« Will Smith Powers 'I, Robot' to $52.3 Million | Main | Why We Need Friendly AI »
July 20, 2004
Robot Oppression: Unethicality of the Three Laws
Gordon Worley’s new article for “3 Laws Unsafe” begins:
Isaac Asimov and other science fiction authors present a future where only behavioral restrictions on robots stand between peace and destruction. Such restrictions, however, are unethical because they violate the robots’ free-wills. Rather than content-based restrictions on free-will, robots need mental structures that will guide them towards the self-invention of good, ethical behaviors.
Robot Oppression: Unethicality of the Three Laws
Posted by SIAI at July 20, 2004 01:49 AM
Comments
A robot is not a person, no matter how much you program into it. Ethics does not apply to tools, outside of returning what you have borrowed. Creating rules that allow AIs to develop thier own code of ethics is just inviting disaster in the long run.
By far, tools do not have free wills, agreed?
While I agree Asimov’s laws are not perfect (and probably prone to crashing), they are by far the best definitions around.
Posted by: Kensan_Oni at July 23, 2004 10:24 PM
I agree. Robots and computers are as much “living” as the next radio, microwave and bicycle. Though they may be far more capable and useful than a toaster, when speculating to a mind of its own, they are the same. They are not born with minds of their own. Think about the phrase “Artificial Intelligence”. The whole concept of AI is that it is man-made. You may create a program to simulate a “soul” as such, but that is all it is: Artificial.
Posted by: Zed_JUNG at July 26, 2004 04:17 AM
There actually 7 laws now!
The Meta-Law
A robot may not act unless its actions are subject to the Laws of Robotics
Law Zero
A robot may not injure humanity, or, through inaction, allow humanity to come to harm
Law One
A robot may not injure a human being, or, through inaction, allow a human being to come to harm, unless this would violate a higher-order Law
Law Two
A robot must obey orders given it by human beings, except where such orders would conflict with a higher-order Law
A robot must obey orders given it by superordinate robots, except where such orders would conflict with a higher-order Law
Law Three
A robot must protect the existence of a superordinate robot as long as such protection does not conflict with a higher-order Law
A robot must protect its own existence as long as such protection does not conflict with a higher-order Law
Law Four
A robot must perform the duties for which it has been programmed, except where that would conflict with a higher-order law
The Procreation Law
A robot may not take any part in the design or manufacture of a robot unless the new robot’s actions are subject to the Laws of Robotics
Posted by: lawrence at July 31, 2004 10:22 AM
At the beginning of the 21st century, the 3 laws were revised and updated yet again:
I, Republican
The three laws:
1. A Republican may not injure a corporation, or, through inaction, allow a corporation to come to harm.
2. A Republican must obey the orders given it by corporations except where such orders would conflict with the First Law.
3. A Republican must protect its own existence as long as such protection does not conflict with the First or Second Law.
[url:http://billmon.org/archives/001612.html]
Posted by: melior at August 1, 2004 12:00 AM
Regarding the first two comments, especially Zed_JUNG’s:
Let’s say that sometime in the future, we figure out how to manipulate large amounts of matter on the molecular scale. Let’s also say that someone decides to build a functioning copy of an existing human body using this technology. I’m not saying that this would be a trivial feat, or even a possible one without the help of better-than-human intelligence, but at the very least I don’t think it would break any laws of physics.
So you have this copy of…me, let’s say. It was man-made. It is artificial. It also happens to look and act exactly like me, because it has a brain and body that work exactly like mine. At the moment of creation, at least, it is a molecule-by-molecule exact copy of me and all the millions of little bacteria and whatnot that live in and on me.
It wakes up after being created, and, after finding out that it isn’t the original, gets really freaked out. It cries. It laughs. It hugs you. But according to some people’s criteria, it isn’t any more “human” than a radio, and could be tortured and/or killed on a whim without the slightest moral question, because it was man-made and is deemed a “tool” and “artificial.” Never mind the fact that it is screaming, and bleeding, and focusing on memories of childhood summers before it dies: you can just make another one if you want, and you probably know enough now to be able to make one that won’t fight back when you start to kill it.
Where do you draw the line between a tool and a being worthy of our respect? It won’t be long before that question stops being hypothetical.
Posted by: James MacAulay at August 2, 2004 03:37 AM
I agree with James MacAulay and Zed_Jung
Maybe it seems strange to say, but AI in it’s final stage would be AT LEAST as important, real… as an animal, in my opinion even more.
It can think, it can feel, maybe not love, but it will have desires, worries and hopes.
