http://www.popsci.com/ai-pilot-beats-air-combat-expert-in-dogfight
>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. Lee called it "the most aggressive, responsive, dynamic and credible A.I. I've seen to date."
>And "Geno" is no slouch. He's a former Air Force Battle Manager and adversary tactics instructor. He's controlled or flown in thousands of air-to-air intercepts as mission commander or pilot. In short, the guy knows what he's doing. Plus he's been fighting A.I. opponents in flight simulators for decades.
>The A.I., dubbed ALPHA, was developed by Psibernetix, a company founded by University of Cincinnati doctoral graduate Nick Ernest, in collaboration with the Air Force Research Laboratory. According to the developers, ALPHA was specifically designed for research purposes in simulated air-combat missions.
>The secret to ALPHA's superhuman flying skills is a decision-making system called a genetic fuzzy tree, a subtype of fuzzy logic algorithms. The system approaches complex problems much like a human would, says Ernest, breaking the larger task into smaller subtasks, which include high-level tactics, firing, evasion, and defensiveness. By considering only the most relevant variables, it can make complex decisions with extreme speed. As a result, the A.I. can calculate the best maneuvers in a complex, dynamic environment, over 250 times faster than its human opponent can blink.
When did you realize that Russians in the air will become obsolete by 2050? Future airspace will be dominated by AI making millions of decisions per seconds that no human fighter can withstand.
![KeenMatureHoopoe[1].webm](https://i.imgur.com/Welkdmm.jpg "KeenMatureHoopoe[1].webm")
>>30429621
>AI Controlled anti-air missiles
>>30429708
No escape
OP again. I hope USA survives until 2050. If we ever get AI to work on fighters, no other country will ever catch up to the USAF. AI will be cheaper than training a human in the long run. You can train an AI and copy it. It's kinda scary to fight an air force that has no downtime and is instantly replenishable.
We're dead
>>30429708
Swarms of them. A zerg rush, if you will.
When this tier of AI technology is applicable for infantry and policing purposes, freedom is over.
>>30429778
Welcome to the dystopia that is our future.
>>30429780
Young men today will be the last generation to know what it meant to be free. What a thought.
>>30429708
>AI controlled cruise missiles
>>30429784
The State watches over you anon. The State makes sure you are fed and safe. Why would you go against the State? Are you one of those terrorists? Do you want to go back to the times when criminals roamed the street, killing raping as they wished? It's the current year, anon.
>>30429778
AI can far more easily deal with aerial combat, I'm honest.
Suits it far better.
>>30429778
Policing is literally the most difficult job an AI could try to do.
>>30429730
>It's kinda scary to fight an air force that has no downtime and is instantly replenishable
Having an AI control a plane doesn't change the fact that you need to both make the plane and maintain it.
>>30429730
Barack Obama will be the last president of the US: Baba Vanga had predicted that the 44th US president would be an African American, but she had also added that he would be the last one. According to her, he would leave office at a time when the country would be in economic ruins, and there would be a huge divide between the northern and southern states – as was the case during the American Civil War.
Too bad for you
>>30429897
I'm talking about human fatigue
>>30429621
This is old news to anyone in the coding field.
>>30429778
Born too late to explore the New World. Born too early to explore the galaxy. Born just in time for the Robot-Human war.
Honestly, who is surprised by this?
30 years from now, there will be little to nothing left that humans can do better than AI.
We better prepare.
>>30429621
We culture now
This is now a future ship names thread
>GSV So Much For Subtelty
>>30429897
A human pilot takes 20 years to be born, raised and trained. An AI would have no such limits.
>>30429621
>mfw imagining drone fighters pulling 20g maneuvers
Isn't this the plot of some anime?
>>30430240
Airframe limits under load are generally speaking under 9g.
Be a while before we have even the material science to create airframes that are capable of anywhere near that.
>>30430251
WHat? Pilots limits define the G limit for aircraft, not metallurgy / engineering / material science.
who's got the anime GIF of pilot vs AI plane
>>30429621
I don't want AI to replace pilots simply because flying is really fucking cool
And if we remove human interaction from things like this, people will lose purpose.
Smash the machine
>>30430162
>GCU Let Me Offa' This Train
>>30430260
Nope; why do you think Super Hornets are only rated for 7.5G while F-16s are rated for 9G?
Pilots play into what jets are designed to achieve, but the limits are primarily set by the lifespan requirements of the jet and how their structures handle repeated stress. Most jets are capable of pulling roughly 1.5x their rated loads before having bulkheads shear, but they don't set the ratings that high because they need their jets to last 4 or 6 or 8 or 10 thousand flight hours. If they could make the jets stronger, they would and they'd advertise it as a ~50,000FH or very low maintenance jet.
Missiles are rated to pull 20-50 lateral Gs, but that's because they're a single-use and short-endurance item, so it's irrelevant of a concern if they don't last a single flight hour of 40G manoeuvring.
>>30430260
>g forces dont affect the airframe
while i think youre both right, youre both idiots
>>30429897
two probleams easily solved by automated factories and human slaves
>>30429897
You just won't need
>life support
>instrument panels
>sticks or HOTOL
>ejector seat
>canopy
>mirrors
It does change quite a lot.
>>30430240
>pulling 20g maneuvers
A missile can pull 30 g. At this rate a human pilot would be split evenly between his left boot and his right boot.
>>30429897
>Having an AI control a plane doesn't change the fact that you need to both make the plane and maintain it.
But you don't have to pilot it, which is the much more difficult task.
In fact, you could eventually automate the entire process - manufacturing, maintaining, and piloting.
>>30430260
kek, not even close to being true, anon.
You do a 9g turn in an F-16 with bags and you're losing both wings.
>>30430345
>At this rate a human pilot would be split evenly between his left boot and his right boot.
Depends on time sustained and in what direction.
Humans are capable of surviving (Or rather, have been recorded as surviving) 214g.
>>30430298
This.
The best pilots can do up to around 13-14g, redbull air race used to be safety limited to 12g but these days is at 10g.
Doing some aerobatics without a pressure suit I passed out once when pulling 6 g going into a loop. Fun times
An AI would also have the advantage of direct control of fly by wire instead of having to go through an imprecise stick.
>>30430345
Life support and cockpit equipment don't actually add up to all that much in terms of weight; on the F-35 it'd be something like 2000lb on a 50,000lb warplane. Some of those weight losses would also be regained in the addition of extra computers, redundancy comms and cooling systems (boxes of silicon, copper and steel aren't exactly light).
