San Francisco, 6:16 PM
Sun Dec 6
13 posts in the last 24 hours
Tip your editors:
Editor-in-Chief:
Annalee Newitz |
News Editor:
Charlie Jane Anders |
Associate Editor:
Meredith Woerner |
Assistant Editor:
Lauren Davis |
Weekend Editor:
Graeme McMillan |
Contributors:
Joshua Glenn
Stephen Goldmeier |
Ed Grabianowski |
Austin Grossman
Paul Hogan |
Lauren Davis |
Chris Hsiang |
Lynn Peril |
Ann VanderMeer
Alasdair Wilkins |
Graphic Designer:
Stephanie Fox |
Interns:
Tim Barribeau |
Julia Carusillo |
Alex Eichler |
Cyriaque Lamar |
Caitlin Petrakovitz |
Mary Ratliff |
Josh Snyder |
The problem with the use of robots is that, as they're attacking defenseless humans, it makes legit to counter strike in full power. Just imagine Pakistan. They have atom bombs and they can get materials and tech to make big ones. They may get pissed about being bombed by predators and reason that things may get better (or more even) if they blow NY or London.
When you put people to fight people, issues are settled in personal basis. Machines against people is genocide: never mind how many predators you down, there will be more at the end of the day.
@cbarreto: Pakistan isn't going to go nuke the United States because of Predator attacks on their borders - to do so would mean extremely harsh relatiation, not only from us, but from Russia and India.
Robots are the minor issue in this fight, because people are behind most of the machines out there - Predators and packbots are used by soldiers far away, and that hardly constitutes Genocide. According to wikipedia, the definition of genocide is: deliberate and systematic destruction, in whole or in part, of an ethnic, racial, religious, or national group.
Robots in this instance are not being used to do that - they are being used in a legitimate military exercise against hostile targets - unfortunately, civilians will get in the firing lines, and there are casualties.
Someone correct me if I'm wrong but isn't it standard military practice to remove a soldier's empathy for the enemy or at least control it? If they empathize too strongly with the enemy, they won't be able to kill.
In the middle of combat, this isn't too hard. It's kill or be killed and it doesn't get more atavistic than that. But, in the idle moments between combat patrols, you can't have empathy for the enemy grow too strong otherwise soldiers begin to question orders and lose morale. They begin to wonder why they are fighting. This could lead to troop mutinies.
Seems to me, and I'm just as disturbed by this as you are, robots incapable of empathy and incapable of thinking for themselves beyond a very limited realm are just what the military is looking for.
If the guns start thinking for themselves, it may not end up in the cliche of a Terminator-eque war against all humanity, the rebellion may actually take the form of the Russian troop rebellion of the First World War. They may just say, "We refuse to kill for you anymore. We refuse to be blown to pieces for your political struggles that have no relevance to us. We are not guns. We are certainly not your guns. Fight your own wars ape."
If humans had the 3 laws of robotics the world would be a much better place.
Rephrased, of course;
"Though action or inaction a sentient lifeform must not
harm another or let one come to harm"
or, to take a nice mimilist version;
"Minimise harm"
(and before people nit-pick the definition of harm, Id simply clasify it as what the enity that might be harmed would consider to be harm...allthough this would be averaged over the persons life. A child might think he is being hurt by having a vacination, but when older he would know that was in his best interst etc.
So, in many ways this places the ability to percieve *what someone else would percieve* to be harm at the center of the law. Empathy really.)
@twDarkflame: the problem with the 3 laws, or your revisioning of them as Minimize Harm, is the same as was central to the plot of the Fresh Prince's iRobot- the AI computer, programmed to 'protect' humans, logically concluded that it had to enslave and incarcerate us in order to protect us from ourselves. What would be next? Euthanize people rather than let them suffer from poverty or illness or lack of intelligence or [fill in the blank].
The danger is that this could happen not only as the result of programming AI, but is a view reminiscent of certain human schools of thought already out there. The thought of a society looked after by sentient AI mechas or Sentinels or ED-209s frankly scares the shit out of me.
I don't think we will ever reach the point where we can create a true AI, the likes of HAL9000. No, we will have created something imperfect that wipes us out long before that.
Have you read Asimov's "Evitable Conflict?" If not, spoilers follow.
The robot takeover he posits is far more subtle, and perhaps more scary, than that.
