The prospect of intelligent robots is one that ignites a wide range of emotionally charged reactions in people. Thousands of movies, television, novels and other methods of science fiction story-telling, include cautionary tales of the robot who knows too much. The story is always the same. People create robots and treat them like slaves. The robots then become too smart and rebel against the human race. This is a scary, if not far-fetched, scenario. For one thing, robots are typically far more strong and less puncturable than human beings. And yet, scientists keep pushing the envelope, suggesting fear isn't the only emotion humans feel towards intelligent robots.
Among the scary doomsday depictions of human-like robots, such as Terminator, Battlestar Galactica and the like, are stories that show us a different possibility for the future of robots. Robots in stories like Iron Man and Short Circuit depicts a future where robots can help humans stay alive, and even be our helpful friends. A new discovery in Japan may be the next step in getting closer to unleashing the mystery of what society would be like with humanoid robots among us.
The first two definitions that show up for robot are "A machine capable of carrying out a complex series of actions automatically," and "(in science fiction) A machine resembling a human being and able to replicate certain human movements and functions." According to the first definition, our modern society is filled with robots. Most people in the first world have at least one robot in their homes, such as a computer or a smart vacuum, but we don't really associate our robots with the robots found on the Jetsons and Star Wars, not to mention iRobot.
"So far, robots, including industrial robots, have been able to do specific tasks quickly and accurately," explained Imaging Science and Engineering Lab's Associate Professor, Osamu Hasegawa, from the Tokyo Institute of Technology. "But if their environment changes slightly, robots like that can't respond," he adds. The toughest barrier researchers face in programming robots, has been teaching them to adjust to new problems and environments. They can be programmed to solve one or two specific problems, but it has been difficult to teach them to try to solve new problems based on what they know and don't know. The technology developed by researchers in Japan involves a "Self-Replicating Neural Network." This network enables robots to think, much like humans. It allows them to assess problems as they are encountered and then work to solve them.
Although these robots now have some problem solving abilities, the technology is still very new and limited. "This robot remembers only basic knowledge, and it can apply that knowledge to its immediate situation," explains Hasegawa. The key to advancing this technology is in the network. The current plan is not to expand how much the individual robots can remember, but to eventual connect all of these robots, used for human tasks, to one neural network. The theory is that as one robot learns something new, all of the robots connected to the network will then have access to the new information.