Slashdot Mirror
Ethical Killing Machines
ubermiester writes "The New York Times reports on research to develop autonomous battlefield robots that would 'behave more ethically in the battlefield than humans.' The researchers claim that these real-life terminators 'can be designed without an instinct for self-preservation and, as a result, no tendency to lash out in fear. They can be built without anger or recklessness ... and they can be made invulnerable to ... "scenario fulfillment," which causes people to absorb new information more easily if it agrees with their pre-existing ideas.' Based on a recent report stating that 'fewer than half of soldiers and marines serving in Iraq said that noncombatants should be treated with dignity and respect, and 17 percent said all civilians should be treated as insurgents,' this might not be all that dumb an idea."
Ethics" is such a poorly defined term...hell, different cultures have different definitions of the term. In feudal Japan, it was ethical to give your opponent the chance for suicide...today, many Westerners would in fact argue the opposite: the ethical thing to do is prevent a human from committing suicide as that's seen as a symptom of mental illness.
I've always defined "morality" as the way one treats oneself and "ethics" as the way one treats others. It's possible to be ethical without being moral--for example, I'd consider a person who spends thousands of dollars on charity just to get laid to be acting ethically but immorally. By that definition, the hullabaloo at Guantanamo would certainly be both immoral and unethical--not only were they treated inhumanely, but it was done against international law and against the so-called "rules of war".
These robots would have to be programmed with certain specific directives: for example, "Don't take any actions which may harm civilians", "take actions against captured enemy soldiers which would cause the least amount of forseeable pain", etc. Is this good? Could be...soldiers tend to have things like rage, fear, and paranoia. But it could lead to glitches too....I wouldn't want to be on the battlefield with the 1.0 version. Something like Asimov's 3 Laws would have to be constructed, some guiding principle...the difficulty will be ironing out all the loopholes.
Ride the skies
Actually (according to every philosophy book i've ever read), morals are codes of conduct, and ethics are is more ethereal "right and wrong" concept. The problem is that 'ethics' has been watered down to mean 'morals' because 'business ethics', etc. roll off the tongue more easily than 'business morals'.
Exactly. Robots, machines, whatever you want to call them, do not have ANY moral substance. Humans do. Humans may refuse to do certain things (or may not). Machines won't refuse.
Bottom line is... I'd rather be up against convincing a human maniac than a robot programmed to ACT like a maniac. One still has a rational (hopefully) thought process somewhere in there, and has a moral element. The other can't think and has has no moral element whatsoever (partially, of course, due to not being able to have a rational thought in the first place).
There's a reason that "mind control" scares so many people. Total "mind control" is what you have over machines, is it not?
AHA! So! How is this any different than humans?
Actually, from a psychology angle, it's substantially different. It has been shown many times that humans are psychologically capable of stretching their moral limits further when they can distance themselves from the action (If you want me to get a citation, I'll go get one -- I'm just too lazy to get it right now).
This is easy to see even without evidence. If you were forced to choose, would you rather push a button that drops a bomb on a village full of children 1000 miles away, or be in a plane and drop the bomb yourself? The two actions have identical results, yet distancing yourself from the action makes it easier to justify the moral consequences.
This is why military leaders are able to stay sane. It's possible (though not easy) to give orders that will directly result in the death of thousands of people. However, if a war general had to shoot thousands of people himself, I suspect it would start to wear down on his psychological health.
Now consider that you're a military general who simply has to push a button, and this button tells your robot to take over a village. It's very, very easy to rationalize that any casualties are not your fault, since all you were doing was pushing that button.
I disagree as well.
Your theory that morals are culturally-based or a standardized code of conduct is just one theory of ethics -- constructivism or something like that. Ethics as a branch of philosophy is also called moral philosophy. Ethics in this sense is the study of the status of morality, of morals.
Pretty much in every modern piece of ethics, "moral" and "ethical" are synonyms. In Hume and Hume scholarship, though, "moral" frequently means "absolute," as in "moral certainty," as well meaning the regular "ethical." For example go to the end of Mackie's "Subjectivity of Values" (this is a very popular paper in ethics). Here he uses the terms as synonyms is the same sentence: "...the central ethical concepts of Plato or Aristotle are in a broad sense prescriptive...they show that their moral thought is an objectification of the desired and the satisfying." So moral thought is thought dealing with ethical concepts -- just as anyone would say that moral thought is thought dealing with moral concepts. And here he is saying that when Plato or Aristotle use ethical concepts, that is, speak of good things or actions, they actually are referring to things or actions that are connected to their subjective feelings of pleasure, but they objectify these things in language as "good." So Mackie is saying that ethical concepts are not-universal at all, but are merely expressions of desires of certain people -- this is precisely what you denied. Notice also that Mackie says that the ethical concepts are prescriptive -- which was the second feature you ascribed to morals as distinguishing them from the matter of ethics.
TL;DR:
If you ask a philosophy professor what Kant's moral theory was, he'll tell you about the Categorical Imperative. If you ask a philosophy professor what Kant's ethical theory was, he'll tell you about the Categorical Imperative. The word is just not regularly distinguished in the topic of ethics. The reason "ethically" was originally used in this story was probably for the simple reason that "more morally" sounds weird in its phonemic repetition.