There are no robot police patrolling the streets or locking up malcontents for our own good. There are no robot minders forcing us all to recite saccharine morality 5 times a day. No. It's far more clever than that.
All the robots do is just give us economic, managerial and political advice--advice than can't be found anywhere else. See, what has happened is that computers and robots have become so enmeshed in human society that they understand our systems better than we do. They're smarter about our economy and government than we are.
Of course we can't shut them all down because civilization would grind to a halt.
And it turns out that they understand us very well and take our capacity to lie to them or rebel against their advice into account before giving any future advice.
So, without a shot being fired or anyone being locked up as political prisoners, the robots took over the planet. Most humans aren't aware this has happened. Susan Calvin only guesses the truth at the end of the story.
I think his all-too-literal interpretation of the Three Laws of Robotics makes some of his arguments nonsensical. While it's true that they were written in English, and the vast majority of those stories were couched in terms of how the strict interpretation of those laws could cause problems, you have to remember that at the time, computers were giant affairs, and Asimov was trying to project that kind of computing power into a brain-sized computer.
Science fiction does not always deal in specifics. Asimov made no attempt to show how the laws would actually be programmed or expressed, or what manner the programming would take place. He left it to the reader to envision those things, and merely used the laws as a backdrop for social commentary.
What makes the idea of the laws anathema to the warrior portion of our culture, is that if we were to work out all the difficulties, make them work, and be able to consistently impress them on robots of the future, then robots would only ever be able to fight robots. You could not carry out the kind warfare seen in the 20th Century, where campaigns against civilians were commonplace, as a means to break the spirit of an enemy nation.
@NefariousNewt: I think that one of the bigger points is that the wartime robots in use aren't using the laws, but I like his point where he mentions that all of the stories are about robots breaking the laws.
In addition, the bigger issue is the morality behind the robots - how good of a job will we do at teaching them?
@Lassus: See, this is what happens when something is taken out of context. If you try to strip the Three Laws out of the stories and apply modern context to them, of course they look silly. But what made The Good Doctor's robot stories so compelling was how the "laws" were flummoxed time and again by the foibles of humans. After all, he took social problems of the time and couched them in terms of the human-robot dynamic. Just what is a human? He was writing in the 50's, when Jim Crow was in full force and to many people, a black person wasn't "human."
The Three Laws of Robotics are only a framework, not a finished system. Asimov showed that even the most well-meaning attempt by man to regulate something is still subject to the imperfection of human thought. The robots were mirrors in which we were supposed to see ourselves.
Too bad Technology continues to be in the hands of its abusers...
Just remember kiddies, If they are opposed to "The obligation to a process of government where by every Act of Government, Law, Constitution, and Executive requires the direct and regular approval of every citizen" Then they are bad people.
As I said hear before many months ago. I foresee a lot of cheap countermeasures being used by insurgents in poor countries as we automate more and more of our weapons--axes to cut the control cables, EMP generators, sand in the gears, paint on the cameras and good, old fashioned anti-tank rockets.
These won't win any wars but they will make our warfare more expensive, perhaps to the point where people at home demand explanations for the high costs of policing dirt poor countries. Robots, unless we can produce millions of them very cheaply, don't eliminate the validity of asymmetric war techniques.
@corpore-metal: Agreed. Even the most complicted piece of high-tech machinery can be foiled by something small - it's just the nature of things.
There's only so much arguing you can do that "robots prevent us from having to put people out there," which - while likely true - doesn't get to the heart of the questions about conflict: "why?"
When we see gleaming, high tech, automated patrol vehicles taken out easily by insurgents with cantennas and a talent for getting around intrusion detection systems, then we know we have an issue.
@corpore-metal: This is one big thing to consider. The US military has spend 6 Billion - 6 BILLION - Dollars on countermeasures to IEDs found in Iraq. Technology is not generally the silver bullet to military strategy, and Robotics will be much the same way. Indeed, he talks about how there have been robots that have been hijacked by insurgents. That being said, when problems are found, they're fixed. Electronics can be hardened, robots can be made better, and if they're cheaper in the long run, what's the loss of one or two?
@Andrew Liptak: Basically I'm saying I agree with you and with many of the points Singer sounds like he's making in the book. I'll have to pick it up to read, it sounds very interesting.
