When robots outsmart us
A group of computer scientists look at what will happen when artificial intelligence tops human intelligence
Our artificial intelligence may soon be smarter than us, said John Markoff in The New York Times. That concern, and fears of a backlash against the idea of “smarter-than-human machines,” has the Association for the Advancement of Artificial Intelligence considering self-imposed limits on new technology that could “run amok”—fully autonomous killing drones, say, or unstoppable computer viruses. Not everyone envisions a future “technological utopia.”
The pervasive “fear-filled skepticism of technology” is more of a threat than any robot we could build, said Joe Windish in The Moderate Voice. Let’s face it: “Our machines are already smarter than us,” which is why we are “completely dependent on them to do much of our computation for us.” There’s bound to be some danger and social disruption, but “hey, that’s life.”
Getting rid of human traits could even be useful sometimes, said Phil Bowermaster in The Specualist, like in police work. Yes, the “idea of ‘robocops’ might sound a little scary,” but think about how the Henry Louis Gates Jr. incident would have gone with a robot cop on the beat—accusing a robot of “racial profiling” would “make about as much sense as accusing my lawnmower of sexually harassing the women in my neighborhood.”
Don’t kid yourself, we have “some pretty scary scenarios brewing,” said Ian Gray in Dissociated Press. Two “respected scientists” just published an article arguing that “a self-aware Internet may already exist,” and military scientists are creating “corpse-eating robots” (though they claim the machines are vegetarian). No wonder the AAAI is worried.