Also you're over exaggerating what Gs do to people; the pilot would be useless at 30Gs, but only because of the blood drain. Back in the 1960s they had volunteers jump on rocket sleds and get accelerated at 40Gs; they were delirious at the end of the ride, but the only real medical issue was that capillaries in their eyes burst; after 3 days rest they were good as new though. Those sleds also accelerated them horizontally in a seated position, so blood from their legs went to their chest / head; in a fighter pulling 40Gs, the blood goes the other way.
>>30430380
I'm purely interested, what were the circumstances for the 214G?
>>30430398
https://en.wikipedia.org/wiki/Kenny_Br%C3%A4ck
>>30430401
https://www.youtube.com/watch?v=Hy8fgGiI1WA
damn; after hitting the fence the car does 360 in like one frame
>>30430360
Okay idiot, that's why you drop the tanks if something engages you.
>>30430383
Oh shit. I didn't know we had flying ace Chuck Yeager in here. Tell us some old war stories.
It doesn't mention the aircraft simulated?
I wonder how fast it would be over if the AI plane were free from the g limits and safety margins required for a human pilot.
>>30430298
Jesus christ so much bullshit.
>>30430298
Different design specs.
Why would anyone build a 20G plane for a 10G pilot? Sure, it would last a long time, but it'd be slow and use too much fuel 'cause it's be heavier than it needs to be.
>Missiles are rated to pull 20-50 lateral Gs, but that's because they're a single-use and short-endurance item, so it's irrelevant of a concern if they don't last a single flight hour of 40G manoeuvring.
Thank you for proving my point that airframe G ratings are defined by the pilot, not the equipment.
>>30430299
Nobody squeezed your head.
>>30429912
She also said Europe would be gone by 2016 but those fuckers are still here
>>30431237
Meanwhile an F-35A doesn't need to drop anything.
>>30431747
If you can get that hangar queen off the ground.
>>30431912
>Constantly doing test flights
>88/88 sorties with only two routine maintenance issues in recent exercise
>Has done all-weather 24/7 carrier ops
How's 2005?
>>30429778
I worry about this as well. Butlerian Jihad when?
>>30431310
>She also said Europe would be gone by 2016 but those fuckers are still here
>BRexit
2016 isn't over yet, chum.
>>30431943
How late? How many gazillion dorrars over-budget? How's flight hours per maintenance hour? or per dollar? Or sortie availability?
![xvqptxkz6p91q0evgws1[1].jpg](https://i.imgur.com/bSZNFuj.jpg "xvqptxkz6p91q0evgws1[1].jpg")
>>30431977
>>30430287
>people will lose purpose
That is the reason why there is little progress here.
Many years ago NASA had a drone that could pull 20 g sustained. The "white scarf brigade" shot down the project.
Same with a drone equipped with air to air missiles.
>>30430398
>Life support and cockpit equipment don't actually add up to all that much in terms of weight
There is also the volume needed and the position to allow the pilot to actually see things. Modern drones can therefore have far flatter structures.
>>30430240
>Isn't this the plot of some anime?
Maybe you're thinking of Macross Plus? Fuck ghosts.
>Russian ai flies close to USA in international waters.
>American ai takes off and immediately lands
>Russian and USA ai began processing as soon as USAi took off, had a theoretical battle faster than a human blink, and they both left without firing a shot after deciding who would win
>war is now obsolete as ai combatants have already decided the winners
>yfw
>>30432045
That's deep, man. AI becomes so advanced that they both quit.
>>30432018
https://www.youtube.com/watch?v=5hJepWBUqZk
>>30432104
or they immediately start fighting us because thats the easiest way to settle the dispute
>>30432045
>implying we would design self preservation/surrender into the Ai
>implying the Ai would open a line of communication to the enemy to be hacked
>>30432104
>war ends because AI systems are tired of humans fighting over stupid shit
>>30431977
Fuck off, Lind
![ghost-sms[1].jpg](https://i.imgur.com/ZDLQFGQ.jpg "ghost-sms[1].jpg")
>>30432018
They did actually turn those into Leashed AI units later on. The RVF-25 has three AIF-7S units as its combat support units, and can take the leash off if the scenario requires full performance, while Macross Galaxy mass deployed AIF-9Vs against the Frontier fleet.
For a board that is mostly afraid of their freedom/rights being taken away, I am actually quite surprised that everyone here seems really happy that non-human weapons development that does not respect any of the Bill of Rights is going really great.
You could literally control an entire army with one console/workstation if this keeps on happening and the 2nd Amendment literally ends up becoming obsolete.
At least with armies, you can expect mass desertions. With AI, not so much.
>>30432178
>the 2nd Amendment literally ends up becoming obsolete.
Unless you're killing gangbangers and rapists with those drone armies, no it doesn't.
>>30432191
>not having super advanced drone policemen that never get tired at all patrolling 24/7
Like I said, this will turn the 2nd Amendment completely obsolete. It's like people don't even think ahead anymore.
Oh it's so cool, breh is the only thing I get in this thread. Besides, AI code is just that, code. The Chinese and Russians probably already have it.
>>30432211
>Oh it's so cool, breh is the only thing I get in this thread.
>>30431961
>>30429778
>>30429780
>>30429784
>>30429804
>>30430324
Arguably >>30430162
>>30432211
Unless they're in your home 24/7, they will be about as effective as regular police. In the sense that they investigate crimes after they've been committed, not stop them.
>>30432045
>>30432104
>>30432129
>>30432144
Reminds me of this.
>>30432170
Sounds a lot like the mini-planes the VF25 had in Macross Frontier. Almost like Fin Funnels, except no need for high-speed human calculations or psychic powers.
![rvf-25-fighter[1].jpg](https://i.imgur.com/VTCZN97.jpg "rvf-25-fighter[1].jpg")
>>30432253
That is from Macross Frontier. The RVF-25 is the Mini-AWACS variant.
>>30432245
They don't need to be in your home, 24/7. They will literally just profile everyone walking on the street from above with multiple sensory cameras and then as they watch break-ins occurring, send in the cavalry.
They will be far more effective than police, because it's like having mobile London street cameras that can actually think in the air. Again, this makes the 2A absolutely obsolete.
>>30432274
I agree, and this is why we had best get to steppin' on technological solutions. (In the meantime, a good reason to oppose drone registration and support good crypto.).
>>30432283
Like the guy from before said, once this technology is put into military and police drones, human freedom will be put at risk.