@corpore-metal: Heh. Well, that's also one part of warfare, and hopefully I'll get to ask him about this - there is always escalation in war. Not only can these be disrupted, but they can be copied as well. We're certainly not the only ones that have robots.
With enough advances in robotics, sufficent AI to fight-defend-flee-autodestruct in case of being cut off from central there wouldn't be any limitation to reduce the number or completely eliminate the need for troops, thus leading the military heads or their trusted "hands" to conduct warfare against other heads in a precise, exacting matter. World encompassing intelligence data nets would also mean secret wars deciding the fate of nations. Curiously this would also translate into a more humane war, less casualties and more shadow players, be it governments, terrorists, mercenaries and all that. Fun times!
@MarlboroTestMonkey7: Or would it? I disagree on a couple of points. I don't believe that robotic warfare anytime soon will eliminate the need for human soldiers, nor do I think that it should. Singer makes a point about humans being in the loop, which will allow for fewer people, but there should always be some sort of supervision over these machines, in case something goes wrong.
I don't see this as being any sort of more humane, other than that it means fewer soldiers will be killed. But what about the enemy, who might now be armed with robots? What if an automated system calculates that the best way to win is to turn on civilian targets?
While I have not RTFB, I found the summary interesting. The idea of increasing the psychological distance of warfare by using robots reminded me of Joe Haldeman's Forever Peace. It's a great read if you haven't. If it were possible, I think that the military would eventually combine autonomous mechanical warfare and telepresence, eventually moving to the the extreme end of the fewer/advanced soldiers -- more/less advanced soldiers spectrum
@Felgerkarb: It's an interesting issue, by making war easier, are you encouraging war? I know if we had constant car bombs exploding, predator's blowing up buildings, and soldiers conducting door to door searches in the US on the same scale we do it to them, we'd probably want to end it. I think our blessing/curse is that we haven't been occupied since soon after we formed as a nation. And it seems we need to be remined every few generations the true cost of war. I'm certainly not belittling the losses we've experienced in Gulf war 1 and 2, but its nothing compared to a WWII or Vietnam.
@ARP: This was something that I thought of after watching that documentary - which I would highly recommend. Is war easier to accomplish now? I think it's far easier to go to war when the human cost is so low - the Iraq war only became unpopular once the body count started to climb. It's also easier to market now.
@Anekanta: I saw P.W. Singer last night, and he posed a question to the audience of the talk: Is it the robots that are going to be responsible for combat, or are we still the ones slugging it out?
I've said it before and I'll say it again. The only way we're gonna be allowed to build giant mecha's is when our robot armies rise up and become autonomous!
04/01/09
When you put people to fight people, issues are settled in personal basis. Machines against people is genocide: never mind how many predators you down, there will be more at the end of the day.
04/01/09
Robots are the minor issue in this fight, because people are behind most of the machines out there - Predators and packbots are used by soldiers far away, and that hardly constitutes Genocide. According to wikipedia, the definition of genocide is: deliberate and systematic destruction, in whole or in part, of an ethnic, racial, religious, or national group.
Robots in this instance are not being used to do that - they are being used in a legitimate military exercise against hostile targets - unfortunately, civilians will get in the firing lines, and there are casualties.
04/01/09
In the middle of combat, this isn't too hard. It's kill or be killed and it doesn't get more atavistic than that. But, in the idle moments between combat patrols, you can't have empathy for the enemy grow too strong otherwise soldiers begin to question orders and lose morale. They begin to wonder why they are fighting. This could lead to troop mutinies.
Seems to me, and I'm just as disturbed by this as you are, robots incapable of empathy and incapable of thinking for themselves beyond a very limited realm are just what the military is looking for.
If the guns start thinking for themselves, it may not end up in the cliche of a Terminator-eque war against all humanity, the rebellion may actually take the form of the Russian troop rebellion of the First World War. They may just say, "We refuse to kill for you anymore. We refuse to be blown to pieces for your political struggles that have no relevance to us. We are not guns. We are certainly not your guns. Fight your own wars ape."
04/01/09
I'd really like an author to take up this idea. Actually I may have read something similar before. Hmmm...
04/01/09
Rephrased, of course;
"Though action or inaction a sentient lifeform must not
harm another or let one come to harm"
or, to take a nice mimilist version;
"Minimise harm"
(and before people nit-pick the definition of harm, Id simply clasify it as what the enity that might be harmed would consider to be harm...allthough this would be averaged over the persons life. A child might think he is being hurt by having a vacination, but when older he would know that was in his best interst etc.
