AI Pilot Beats Retired Air Force Colonel

[yt1993]

Source: http://www.popsci.com/ai-pilot-beats-air-combat-expert-in-dogfight

 

Quote

A pilot A.I. developed by a doctoral graduate from the University of Cincinnati has shown that it can not only beat other A.I.s, but also a professional fighter pilot with decades of experience. In a series of flight combat simulations, the A.I. successfully evaded retired U.S. Air Force Colonel Gene "Geno" Lee, and shot him down every time. In a statement, Lee called it "the most aggressive, responsive, dynamic and credible A.I. I've seen to date."

And this coming a several weeks after Elon Musk's statements on his potential fears of AI.
If you would think that this is what the US government would actually let us know regarding their AI program, it only makes you wonder what's going on under the veil of secrecy.

This just drives home the point that an AI could actually be better than humans at more than just playing chess or poker.

[BbsMentos]

That's a bit scary I'll admit it

[You_are_a_cunt]

Skynet, thy name is ALPHA

[BingoFishy]

8 hours ago, revsilverspine said:

Skynet, thy name is ALPHA

Pun intended?

[You_are_a_cunt]

1 minute ago, BingoFishy said:

Pun intended?

Didn't even cross my mind until you pointed it out

[RagnarokDel]

To be fair, while a retired colonel has decades of experience, he's probably rusty and doesnt have the reflexes he used to at 30. I wonder how it would fare against an active fighter jet pilot.

[Dabombinable]

Its not an AI...its just an extremely complex program.

[nobelharvards]

AIs can take more G-force than a human.

 

AIs can learn faster than a human (as seen by Google's AlphaGo constantly playing itself).

 

AIs do not suffer from aging as much, and can be easily copied to newer hardware without skipping a beat. An experienced person teaching a newer person will never fully transfer their skills or knowledge, it has to be learnt by the newer person over time.

 

AIs do not have emotions. Human emotions constantly cause irrational decisions.

 

AIs can be run all day every day, forever, with a power supply. Eating takes actual effort on the humans' part. Humans "waste" ~1/3 of their time sleeping.

 

Being human sucks.

 

It should also be pointed out this is all being run by a Raspberry Pi, not a super computer.

[Bittenfleax]

27 minutes ago, Dabombinable said:

Its not an AI...its just an extremely complex program.

It is machine learning though??

 

(Not saying that Machine Learning is AI)

[GER_T4IGA]

Sounds pretty scary, thanks for sharing.

[yt1993]

38 minutes ago, nobelharvards said:

AIs can take more G-force than a human.

 

AIs can learn faster than a human (as seen by Google's AlphaGo constantly playing itself).

 

AIs do not suffer from aging as much, and can be easily copied to newer hardware without skipping a beat. An experienced person teaching a newer person will never fully transfer their skills or knowledge, it has to be learnt by the newer person over time.

 

AIs do not have emotions. Human emotions constantly cause irrational decisions.

 

AIs can be run all day every day, forever, with a power supply. Eating takes actual effort on the humans' part. Humans "waste" ~1/3 of their time sleeping.

 

Being human sucks.

 

It should also be pointed out this is all being run by a Raspberry Pi, not a super computer.

I agree with you entirely,

Except the one fact regarding emotions - there are many studies showing that humans actually having emotions is what lets us make actual decisions in life.
There were studies done with people who suffer from a rare medical issue where the part of the brain that is in charge of emotion is damaged.

http://bigthink.com/experts-corner/decisions-are-emotional-not-logical-the-neuroscience-behind-decision-making
 

Quote

A few years ago, neuroscientist Antonio Damasio made a groundbreaking discovery. He studied people with damage in the part of the brain where emotions are generated. He found that they seemed normal, except that they were not able to feel emotions. But they all had something peculiar in common: they couldn’t make decisions. They could describe what they should be doing in logical terms, yet they found it very difficult to make even simple decisions, such as what to eat. Many decisions have pros and cons on both sides—shall I have the chicken or the turkey? With no rational way to decide, these test subjects were unable to arrive at a decision.

 

So a Data-like robot wouldn't actually be able to compute and make perfect decisions per-say, speaking purely regarding the emotion point - which could probably be fixed using smart programming.

[Deletist_Jerk]

I was hoping I could forget about the movie Stealth once and for all....

 

 

[NumLock21]

I misread the title and thought, why would a pilot literally beat up a retired air force colonel!?

 

[AlTech]

Did anybody expect anything different? COs I didn't

[Bouzoo]

So... almost this

 

[Tech_Dreamer]

AI LMAO!

[vanished]

Wow, very cool.  Of course, it's another killing device so that's not great, but I can't help but admire the tech.

[DXMember]

i guess there's reason why he's retired, eh?

badum-tsss

 

 

))) flying fighter jets is no different than playing chess, there's just more pieces to move and that's what machines are good at

it's still the people we have to fear, they give guns and missions to the AI, it doesn't come up with it's own....

[DXMember]

16 hours ago, yt1993 said:

Except the one fact regarding emotions - there are many studies showing that humans actually having emotions is what lets us make actual decisions in life

well... not exactly... those people still can make decisions like which car you wanna buy - a Volkswagen or a Porsche?

and they'd be like - oh... the Volkswagen has more seats and more kilometers per liter of petrol, the Porsche is more expensive but it's faster, well we have speed limits in the city either way and I have a family - so I'll take the Volkswagen.

 

But they can't however make a decision on the question - would you like a red Volkswagen or a blue one?

because there's no way to rationalize color, in the same way there's no way to rationalize which pair of jeans to wear today

the what to eat part can be rationalized if you start looking at stuff like nutritional value, vitamins - but it take a long ass time to do so especially if you have to do that every morning

 

the emotional part helps us make rapid decisions without much effort which not always are rational - like buying a Porsche in the first example because it's pretty

[Albatross]

NO! Don't with the AIs and robots! I don't want that kind of future, too scurry for me!

[MrDynamicMan]

21 hours ago, nobelharvards said:

AIs can take more G-force than a human.

 

AIs can learn faster than a human (as seen by Google's AlphaGo constantly playing itself).

 

AIs do not suffer from aging as much, and can be easily copied to newer hardware without skipping a beat. An experienced person teaching a newer person will never fully transfer their skills or knowledge, it has to be learnt by the newer person over time.

 

AIs do not have emotions. Human emotions constantly cause irrational decisions.

 

AIs can be run all day every day, forever, with a power supply. Eating takes actual effort on the humans' part. Humans "waste" ~1/3 of their time sleeping.

 

Being human sucks.

 

It should also be pointed out this is all being run by a Raspberry Pi, not a super computer.

I disagree. An AI would be different from a human in that situation because it could take all of the factors into account. The un-emotional human simply doesn't care enough to logically take everything into account, and AI would choose whichever one had the slightest 1% advantage. In a theoretical situation where the pros and cons balance each other exactly ( two identical pieces of chicken, for example) it can RNG and decide.