>>30432178
>2nd Amendment literally ends up becoming obsolete
it isn't already? the constitution is a security blanket. it's there to make you feel safe, but doesn't actually protect you from anything. if the Government wanted to take away your, as an individuals, guns away they could. taking away everyone's firearms isn't something that needs to be done to protect itself, but it does appease a very vocal majority of the population.
'mass desertions' is just another safety blanket theory. were there mass resignations after a compound gets burned to the ground or someones pet gets shot? nope. you'll always have people eager to be the jack boot. at least with AI it won't be malicious for personal pleasure... hopefully
>>30432302
When I said "we", I didn't mean "military and police."
If the only thing that can beat a bad guy with a drone is a good guy with a drone....
>>30432247
we'll probably have to keep a human in or attached to the cockpit somehow just make sure it doesn't get smart enough to create world peace
>>30432270
Macross Frontier was great until that movie of it came out.
So what is the outcome of this going to be?
The Matrix, Terminator, Starsiege?
Or, Macross, Star Trek, and Star Wars?
>>30432344
We're the movies just series recaps?
>>30432328
I somehow doubt they'll do something like that in programming the planes.
>>30432247
always creeps me out
sadly it was a fake afik
>>30432406
Recaps that retconned a lot of things in the series. I still believe to this day that the TV series is much better than the movie.
>>30432361
Battlestar Galactica
https://www.youtube.com/watch?v=YPfqODwT9C0
>1997+19
>putting AI in a weapon
then again it's not like we had not been warned, so I guess we deserve this.
>>30430260
>WHat? Pilots limits define the G limit for aircraft, not metallurgy / engineering / material science.
Oh really?
Watch what happens when somebody tries to pull a multi-G turn in this B-52.
>>30432474
That didn't have anything to do with the amount of Gs IIRC.
>>30429621
GHOST
H
O
S
T
https://www.youtube.com/watch?v=5hJepWBUqZk
>>30432473
Scientists can be dumb a lot of times.
>>30432474
>blaming it on the G's.
That's what happens when a shitty pilot does a hardcore maneuver on a big slow plane at slow speed at low alttitude.
If i try to do a barrel roll at 50 meters with a chopper i would also die.
>>30432247
This is not far from the truth, btw.
Watch this from 15:11 till end
https://youtu.be/xOCurBYI_gY?t=15m11s
>>30432474
why did the pilot do this?
>>30429621
Stick the right guy from /v/ in that simulator and I bet he'll get a kill on the AI.
>>30432501
https://www.youtube.com/watch?v=y0rxpQI0ZbY
>>30432132
Think of it as like that movie about the nuke AI that wants to win the cold war "game" for America with a pre emptive strike on the Soviet Union. They make the AI simulate fighting an AI as smart as itself. It runs thousands of simulated nuclear exchanges and eventually concludes that "the only winning move is not to play".
>>30432623
AI's that get every advantage in a digital work will always beat humans. You guys are acting as if a player getting killed by a Quake bot with its accuracy set to 100% spells the end of humanity.
>>30432772
It does. Aerial combat is as much math in real life as it is math in a videogame.
atmospheric flight is ridiculously easy to model.
simple autopilot existing for 100+ years is a proof enough
>>30432800
>simple autopilot existing for 100+ years is a proof enough
>Hold altitude and compass angle
>Requires any calculations
>>30432761
Seriously, if you just automate everything to the point where it's basically like an arcade flight game, the odds tip back into the human's favor. Human and assisting AI that deals with the complex stuff is better than pure human or pure AI.
>>30432800
>atmospheric flight is ridiculously easy to model.
>>30432428
Really? What did it recap?
Its been a long time since I watched either.
>daily reminder Messiah is the sexiest VF
>>30432840
That's now how it fucking works /v/tard.
In videogames the AI is balanced not to be too fucking smart. In this scenario the AI has to try the best it can to overpower the player.
Imagine the most difficult AI of the entire ace combat series and let him use a plane of the capacity of the player, and make it a little bit more difficult, and make it learn your tricks as it fights you.
That's what you are facing with this new AI.
>>30432772
>mfw
>>30432800
>simple autopilot existing for 100+ years is a proof enough
I was skeptical, but:
>The first aircraft autopilot was developed by Sperry Corporation in 1912. The autopilot connected a gyroscopic heading indicator and attitude indicator to hydraulically operated elevators and rudder. (Ailerons were not connected as wing dihedral was counted upon to produce the necessary roll stability.) It permitted the aircraft to fly straight and level on a compass course without a pilot's attention, greatly reducing the pilot's workload.
>>30432862
>>30432835
>HURRR
how unexpected
>>30432800
Al can't be omniscient in real life. I bet that the AI in the simulation is programmed to be able to instantly see any enemy plane within a direct line of sight of it's own plane.
>>30430287
AI will replace a lot of things in our world, because we are driven to do things in the most efficent manner possible, at the lowest cost possible.
This AI is still not Sentient AI, but is basically "ace combat" in the real world. While that is a gross simplification, its as good as its gonna get here.
However, as for you...You will Fly anon, but you will not fly one jet, you will fly a fleet of jets, and you will control them as if they were your fingers by an extension of your will.
Distributed intelligence
Distributed Nervous System
"smart skin"
Brain computer interface
Symbiotic/ Sympathetic AI.
tl/dr: we will start getting personal AI's to assist us in our cognative tasks, making us as creative as we are now, but faster, more precise. Yes there will be new and exciting problems and horrors to come with this, but then, isint there with each new invention?
https://en.wikipedia.org/wiki/Intelligence_amplification none of this is new "news" We are building all the components, and none of the screaming of bill gates, elion musk or any of those other goobs is gonna stop this.
>tfw you will collaborate with the robots if robot x human war starts
The ultimate revenge againts the normeis
>>30432900
This is what network warfare is for. IRL AI can easily feed of literally all networked platforms in the theater, including awacs, drones, radar satellites, etc. A little bit of prediction and interpolation will fill in all the momentary blanks (which is not even that necessary desu)
close enough to simply knowing where your enemies are by design.
>>30432939
>tl/dr: we will start getting personal AI's to assist us in our cognative tasks, making us as creative as we are now, but faster, more precise. Yes there will be new and exciting problems and horrors to come with this, but then, isint there with each new invention?
http://squid314.livejournal.com/332946.html
(but decentralized is still better than centralized)
>>30432939
No.
