The nice thing about the weekend is that you have a chance to get caught up on all that reading you have put off the rest of the week. For example, Monday through Friday, I would probably just scroll on past a story with this headline: "Can Robots Make Ethical Decisions?" Not today. My weekends may at one time have been made for Michelob, but now they are much better suited for this kind of high-minded speculation.
It only took the author of this article a paragraph and a half to bring up both Isaac Asimov's Rules of Robotics and Blade Runner. Interesting, since the actual focus was a recent paper published in the International Journal of Reasoning-based Intelligent Systems describing a method for computers to prospectively look ahead at the consequences of hypothetical moral judgments. The authors of the paper, Luís Moniz Pereira of the Universidade Nova de Lisboa, in Portugal and Ari Saptawijaya of the Universitas Indonesia, claim that philosophy is no longer exclusively for human beings. These two science-types used a series of events called "the trolley problem" to explore a pretty standard moral dilemma: is it permissible to harm one or more individuals in order to save others?
Mister Spock, perhaps the most seamless interface between man and machine, has stated in many different realities that " the needs of the many outweigh the needs of the few." If this is true, then we should expect any good and moral machine to come up with the same response. Roy Batty got it too. HAL didn't. Of course, these aren't Predator drones or the new, fully automatic H-II transfer vehicle. As much as they owe to science fiction, they are here now. I can't help thinking that we would have been better off starting out with questions just a little lower on the moral ladder, like a TV tuner that could be quizzed about the choice of recording reruns of "Hawaii Five-O" versus old episodes of "Mannix." Come to think of it, that's exactly the kind of moral dilemma that I would like a machine to help me with.
No comments:
Post a Comment