Rosie, the Jetsons’ maid robot, is a sweet, nurturing cartoon robot. Not only does she bring George his slippers, she washes his clothes, teaches his son to dribble a basketball, and sings while vacuuming the rug.
The Stanford Artificial Intelligence Robot (STAIR) looks more like tubular shelving on a Segway than a robotic maid. It finds objects with its stereoscopic camera eyes and grabs them with a robotic arm. Perhaps not unlike an early model of Rosie, STAIR can interpret relatively ambiguous vocal commands, navigate around unfamiliar environments and objects, and solve problems.
“STAIR, please fetch the stapler from the lab,” says a researcher in a recent video. “I will go get the stapler for you,” replies STAIR. Avoiding obstacles, STAIR wheels into the next room and scans it looking for the stapler. Grabbing the stapler, it returns to the researcher. “Here is your stapler,” says STAIR, “Have a nice day.”
h+ contributor Ben Goertzel, an organizer of this year’s Second Conference on Artificial General Intelligence, characterizes general intelligence as “the ability to solve a variety of complex problems in a variety of complex environments.” STAIR shows the evolutionary transition that is occurring in artificial intelligence today — from the narrow AI of expert systems to more generalized intelligence. As the STAIR video demonstrates, the multitalented STAIR walks, talks, sees, hears, and solves problems in an obstacle-laden lab environment.
Andrew Ng, the assistant professor of computer science at Stanford who led the development of STAIR, is optimistic that the many disciplines of AI are now mature enough to be integrated “to fulfill the grand AI dream.” And no, this is not just a robotic maid to fetch staplers or slippers, but rather computers that are as Intelligent as people.