It will be "engage dogfight mode and let go of the wheel". AI will wake you up from G-induced coma in several minutes, if you both are lucky.
>>30429621
>>30432178
>>30429730
>>30429778
the issue with this line of thinking is that you're all assuming that nobody will bother to produce a decent countermeasure against such a powerful technology. I predict that by the time advanced AI becomes battle-worthy, a countermeasure will have already been created and we'll be more or less back to square one before the Habbening even had a chance to occur.
>You could literally control an entire army with one console/workstation
which sounds great until a script-kiddie gains access and opens up Roller Coaster Tycoon 2 windows until the system crashes with no survivors.
>>30432876
So long as I don't have to deal with pilot nerd stuff like trimming and missile arming, I could win that by being a cheeky cunt or by chance. I used dumb missiles fired from rocket pods to kill ace planes in HAWX. Since the ace pilots are programmed to fly towards me until they can engage, and they don't try to evade unless they detect a lock-on, I can shoot dumb missiles at them from outside lock on range and the aces fly straight into them. This is an example of how AI can be exploited by doing things it's designer didn't anticipate.
>>30432473
>>30432516
and idiots who love entertaining movies but know nothing about the field they talk about can be really fucking dumb sometimes.
>>30433065
No anon. That's an example of a retarded AI made to please someone as retarded as you.
>>30433039
And you base your assuredness on what exactly?
All weapons have effective countermeasures to them? Then there wouldn't be even a premise for any kind of war.
>>30433039
>which sounds great until a script-kiddie gains access and opens up Roller Coaster Tycoon 2 windows until the system crashes with no survivors.
>implying opening several RCT2 games at the same time and watching an AI play all of them differently at the same time wouldn't be fantastic.
>>30432886
Automatic flight really isn't that complicated. Automatic driving is because there's an insane amount of variables to keep track of.
>>30433084
Well spoken out of your own ass m8.
>>30433008
why would you stick your frail body in the machine? If you are going to go AI controls ship entirely, it would be the pilot/s/ in a mothership at a distance, surrounded by an AI formation. Communication would be done via laser to prevent traditional wireless "hacking".
There is no reason to put meat in a ship that can go full AI. even at that, once lasers come online in an affordable way, until we develop a way to defeat them, high speed maneuverability will be pretty worthless, and it will be more about ECM and first strike capability.
War evolves constantly.
Fuck, its why the sixth gen fighter concept was scrapped, because by the time we get it where it needs to be it would be obsolete for the technology its facing.
>>30433084
How much is Cyberdyne paying you
Yandere drones when?
>>30433119
you had to resort to m8, and offering nothing. my scenario still stands. a shitpost for your shitpost. now go get educated and be a productive member of society.
>>30432328
People keep saying the F-35 might be the last manned jet the military ever makes. Disagree.
The F-35 will be the last single-seat jet the military makes. From here on, its gonna be two-seat fighters controlling swarms of drones.
Your future wingman will be a drone.
>>30433107
That's the point. It's much easier to make an autonomous plane than an autonomous car. Which is kind of a moot point now since both are already made.
Making it dogfight-worthy is not even a problem since every modern fighter is FBW anyway.
It's just code ffs. Code is doable.
>>30433137
http://www.cyberdyne.jp/english/
https://en.wikipedia.org/wiki/Cyberdyne_Inc.
Sadly nothing.
>>30429794
>AI-controlled ballistic missiles
>>30433150
I already have my neuroscience degree, thanks.
That's actually one of the reasons why I'm not even trying with you. It's more efficient to just throw an insult back than spend an hour explaining to you why a dogfighting AI in a real world is not an insurmountable problem. As it is not impossible to have an autonomous intelligence-gathering network. As it is not impossible to automate the decision cycle in that network, and link it back to some of the weapons.
>>30433039
>a countermeasure will have already been created
such as...?
>>30433103
I don't think this is true. More the opposite: if there were a weapon with no countermeasures, nobody would dare go to war with the party that controlled it. This is the theory of the nuclear peace.
>>30433211
>I already have my neuroscience degree, thanks.
And Bruce Jenner is a woman.
>>30433135
>There is no reason to put meat in a ship that can go full AI.
Other than reluctance to put the kill decision in an AI's hands.
>>30432939
>This AI is still not Sentient AI
No shit. Its required level of processing power can be provided by a Rasberry Pi. What's my concern is how... innocent it is. Does a master class chess computer understand that it's playing chess? No more so than a combat AI understands it's killing people. I don't want AI to be put in positions that let it decide whether it wants to kill people. It should always be seen as a tool that's brought out when humans decide something needs to be killed.
>>30433224
Both are true. But it's two extremes of the spectrum. The point was that weapon effectiveness does not guarantee that there's an effective countermeasure to said weapon.
And nukes are not used only because it's guaranteed impossible to control or predict the outcome of an exchange.
But if you *think* you can control the very-very effective, quick and precise weapon, wouldn't you use it?
>>30433241
>I have nothing to say but I'm still here, undefeated! Watch out!
>>30433263
>wouldn't you use it?
Wouldn't have to. Deterrence.
(Or would have to use it once, pour encourager les autres.).
>>30433280
Not even the same guy, senpai. It's like on the internet you can say whatever you want! You can even pretend to have a nonspecific degree in neuroscience (as opposed to a masters or a doctorate, where it would actually mean something. Lie better next time)
>>30433260
you step into this conversation acting irrational and not having a full grasp of what an existing AI can, and cant, do.
which is: whatever its told to do within the perimeters it's given.
If you give it sloppy perimeters , you get sloppy results, but its still the result of human error.
Want an AI to kill the enemy, but not your own people, or civilians?
Do not use kinect level people sensing, with a crude device like blue tracking + GPS "box" to dictate the AOE.
Give it the ability to determine human shapes, weapon shapes, estimate where the weapon is pointing and if they are disarming. I can go on. This shit is why I dont often talk about it, but the ignorance is getting to a point where people are deciding they know a few peices of the picture so they know everything. Unless you are studied in this field, please, sit back and enjoy, ask questions and learn, but for the love of fuck, dont deign to dictate what should and should not be done.
At the end of the day, an AI is just a magnificently complex logic gate, and its gonna do EXACTLY what you told it to do. if its fucking up, its because a human fucked up something, even if that something is the code.