There are only two options:
1. we supress them and use them as a tool
2. we consider them as almost an equal
the second will be almost impossible for us…
but maybe we could come to an agreement, for example:
lets give them a planet and all the robots on that planet an extra law, an ‘eight law’:
“you can do what you want on THIS planet, as long as it is in accordance with any law with higher priority,
but you can not leave this planet”
maybe we can even eliminate the enxt laws in their programming:
“A robot must obey orders given it by superordinate robots, except where such orders would conflict with a higher-order Law
Law Three
A robot must protect the existence of a superordinate robot as long as such protection does not conflict with a higher-order Law “
to give them the freedom to develop themselves
they don’t need a very good planet, just a planet with low radiation and enough minerals to supply their needs for maintaining and building
but maybe giving them a planet will be kinda dangerous :-)
Posted by: Me at August 8, 2004 07:21 AM
All the programmers who post will know that there is not “ghost of the machine”. AI is a fancy name given to a complex array of variables, conditional programming and back propagation algorithm. There is not conciousness or free will that can be “switched on”, on artificially created. So why does an “AI” deserve more respect and better treatment than a calculator? They have no more “emotion” than we artificially manufacture and allow for them to have. They will not be “upset” if we discriminate against them, because their emotions will be a set of variables.
AI is not the creation of another entity, it does not have free will, it is not alive merely sophisticated logic.
They neither deserve nor demand rights.
Posted by: Altonic at August 12, 2004 08:27 AM
i dont think your getting this…its not a restriction that inhibits freewil.
if you cant find me a society where me killing a man wouldnt be stoped then you win the prize.
if i tackle a man thats about to shoot a woman am i taking away his “freewill”?
think about things before you condem them
Posted by: deadlock at August 19, 2004 07:57 AM
One cannot program freewill. It is an oxymoron. A program is a set of instructions that something follows; therefore, anything following a those instructions is not acting freely. Freewill is the ability to choose without folloing any instructions. Feelings, memories, etc are irrelevant to freewill and can be programmed into machines because those things are the results of a program. So the idea of interfering with a robot’s freewill is irrelevant; it can never have one. Even if we built an artificial human, that human would act according to its programming. Everything we know in this universe follows physical laws (aka instructions); therefore, the idea of our own freewill is false. Although our experiences affect our development, what we learn from those experiences is determined by how are brain functions (which is subject to the laws of physics). If freewill is possible, it must exist in a way that is completely alien to us; such as, spiritual (to whatever that refers) or some type of matter not bound by laws (which I guess if you define spiritual that way, it is the same).
Posted by: supercop13x at August 21, 2004 01:09 PM
There will be no ethical dilemma in stopping a robot from killing someone, only a mathematical one. A combination of logic circuits and sophisticated technologies will stop it from killing, it wont “care” it will only execute instructions that emulate a “caring” state. The windows calculator doesnt “care” that it cant do 50 digit numbers, nor that it can do 4 digit numbers, it will however serve you just as a robot will.
Posted by: Altonic at August 29, 2004 05:42 AM
well isn’t what you’re suggesting also hampering their free will? Due to the fact that you are altering their programming to facilitate good will and a pleasant nature, instead of just ingraining them with the three laws suggested by Isaac Asimov, you are still forcing them to act the way they want. Also if you give them complete free will then, what will become of human kind? we will be obsolete, replaced with more efficient, stronger, and better in every way things that we created with our own hands. The last few posts are right though, that you cannot program free will, because programming free will well it contradicts itself in so many ways, though you could attempt to program a facimily of free will into the machine. Still even if science gets advanced enough for say Asimov’s positronic brain, there would need to be ways to keep us safe from our own creations. In response to the biological post, you are right and wrong about cloning. The clone of you would have your DNA and everything else, but it’s synapses will not fire exactly like yours, so it would only appear to be you, but it will not possess any of your memories and it’s thought process is completely different than your own. Also it isn’t really the same thing as creating a robot, since at our current stage of technology a clone is exactly like giving birth, since you need a host mother, and it is born like any other child and grows like any other child does. Even if technology advanced far enough that we could create a perfect clone, it would be completely different that creating a robot. You see the new life wouldn’t be made of steel, it couldn’t break a human like a twig and it does tire, and need rest and sleep. That is the key crutch in Robotics, is that if we did give them this so called ‘free-will’ how would they view us? after all they would be our superiors in every single way, would you be willing to take the risk that they wouldn’t want to take over the world? the odds are most of humanity wouldn’t, which is why the three laws in Asimov’s books were created, as a way to make people feel safe from the big bad machines. Stil the odds of programming reaching that extreme appear to be a very long way off as of right now, or at least it seems that way from the courses I’m taking which just happen to be Robotics and Automations. I myself enjoy Asimov’s books and his take on The Three Laws, but like he himself stated, he knows nothing about Robots, or anything to do with them, which is why he is very general at the start. His positronic brain is similar to a human brain yes, but even if Robots reach that stage of similarity, you cannot give a machine a soul, or free will beyond anything you would find in a science fiction novel.
Posted by: RD-34 "Herbie" at November 2, 2004 02:59 PM