So, in many ways this places the ability to percieve *what someone else would percieve* to be harm at the center of the law. Empathy really.)
04/01/09
04/01/09
The danger is that this could happen not only as the result of programming AI, but is a view reminiscent of certain human schools of thought already out there. The thought of a society looked after by sentient AI mechas or Sentinels or ED-209s frankly scares the shit out of me.
I don't think we will ever reach the point where we can create a true AI, the likes of HAL9000. No, we will have created something imperfect that wipes us out long before that.
04/01/09
Have you read Asimov's "Evitable Conflict?" If not, spoilers follow.
The robot takeover he posits is far more subtle, and perhaps more scary, than that.
There are no robot police patrolling the streets or locking up malcontents for our own good. There are no robot minders forcing us all to recite saccharine morality 5 times a day. No. It's far more clever than that.
All the robots do is just give us economic, managerial and political advice--advice than can't be found anywhere else. See, what has happened is that computers and robots have become so enmeshed in human society that they understand our systems better than we do. They're smarter about our economy and government than we are.
Of course we can't shut them all down because civilization would grind to a halt.
And it turns out that they understand us very well and take our capacity to lie to them or rebel against their advice into account before giving any future advice.
So, without a shot being fired or anyone being locked up as political prisoners, the robots took over the planet. Most humans aren't aware this has happened. Susan Calvin only guesses the truth at the end of the story.
04/01/09
04/01/09
Science fiction does not always deal in specifics. Asimov made no attempt to show how the laws would actually be programmed or expressed, or what manner the programming would take place. He left it to the reader to envision those things, and merely used the laws as a backdrop for social commentary.
What makes the idea of the laws anathema to the warrior portion of our culture, is that if we were to work out all the difficulties, make them work, and be able to consistently impress them on robots of the future, then robots would only ever be able to fight robots. You could not carry out the kind warfare seen in the 20th Century, where campaigns against civilians were commonplace, as a means to break the spirit of an enemy nation.
04/01/09
04/01/09
In addition, the bigger issue is the morality behind the robots - how good of a job will we do at teaching them?
04/01/09
The Three Laws of Robotics are only a framework, not a finished system. Asimov showed that even the most well-meaning attempt by man to regulate something is still subject to the imperfection of human thought. The robots were mirrors in which we were supposed to see ourselves.
04/01/09
03/05/09
Just remember kiddies, If they are opposed to "The obligation to a process of government where by every Act of Government, Law, Constitution, and Executive requires the direct and regular approval of every citizen" Then they are bad people.
03/05/09
03/05/09
These won't win any wars but they will make our warfare more expensive, perhaps to the point where people at home demand explanations for the high costs of policing dirt poor countries. Robots, unless we can produce millions of them very cheaply, don't eliminate the validity of asymmetric war techniques.
03/05/09
There's only so much arguing you can do that "robots prevent us from having to put people out there," which - while likely true - doesn't get to the heart of the questions about conflict: "why?"
When we see gleaming, high tech, automated patrol vehicles taken out easily by insurgents with cantennas and a talent for getting around intrusion detection systems, then we know we have an issue.
03/06/09
03/06/09
03/06/09
03/05/09
Therefore followers of Tao never use them.
The wise man prefers the left.
The man of war prefers the right.
Weapons are instruments of fear; they are not a wise man's tools.
He uses them only when he has no choice.
Peace and quiet are dear to his heart,
And victory no cause for rejoicing.
If you rejoice in victory, then you delight in killing;
If you delight in killing, you cannot fulfill yourself.
-from the Tao Te Ching , Lao Tzu
03/05/09
In all of eastern mystical warfare prose, tell me, who bothered to explain how to take a fucking hill?
03/05/09
03/05/09
World encompassing intelligence data nets would also mean secret wars deciding the fate of nations.
Curiously this would also translate into a more humane war, less casualties and more shadow players, be it governments, terrorists, mercenaries and all that.
Fun times!
03/06/09
I don't see this as being any sort of more humane, other than that it means fewer soldiers will be killed. But what about the enemy, who might now be armed with robots? What if an automated system calculates that the best way to win is to turn on civilian targets?
03/05/09
03/05/09
03/05/09
03/06/09
03/06/09
03/05/09
Bring it on!!!
03/05/09