Are you joining the Military, and if so, what branch?

Probably wont join. But if I did, it'd be the Marine Corp. And if I'm right(Though I doubt it.) those marines are running a training excersise. Which would explain why the barrel has the plug on it.

Unless I'm of a good military age when WWIII comes around... nope

Everyone is either over estimating the capabilities of programming or underestimating it.

The ones who are underestimating it are thinking of generic algorithms, if then, if you see this happen then do this... ect. In reality programming can do whatever we want it to do, in the case of bob for example, we can emulate our brain's own system by programming individual neurons, and giving it algorithms to adjust the synaptic connectivity, allowing it to learn. If we are capable of something, then so is a program. Not only do robots have the potential to think out of the box, but their brain has instant access to a computers calculation abilities, among other things. Eventually, it is possible they could become self aware.

Now, in regard to the idea that such a robot could question orders, it would take a very poorly thought out program for it to have this ability. For one, every program is completely slave to it's algorithms. These are definite, and they control the robot's thinking and actions on the most base levels. For example, if we defined in the program that it could not "team kill" then it wouldn't be able to do this. As a precautionary measure, in the event that a human of the same faction is killed by the robot's actions, the algorithm would delete it and shut down the robot to be destroyed as faulty machinery. On top of all of this, there is another factor t consider. Since in a neural network the only way to learn is by the adjustment of synaptic weights, and this is preformed by an algorithm, the algorithm could control what and how it learns. For example, the network could be trained before combat, and issued to all of the robots. During combat it would use a type of reinforcement learning that only allowed it to learn how to better fight the enemy, and the enemy could not be changed.

Above all else, the human fighting force will never cease to exist. Such robots will fight alongside our soldiers, not for them.

I rest my case.

CivBase wrote:So which one of you Brits is joining the RAF?
SAS?
Royal Navy?

i salute to them but the real heros are the royal marines

and im not joining so who put im a brit (yes)

PiEdude wrote:
It's not the movies.

I mean, think about it.
For almost every case of slavery in history, there has been some kind of reisistance or rebellion.

Except that the rebellions in North America usually failed, because of population percentages and labor types.

PiEdude wrote:So what happens when the slaves all have guns, and we don't anymore, because we gave all of them to the slaves?

Except these slaves won't be intelligent enough to realize what slavery is. They won't be able to rationalize their existence out.

A machine doesn't need human intelligence to kill. Just parameters and a CPU.

If i have to join the military for some reason in the future, i would choose to be a combat engineer for the army.

Baconsen wrote:If i have to join the military for some reason in the future, i would choose to be a combat engineer for the army.

i would have to be RAF

Dud Doodoo wrote:Everyone is either over estimating the capabilities of programming or underestimating it.

The ones who are underestimating it are thinking of generic algorithms, if then, if you see this happen then do this... ect. In reality programming can do whatever we want it to do, in the case of bob for example, we can emulate our brain's own system by programming individual neurons, and giving it algorithms to adjust the synaptic connectivity, allowing it to learn. If we are capable of something, then so is a program. Not only do robots have the potential to think out of the box, but their brain has instant access to a computers calculation abilities, among other things. Eventually, it is possible they could become self aware.

Now, in regard to the idea that such a robot could question orders, it would take a very poorly thought out program for it to have this ability. For one, every program is completely slave to it's algorithms. These are definite, and they control the robot's thinking and actions on the most base levels. For example, if we defined in the program that it could not "team kill" then it wouldn't be able to do this. As a precautionary measure, in the event that a human of the same faction is killed by the robot's actions, the algorithm would delete it and shut down the robot to be destroyed as faulty machinery. On top of all of this, there is another factor t consider. Since in a neural network the only way to learn is by the adjustment of synaptic weights, and this is preformed by an algorithm, the algorithm could control what and how it learns. For example, the network could be trained before combat, and issued to all of the robots. During combat it would use a type of reinforcement learning that only allowed it to learn how to better fight the enemy, and the enemy could not be changed.

Above all else, the human fighting force will never cease to exist. Such robots will fight alongside our soldiers, not for them.

I rest my case.

If one sets off a nuke in NYC, it could think 'I did not kill them, the nuclear device did.' and it would be free.

Im not going to answer this poll yet... it all depends on whether I get a scholarship or not. If not, I need the money to go to college.

extrem back up im joining the CG

Coast guard is pretty hard to get into

Angatar wrote:

Dud Doodoo wrote:Everyone is either over estimating the capabilities of programming or underestimating it.