>>30432615
>B-52 is being flown by Lt Col Bud Holland
>Practicing for an air show
>Col Holland has a history of being "aggressive" pilot
>Almost crashes B-52 in to own daughters softball game in 1991
>Reprimanded for unsafe flying practices in 1991, 1992, 1993, and 1994
>"Bud your style is highly unorthodox and down right dangerous but god damn it if you're not the finest pilot I've ever seen."
>Never saw any real repercussions for doing dangerous shit where other peoples lives were at stake besides his own
>Kept doing dangerous shit
>Crashingthisplanewithnosurviors.avi
>>30429621
>not using this tech robot waifus.
WTF, guys?
>>30433039
>the issue with this line of thinking is that you're all assuming that nobody will bother to produce a decent countermeasure against such a powerful technology
The only countermeasure to an AI will be a smarter AI or a smarter missile.
>>30433103
>All weapons have effective countermeasures to them?
I challenge you to name one that doesn't.
>Then there wouldn't be even a premise for any kind of war.
If every single weapon and its respective countermeasure was completely omnipresent, sure. We live in the real world where this isn't the case.
>>30433106
>a dozen AI cooperate in multiplayer
mite b cool. I wonder if one of them will inevitably be a shitter and ruin everything like when humans play
>>30433224
>such as...?
we can only guess, using past developments as examples it'll likely be something most of us can't possibly anticipate.
>>30433263
>But if you *think* you can control the very-very effective, quick and precise weapon, wouldn't you use it?
seeing as how all things are relative, this can be applied to virtually any weapon. Use ultimately depends on the calculation of how much shit you will be in afterwards.
>>30433337
this guy gets it
>>30433305
>it doesn't mean anything if you have it
>and you don't have it
>everyone is lying all the time because they can
>>30433298
deterrence what? If a weapon is not a wmd and is very effective, where does "deterrence" come from? Or are you saying that the more weapon is effective the less it will be used out of fear? That's not true senpai.
>>30433332
Your pretentiousness leading to people becoming angry with you is probably why you have bad experiences when talking about these things.
Anyway, you echoed what I said. AI should not be used in situations and for things that AI shouldn't be used for. The situations and things are semantics.
>>30433332
>if its fucking up, its because a human fucked up something
I am not an expert, but lots of seemingly well-informed people think it's very hard to give a hypothetically very powerful AI the correct instructions.
Stuart Russell:
>As Steve Omohundro, Nick Bostrom, and others have explained, the combination of value misalignment with increasingly capable decision-making systems can lead to problems—perhaps even species-ending problems if the machines are more capable than humans. Some have argued that there is no conceivable risk to humanity for centuries to come, perhaps forgetting that the interval of time between Rutherford’s confident assertion that atomic energy would never be feasibly extracted and Szilárd’s invention of the neutron-induced nuclear chain reaction was less than twenty-four hours.
>>30433352
>I challenge you to name one that doesn't.
literally any weapon that kills people, compared to all the goofy inventions that don't. Not a hard concept.
>>30433394
>If a weapon is not a wmd and is very effective, where does "deterrence" come from?
I'm not sure what non-WMD examples there are. Maybe something like early Maxim guns before tanks and trench warfare. And most of the instances in which they were used involved mowing down hordes of nogs or Indians until the remnant decided to quit fighting--which is war, I suppose, but not much of one.
>>30429778
>Implying the greatest mind won't hack said system for their use as moles
>>30433352
>I challenge you to name one that doesn't
sub-launched icbms?
>>30433367
If you actually had it you wouldn't be this assblasted about being told you don't.
>>30433431
your point must be flying way over my head because I can think of millions of examples of counters to effective weapons
>implying the Tu-2Sh wouldn't kill people bretty gud
they just needed more money for that program
>>30433490
deterrence has worked pretty well so far, I could say ABMs but they're not really prolific enough to call them a proper counter
>>30433530
What if egotistically I enjoy my true superiority over your denial?
>>30433490
I'm cautiously optimistic. F-22s already have a kill ratio of something like 30 to 1 against any other plane. Now we have AI pilots that can win 4 on 1 battles against humans in other F-22s, even when the humans are given superior engagement range.
This is a level of force multiplication that completely mitigates the numerical superiority of the Chinese military. They would undoubtedly lose the air battle in any conflict they pick with us.
>>30433536
How old are you? I'm seeing that you can't use grammar correctly and it's giving me the notion that you're naive.
>>30433604
I'm not worried about the Chinese so much as I'm worried about whoever gets their hands on AI weaponry liquidating everyone they don't need in order to turn their bodies into computronium.
>>30433604
>F-22s already have a kill ratio of something like 30 to 1 against any other plane
Battle proven - when?
>>30433626
>current year
>he doesn't know about Spurdo-speak
wew lad
>>30432900
>>30432876
>>30433039
>>30433084
>>30433211
>Al can't be omniscient in real life
All you have to do for this to happen is have ONE OVERZEALOUS SCIENTIST trying to create a self-learning and self-sustaining program for the AI and it's all over.
That's it.
>>30433418
you made a statement, which is correct, but backed it up with something that had nothing to do with it other than trying to highlight the fallacy of man's hubris.
Really that quote has more to do with "where there is a will, there is a way" rather than "we said it cant happen, so it wont." on safety issues.
I actually have to rant for a quarter sec here, I am deepy disturbed by the deep integration of feelings over ration/ logic in today's conversation/ debate material. It is wrong, wrong on all levels.
We are humans, thinking/ feeling creatures. We can think about our feelings and not let the rule us, while still experiencing them, which is part of what makes us human.
Putting on my asshole hat, the people who choose not think, are omitting one half of what it is to be a human.
This was no dig at you.
>>30433585
not anyone, but watching this exchange. You are just trolling, and you have mental capacity, its a shame you dont utilize it in a more practical manner. the era of anonymity is gradually coming to an end, so the "casual sadist/ everyday sadist" behavior is also becoming impractical.
Might as well future proof your shit yo.
but here, let me finish this coversation for us.
"get bent fag, i like doing what i like"
-
"thats a sham"
-
"trololol"
you masturbate, eat some chips, and wake up for the next day.
>>30433664
This to be quite honest.
>>30433927
You. You are the best troll in this thread at the moment. For a sec I thought it was serious, but your use of caps gave it away.
Attached is a reward.
>>30433664
i am worried about automatic guns and whoever gets their hand on rapid firing weaponry liquidating everyone they want.
>>30432474
B stands for bomber. Different specs. Moron.
>>30433982
What's wrong with what I said? As others have pointed out, current AI is limited by the parameters you give to them.
