To Front Page   >     >   You Are Here

Subscribe:   XML icon     Add this feed to your Bloglines account

Sat - November 1, 2003


A Robot May Not Harm A Human Being. Is The Opposite True? 



"A robot may not injure a human being, or, through inaction, allow a human being to come to harm."

—Isaac Asimov, "I Robot", First Law of Robotics 

Can a robot think? Can a robot cry? Can a robot long for a better, more satisfying existence? Do the answers to these questions influence whether robots should have legal rights?

These are the questions posed in a thoughtful article by Glenn Reynolds (yes, that Glenn Reynolds) on Tech Central Station the other day. With some experts predicting that sentient, humanoid robots are "as inevitable as airplanes" by 2055 and that we will have "robots like C-3PO walking around and filling jobs as early as the 2030 time frame," it's not too early to ask: Are robots people too?

Turns out, that's a surprisingly difficult question to answer, not only because it asks us to define the source from which such rights spring but also because it challenges our preconceived bias that human beings are unique. For example, one theory advanced to support animal rights is that animals feel pain. But surely that isn't the reason we grant some animals rights. Cats and dogs react when people accidentally step on their tails. Cows react when exposed to cattle prods. Fish react when hooked or netted. All three animals respond to painful stimuli, yet we treat their "rights" in these situations very differently. To say that a sentient robot might react to pain doesn't tell us whether the robot should be accorded the relatively broad legal protections given pets or the modest ones offered fish.

We might also grant sentient robots rights because they have a "sense of self" or can demonstrate independent thought. But again, this compels us to make uncomfortable distinctions. Certain animal species, such as chimpanzees, display complex emotions, communicate using language, and demonstrate a sense of self. They use tools, solve problems, and even have a social structure—yet our society routinely uses them in scientific experiments. This is so even though an anencephalic child born without a brain is accorded a plethora of rights. Is a sentient robot like the child or the chimp? Are the robot's rights superior to theirs in any circumstance?

For human beings, rights are frequently relative—conditioned upon the nature of a legal relationship. A child's rights differ from an adult's rights. A landlord's rights differ from a tenant's rights. And so on. But how do you discuss rights for sentient robots without first deciding that robots are more than mere property? Your car today has no right to determine where it is going or how early in the morning you'll start it, and you probably like things that way. The upshot of giving robots rights is that such rights will one day prevail against a human being's contrary desires. But if robots—as mere human creations—have no rights at all, then we'll have to explain why creating sentient slaves is acceptable to us. Or why a test tube baby, created in a lab, is fundamentally different from a sentient robot, created in a factory.

These will not be easy issues, intellectually or emotionally. We're not a species comfortable with sharing our privileged status. Yet perhaps in facing the complex challenge of defining rights for artificial intelligence, we'll learn to better appreciate a human quality that truly unique makes us unique.

We're the only known species with a code of morality.

 

  To Front Page     |   Email This  



©