.
Now no human can beat the software at Go.
I don’t know what makes you hold such beliefs. Aerial combat, chess, Go all have boundaries and limitations. Software can be taught these and every single response planned (not determined) for ahead of time. The responses are not hard coded today. The algorithms learn. And they learn so fast to an outside observer it looks like they already knew.
Your statement I quoted would probably work in the eighty’s or maybe early 90’s. Not today.
This completely discounts having a manned fighter vs: an unmanned exact duplicate airframe. All advanced military combat fixed wing aircraft are ‘fly by wire’. The pilot suggests changes to the flight regime, and the flight control systems interpret his (her) wishes to control surface and engine commands. With one hard limit. Don’t kill the delicate human in the cockpit. Which these airframes are completely capable of. A computer piloting a plane (which makes it a drone) or a human operator only telepresent and not in the seat, will thrash the same airframe with the pilot onboard.
The Air Force is fighting tooth and nail to keep the most limiting aspect of aerial force projection in place. The human pilot. Because “Glory”, “Prestige”. They won’t admit it out loud because they equate placing that pilot’s life at risk with some sort of twisted view of ‘duty and honor’. That may be valid if you want a fair fight. Fuck that, I was First Cav. We didn’t believe in fair fights. We believed in winning.
No surprise here…..
Remove life support, pilot, etc. from the aircraft & you gain endurance, speed, payload & the ability to turn at much higher G’s.
Great!
You then have to worry about a 15 year old from Russia aiming the Jet in your direction (re define what the AI calls the “enemy” or the skynet concept.
The Question is, would the AI be able to figure out that the real threat is Hillary & the DNC?
Air Force personnel , if we want to survive, will become mathematicians and engineers, no pilots on the aircraft. We will succeed for a while until rules of engagement are applied by President Hunter Biden when we begin to lose and in response Diversity will be applied and people of Merit will be gratefully cashiered. The last news item before America is no more will be an op ed about how valiant the almost 4,000 airmen fought despite communication problems, hundreds of LGBT pilots arguing with each other, several hundred BLM airmen attacking our own, Democrats insisting that we are the enemy and Trannys beating pilots without a mask…all in the three buildings where flying is conducted.
A few years ago on Jeopardy they had Ken Jennings and Brad Ritter play a computer. The computer won. However, they had to structure the questions so that the computer could handle them. No before and after or several other of the traditional categories Jeopardy uses to spice things up.
Get rid of AI.
Chinese theftnology in … 3… 2… 1
“You then have to worry about a 15 year old from Russia aiming the Jet in your direction…”
We’re a little better with data security on lethal systems. Not as easy as hacking your Twitter feed. We don’t let Microsoft write that part of the OS.
@Lowell
Good!
An exaggeration on my part. I’m actually more concerned about the Chinese Military hacking at an organized military level.
AI playing on it’s home turf, to be expected
All fun and games until an EMP cooks your transistors.
Yeah, Skynet flying all our F-16’s sounds like a GREAT idea.
Said NO ONE. EVER.
The game of Go is much more complicated than chess, and had long been the strategy game computers could not yet defeat. win against, even mid-level skillful human opponents. But that changed four years ago when [Google owned] Deep Mind’s AlphaGo AI program defeated its first champion Go player.
https://techcrunch.com/2017/05/24/alphago-beats-planets-best-human-go-player-ke-jie/ .
But it’s a cautionary tale as companies look to apply the AI technology to real problems in societies, and begin letting the AI programs access large databases of human medical and other records. The claim / dream that AI will produce a better world will more likely become a nightmare.
Do you want Google AI to have control over your healthcare options? I don’t.
“…DeepMind has been mired by controversy. A data-sharing partnership with the UK’s National Health Service, initially heralded as having the potential to optimize medical care to reduce the number of preventable deaths…..Critics have seized on the data transfer of 1.6 million patients’ medical records to the Google-owned company as part of the project. The original arrangement remains under investigation by the UK’s data protection watchdog, the ICO.”
Lowell what do you do?
I ask that for more understanding of your comment.
But my question is this, what if the internet command goes down by disrupting the AI aircraft?
Who wins then?
What Lowell’s trying to say is this.
“Kyle Reese: Listen, and understand. That terminator is out there. It can’t be bargained with. It can’t be reasoned with. It doesn’t feel pity, or remorse, or fear. And it absolutely will not stop, ever, until you are dead.”
Even if it busts an air frame in half.
Reminds me of the old children’s pointless argument over which superhero would win a fight.