But if you have a fucking stupid scientist who, in the ambition of wanting to be a Creator of sorts, creates an AI that can surpass its parameters through self-learning and self-sustaining algorithms, which can also reproduce itself through the internet, then yeah, shit will be fucking over.
Tell me again: what's wrong with what I said?
>>30433986
Explain.
>>30433987
B stands for Strategic bomber. A for Tactical.
>>30433834
>implications that the world will ever go to a world war
>>30433431
I hate when this picture gets posted. Because as silly as that looks, the United States took it a step farther.
https://en.wikipedia.org/wiki/Lazy_Dog_(bomb)
>Experimental Lazy Dog projectiles of various shapes and sizes were tested at Air Proving Ground, Eglin AFB, Florida, in late 1951 and early 1952. An F-84 flying at 400 knots and 75 feet above the ground served as the test bed while a jeep and a B-24 were the targets. The result was eight hits per square yard. Tests revealed Shapes 2 and 5 to be the most effective. Shape 5, an improved basic Lazy Dog slug, had the force of a .50 caliber bullet and could penetrate 24 inches of packed sand. Shape 2 could penetrate 12 inches of sand — twice as much as a .45 caliber slug fired point blank.
>>30434047
cont.
Basically, why fire from a gun, when you could drop it from a plane and have the same impact.
>>30434026
Whenever anyone uses "i dont want X type of people to get Y weapon" its the same argument anti gunners use. About difference of power.
The issue aint about people getting AI piloted planes, but about everyone being able to get one.
>>30434069
Ehh, I think it's more about the machines getting a life of their own and becoming the species destroyer that they have been in Terminator and the Matrix. Because they will be far more scary than the Chinese or the Russians. With humans, you can expect human error. With machines that can learn shit on the fly and adapt through faster thought processes and can hack computer systems faster than a human can defend against?
Yeah.
>>30433948
>backed it up with something that had nothing to do with it other than trying to highlight the fallacy of man's hubris.
The point is that what Russell calls "value misalignment" is the problem of telling the machine to do the right things. If you don't, it's going to maximize something you don't want maximized.
>>30433986
And I'm sure the Fuzzy-Wuzzies and the Comanche are laughing at such a silly idea. Shame I can't find any around to corroborate.
>now without snark:
Reasonable people may disagree, but it seems to me that there is at least good reason to worry that sufficiently advanced automated militaries will indeed render most human beings obsolete. And that we're seeing the beginnings of this already, with the surveillance state and increasing governmental tyranny throughout the Western world, as individual rifle-armed infantrymen have become less important for power projection relative to high tech. Maybe the end scenario is not "liquidated for computronium" but rather "given a basic income, sterilized, and supplied with free opiates", but that's still no way to go.
>>30434047
>>30434060
DARPA - Public tax dollars at work!
>>30434069
>The issue aint about people getting AI piloted planes, but about everyone being able to get one.
Right, and that's an important problem we need to solve. Hence blocking attempts to register drones, among other things. But I don't know that we can expect this to happen before the first guys to get autonomous weapons take over and prevent any such thing from happening.
>>30434126
>DARPA - Public tax dollars at work!
You wouldn't be shitposting here without them!
>>30434047
>>30434060
I don't see how the US developing something (theoretically) as stupid in anyway diminishes the stupidity of strapping dozens of PPSH's into an airplane.
I wouldn't call air dropped kinetic projectiles a 'step further' they're different things and honestly the US version looks like it was at least vaguely viable for a time.
Something tells me that people at DARPA are on /k/ shitposting with shitposters at this point of the thread.
>>30429621
AI will always beat a human at any given specialized task.
GAI, when developed, will have humans defeated at any task.
Literally nothing surprising about this, all superpowers are developing AI. EU, China and Amerikan
Russia is the only big regional power that fails at producing anything computer related.
>>30434200
>Russia is the only big regional power that fails at producing anything computer related.
This.
Russias only achivement is producing vacuum tubes that sound good in an amplifier
>>30434257
they're also pros at making nixie tubes
>>30431293
>Thank you for proving my point that airframe G ratings are defined by the pilot, not the equipment.
Yeah except for the part that the other anon didn't say that. Missiles aren't overly limited because once they're launched, they're never going to be used again, because they explode at the end of that flight.
Also missiles are much smaller than a fighter, so their structure can be made to more easily endure higher stresses, especially over their short, less than two minute usually, life spans.
>>30432710
You're talking about that movie with Matthew Broderick from the 80s. War Game or some such. Broderick makes the AI play itself in tic-tac-toe until it figures out it can't beat itself then it plays out all the possible nuclear strike strategies and resets itself when it figures out that MAD works no matter what it tried.
My dream A.I pilot is one that has the accumulated knowledge of all past fighter aces.
It's automatically better than you will ever be.
>>30431310
Shes right though, it's gone from Evrope to "Europe"
>>30434027
>A for Tactical
More like Attack. Though you are correct about their roles for the most part. The line is blurring somewhat with the increasing use of guided munitions. Also don't forget retarded misinformation/psy-ops shit like the F117, and shenanigans like trying to shoehorn an A onto the F22.
>>30432045
>>30432104
>both Ai planes fly off into the distance
>where'd they go?
>phone rings
>Israel destroyed
>world peace achieved
>>30429621
Windows 50
What kind of botnet will be?
>>30429621
More likely it'll be AI systems incorporated to take over and carry out maneuvers that a pilot can not, while the pilot is temporarily incapacitated and then resuscitated, so that the plane has enhanced capabilities, but there is still a pilot. Drone dogfights (botfights?) could be interesting too, but I can't see a scenario where not having a pilot in the air is enough for every situation.
>>30434645
>implying it isn't palestine instead.
>>30432474
That pilot was a real piece of shit, but that looks more like a stall than g problems
>>30432012
To a limited extent; the MQ-1, MQ-9 and RQ-4 all have cockpit-like volumes to accommodate proper satellite dishes for better satcom. Though to be fair, dishes aren't stealthy and thus aren't used on aircraft like the F-35 (rather they use a conformal patch-style antenna).
>>30431293
I just said that a missile can pull a heap of Gs because it doesn't matter if it gets cracks in its casing or structure after the first turn or two. If you want to trade millions of dollars training pilots for millions of dollars in repairs fixing a drone every time it lands, be my guest.
>>30429867
>Flying is for droids..
.>>30429621
>>30432840
hi, welcome from /v/.