The ones who are underestimating it are thinking of generic algorithms, if then, if you see this happen then do this... ect. In reality programming can do whatever we want it to do, in the case of bob for example, we can emulate our brain's own system by programming individual neurons, and giving it algorithms to adjust the synaptic connectivity, allowing it to learn. If we are capable of something, then so is a program. Not only do robots have the potential to think out of the box, but their brain has instant access to a computers calculation abilities, among other things. Eventually, it is possible they could become self aware.

Now, in regard to the idea that such a robot could question orders, it would take a very poorly thought out program for it to have this ability. For one, every program is completely slave to it's algorithms. These are definite, and they control the robot's thinking and actions on the most base levels. For example, if we defined in the program that it could not "team kill" then it wouldn't be able to do this. As a precautionary measure, in the event that a human of the same faction is killed by the robot's actions, the algorithm would delete it and shut down the robot to be destroyed as faulty machinery. On top of all of this, there is another factor t consider. Since in a neural network the only way to learn is by the adjustment of synaptic weights, and this is preformed by an algorithm, the algorithm could control what and how it learns. For example, the network could be trained before combat, and issued to all of the robots. During combat it would use a type of reinforcement learning that only allowed it to learn how to better fight the enemy, and the enemy could not be changed.

Above all else, the human fighting force will never cease to exist. Such robots will fight alongside our soldiers, not for them.

I rest my case.

If one sets off a nuke in NYC, it could think 'I did not kill them, the nuclear device did.' and it would be free.

Nonsense,
1. The robot doesn't have the ability to learn anything other than ways to kill the enemy. It does not have a human brain and higher level functions that allow it to change it's focus. The way it thinks is predefined by the programmer, and it ONLY has the ability to learn to kill the enemy better. The interpretation of the enemy is not up for question on any level, for it is predefined.
2. A good programmer would be able to identify lethal action (setting off a nuke in an area that is filled with friendlies is a very easy to recognize lethal action) and stop this command from passing on to the motors.
3. The robot would once again, only have the abilities it was originally given, while it can come up with creative solutions to killing enemies, it once again cannot do anything outside that, save it's predefined abilities. While you might argue that it's "creative solution" could be indirectly killing friendlies, keep in mind that if a friendly is killed because of it's actions the robot is destroyed.
4. Higher brain function is not needed in a killing robot. Humans are the only species that have the ability to explore one's own brain and mind, this is only achieved at a very high state of awareness. Such robots would in no circumstances require this same higher level function to be efficient killing machines.
5.Simply put, I see no reason why a robot would be trained with knowledge of handling nukes.

true robots would have to be programed (hopefuly bu use and not someother intergalatic species) and were not stupid enough the give the robot such abalites to kill freindles and the only ecpetion to that is if a hostile was attakcing another freindly and the only way of killing the hostile had to deal with injury/killing the freindly so whick either the robot would decise a way to kill the hostile without hurting the hostile i could never see a "Terminator" robot rebblion happening.

Say you go into a urban enviroment, and civilians and enemies are mixed. How would it define an enemy?

P.S. I am winning this argument, as it seems.

Angatar wrote:Say you go into a urban enviroment, and civilians and enemies are mixed. How would it define an enemy?

"If you're not with me you're against me."

"Wipe them out, all of them."

I program my robots will "kill all human" routines


because I am a fucking rebel

KrAzY wrote:I program my robots will "kill all human" routines


because I am a fucking rebel

then you and your allies would be targeted aswell

Onyxknight wrote: then you and your allies would be targeted aswell

The drones fight alone on suicide missions (unless they're expensive)

allies?


why would I be targeted? I would program it to kill all humans... not all lifeforms

KrAzY wrote:allies?


why would I be targeted? I would program it to kill all humans... not all lifeforms

>.> i found ET!!!!!! you shall go home NAO!!!!!!!

Angatar wrote:Say you go into a urban enviroment, and civilians and enemies are mixed. How would it define an enemy?

P.S. I am winning this argument, as it seems.

Then it wouldn't fire unless it had clear shots...

As I said, an enemy is predefined in it's brain. It will ONLY attack enemies, or not attack at all.

Seeing as you have absolutely no ground except for movies, and you haven't made a good point yet, I don't see how you're winning...

KrAzY wrote:because I am a fucking rebel

lol!

Dud Doodoo wrote:

Angatar wrote:Say you go into a urban enviroment, and civilians and enemies are mixed. How would it define an enemy?

P.S. I am winning this argument, as it seems.

Then it wouldn't fire unless it had clear shots...

As I said, an enemy is predefined in it's brain. It will ONLY attack enemies, or not attack at all.

Seeing as you have absolutely no ground except for movies, and you haven't made a good point yet, I don't see how you're winning...

What exactly is a 'defined enemy'?

Enemies that have been defined as enemies

...

How will it identify and enemy?