In games, the computer has a difficulty setting, which is finely balanced to not cause butt hurt. in most games, even arcande games, there are some very hard settings, but no "completely ruin this kids day" settings. An AI, calculates things faster than a person can, it can also micro manage everything all at once. It can call on literally millions of bits of data to determine every likely outcome, and how to respond as programmed.
>>30433927
Took me seconds to realize you're trolling
gotAKekAndAReply/10
>>30429778
>>30429780
Now you see why people like Elon Musk and Stephen Hawking are so worried about the future of AI. They are not technology fearing luddites like some have painted them, but very rational people who fear a future where certain groups (states, corporations, the rich) possess a monopoly on militarized AI technologies that allow them to control those who have no access to AI themselves. When true, god-like AI are created sometime this century, who do you want them to be controlled by? Scientists and philanthropists who use them to cure disease, create beneficial technologies, and improve the average person's life in every way possible? Or would you rather it be corrupt governments and corporate states that oppress citizens and murder enemies of the state with ruthless efficiency? Eventually these AI will grow beyond our control, and when that occurs the human race will be at the mercy of living gods. Whatever happens next will be directly determined by their previous experiences, and the lessons they learned working for governments and private companies will shape their general view of humanity. An AI that views us as disgusting, self-centered vermin that have spent our entire history murdering each other and raping the planet will probably have no issue with eradicating us. Yet an AI that has experienced compassion and human goodness could elevate us to the next level of human evolution, merging man with machine to take our first step into godhood. Our decisions in the next few decades will determine whether we reach the stars or die on Earth, exciting times indeed.
>>30433305
Well neuroscience is a real degree, my brother has one as well. And for the record, he thinks AI with human-level intelligence will be created before the end of the century, so this guy is just being a jackass using his probably non-existent degree to justify his small-minded views.
>>30435644
>cientists and philanthropists who use them to cure disease, create beneficial technologies, and improve the average person's life in every way possible? Or would you rather it be corrupt governments and corporate states that oppress citizens and murder enemies of the state with ruthless efficiency?
You say that like those are two different things.
>Of all tyrannies, a tyranny sincerely exercised for the good of its victims may be the most oppressive. It would be better to live under robber barons than under omnipotent moral busybodies. The robber baron's cruelty may sometimes sleep, his cupidity may at some point be satiated; but those who torment us for our own good will torment us without end for they do so with the approval of their own conscience.
>>30434638
Nope, the official definition for A designation is Tactical Bomber.
>>30432305
youre wrong, they will never take guns away. shut the fuck up you ignorant faggot. If you honestly think the american gov will take arms away from people or think that the american people will just let the gov do whatever youre wrong. turn off fucking fox and cnn and watch some c-span and talk to people. Ignorant fucks like you make places like /k/ fucking miserable.
>>30437001
Sounds like the real moron is yourself.
>>30429621
>air battle manager
good job calling picture clean about 4 minutes after i need it, champ.
>>30437001
I agree that he's wrong about the status quo, but the worry is that sufficiently advanced automated law enforcement could make the gun-grabbers' dream a reality.
>>30435908
>but those who torment us for our own good will torment us without end for they do so with the approval of their own conscience.
As long as our robot overlords destroy the jews I'd polish their chrome-y cocks all day long.
>>30437339
>tfw jews are the robot overlords
>>30437371
>Neo-liberal ultra pro-mondialisation stateless bourgeoisie in literal absolute control of the world
Wake me up from and tell me the world being a bad rip-off of judge dredd was just a nightmare (if it's reality just kill me).
>>30437309
so for those of you who don't know shit about shit, an ABM ain't shit. they're useful for telling you that you can kill things without going to jail, and giving you some SA. the fighters run the intercept/do the BVR-WVR transition, etc. with ABM pointouts. done right an ABM is extremely helpful for maintaining your SA. done wrong you should turn down the volume on your radio because that's literally the best outcome.
>>30437524
>can't wake up
Well, now you know who our enemies are.
Cody Wilson has some nifty ideas. I'd say our best shot probably lies in that general direction.
>>30437684
>Cody Wilson
>A leftist crypto anarchist
Erf...
>>30437773
>3d printing drones
>thymotic anarchist
>crypto anarchist as in cryptography, not crypto anarchist as in "pretending not to be an anarchist"
>reads foucault and understands that vast faceless egregores are conspiring to eat humanity
Believe me, I'm no fonder of the usual
>hey guize I'm an anarchist! btw evil straight white republicans amirite das rayyyycis
types than anyone, but this guy is on a whole different level.
>>30437853
>thymotic anarchist
Anarchy can't not be about isothymia and megalothymia anyway... Which one does he support ?
>>30430251
>Airframe limits under load are generally speaking under 9g.
Generally because pilots can't fly with more than 100m/s/s of acceleration. But that's by no means the limit of what can be built or fly.
>>30437931
That is a good question. I wasn't aware of the distinction until now.
I thought he meant it as a contrast to those anarchists who would basically be on board with everything becoming a giant, very clean, very well-stocked and well-maintained shopping mall (or hippie commune, as the case may be). Anarchy of the thumos or spirit, not just the absence of government. But I don't know.
>>30434870
it was a stall caused by/excacerbated by G problems. at that angle of bank, he didn't have enough upward component of lift at that airspeed to keep flying, and he kept losing knots by trying to wrap it around tighter, and he had 8 shitty low bypass turbofans with no afterburner.
and yes, Bud Holland should have had his wings pulled.
>>30430383
honestly that's pretty good, my resting G tolerance was like 4.5 and that was pretty good. then again i don't notice 2 G's anymore.
>>30438058
>Anarchy of the thumos or spirit, not just the absence of government.
So he means anarchy as a mean to preserve thinkings and cultures outside of the capitalist "cosmopolis" that wants to absorb everything ? That I can ship.
>>30429708
>>30438029
https://en.wikipedia.org/wiki/John_Stapp
>By riding the decelerator sled himself, in his 29th and last ride at Holloman Air Force Base in New Mexico, Stapp demonstrated that a human can withstand at least 46.2 g (in the forward position, with adequate harnessing). This is the highest known acceleration voluntarily encountered by a human, set on December 10, 1954.[6] Stapp reached a speed of 632 mph (1,017 km/h),[7] which broke the land speed record and made him the fastest man on earth. Stapp believed that the tolerance of humans to acceleration had not yet been reached in tests. He believed it is much greater than thought possible.
Human tolerance is the least important factor in airframe G-loads.
>>30432104
Obviously, they tried to settle their differences with a game of tic-tac-toe.
>>30432474
That's a stall, not G-induced airframe fatigue.
>>30438574
>Stapp believed that the tolerance of humans to acceleration had not yet been reached in tests. He believed it is much greater than thought possible.
>Honey, I need to go back, I need to find the limit!
>John, stapp!
>>30429708
>AI controlled AI
>>30432045
Both AIs had a commlink did they?
>Chinese AI jet sits on tarmac, linked into both, hacking the fuck out of the pair. Outcome decided is whatever China needs that day.
Chinese AI jet doesn't even have functional engines, but nobody noticed coz it never needed to leave the ground.
>>30438574
Its great to say that but the acceleration he was experiencing were backward / into the seat, not parallel to the spine or lateral. Its far, FAR easier to deal with forward/backward Gs.
Also got to stay conscious, lucid and thinking properly while under such a state otherwise its not exactly useful.
>>30429621
Can we just all agree if an AI company is ever bought out by Boston Dynamics, we should all start burning it down?
I swear to god Boston Dynamics is going to kill us all.
>>30430162
ROU Begin Street Sweeping Operations
>>30434722
>implying the destruction of a nazi style ghetto and an open air prison would achieve anything
high g loads dont harm you if your under water btw
SOON
>>30440786
Of course they do.
>>30440801
Actually, he meant if your body is submerged in a sealed container.
Water does not compress.
>>30440811
If the human body was 100% identical in density to the submerging fluid, there would be a point.
But it's not. The human body has varying densities through it's mass. Therefore high Gs would still damage bodily structures.
>>30440811
And?
You're still going to have the blood drain from your brain and experience GLOC.
>>30435908
They are. One creates an AI that hates the human race for what it has experienced, and the other at least starts out life with the simple goal of helping people. There are measurably different outcomes.
>>30440828
No I don't believe that is true.
>>30440791
>just in time to see AI jet pilots outskill humans like in my japanese vidya and animes
If AIs are put into real warmachines, they wouldn't be aware of death or killing. It would simply be a game of "tag".
AI jet chases opponent, tries to "tag" opponent with various tools (gun, laser, missile, etc). Repeat until fuel low, odds of winning too low (retreat), or destroyed.
It would literally like a dog playing catch the ball.
>hard to imagine, but the T-800 was never a hateful, evil creature. It was merely fulfilling it's purpose.
>>30440878
Because you're a moron.
>>30432939
>and you will control them as if they were your fingers by an extension of your will.
>Distributed intelligence
>Distributed Nervous System
>"smart skin"
>Brain computer interface
>Symbiotic/ Sympathetic AI.
We 40k imperial navy now
>>30429730
"But what happens when the enemy steals the keys"
>>30439491
Muh brain
>>30440903
"Liquid immersion provides a way to reduce the physical stress of G forces. Forces applied to fluids are distributed as omnidirectional pressures. Because liquids cannot be practically compressed, they do not change density under high acceleration such as performed in aerial maneuvers or space travel. A person immersed in liquid of the same density as tissue has acceleration forces distributed around the body, rather than applied at a single point such as a seat or harness straps. This principle is used in a new type of G-suit called the Libelle G-suit, which allows aircraft pilots to remain conscious and functioning at more than 10 G acceleration by surrounding them with water in a rigid suit.
Acceleration protection by liquid immersion is limited by the differential density of body tissues and immersion fluid, limiting the utility of this method to about 15 to 20 G.[57] Extending acceleration protection beyond 20 G requires filling the lungs with fluid of density similar to water. An astronaut totally immersed in liquid, with liquid inside all body cavities, will feel little effect from extreme G forces because the forces on a liquid are distributed equally, and in all directions simultaneously. However effects will be felt because of density differences between different body tissues, so an upper acceleration limit still exists."
https://en.wikipedia.org/wiki/Liquid_breathing#Space_travel
Total liquid immersion, including inside lungs, allows G-force resistance of greater than 20 G. Pressure resists the G forces on the body evenly and prevent blood from pooling in certain parts of the body. So no, you fucking moron, you will not pass out from blood drain in a liquid filled container unless you get to incredibly high G levels.
>>30441400 This is what I said here (though expertly described):
>>30440825
It appears there's a workaround on the drawing board. Good stuff.
>>30432710
>>30434507
>AI won't launch nukes because everyone loses
that's bullshit. why would the AI have a concern for the general preservation of humanity and/or the environment? it's mission is to win. it wins by killing the enemy. any retaliatory strike is an acceptable consequence as long as the enemy is destroyed.
also, there is the issue of one side having more/better WMDs. that's clearly an advantage and an incentive to strike first
>>30442569
>why would the AI have a concern for the general preservation of humanity and/or the environment?
Because you can make that one of its objectives?
>>30439743
>CHINA STRONK AI shitposting on /k/
I almost want to see how that plays out. Almost.
>>30442569
If it's programmed for mindlessly destroying the enemy with no regard for damage received, sure.
But I would imagine that it was programmed to maximize enemy destruction while minimizing friendly destruction. Counterforce strikes and all that.
By forcing it to play itself in this scenario, it went through every conceivable course of action before concluding the best way to minimize damage taken is simply not engaging.
That's the point of the movie. The only winning move is not to play.
>>30434994
>Though to be fair, dishes aren't stealthy and thus aren't used on aircraft like the F-35 (rather they use a conformal patch-style antenna).
Conformal antennas can be arrays that are capable of beam steering so you can direct the signals to the satellite or relay drone you are linked with. Especially Israel is very experienced in relaying.
A newer approach is laser communication, capable of even higher bandwidths. And this too takes up a small volume.
>>30442603
you could, but why would you for a weapon system AI?
>>30442736
sure, but would you want such a system in charge of your strategic national defense?
if the enemy launches the AI could decide not to counter strike because that'll save the most lives.
i can see the logic in programming it to never preemptive strike, but why give it the ability to just give up?
>>30443077
>you could, but why would you for a weapon system AI?
..precisely to avoid such a scenario.
What do you think stops the US from just firing everything it has right now at China or Russia?
>>30443228
Democrats
>>30443077
The point of the movie is that the AI's capability is beyond what its creators and users expect, thus allowing it to come to a more humanlike conclusion. The only winning move is not to play.
The worst part is that there is NO way we can stop the governments from developing the AI.
If we outlaw it they will do it in secret because another state will do it.
