Despite advances in artificial intelligence (AI) research, a robot that can pass the Turing test has yet to be developed. The development of human-like intelligence has eluded AI researchers because it involves skills, such as perceiving the environment and incorporating past experiences into decision making, that machines are still not good at.
Humans can recognize objects and distinguish them from one another even in different lighting and on different backgrounds. However, it takes a massive amount of computing power to execute this innate human ability, but computers are beginning to gain skills that bring them closer to human-like intelligence. For example, Google is developing an omnivorous search box that can recognize images and sounds recorded on a smartphone.
Meanwhile, the U.S. Defense Department is working with University of Michigan professor John Laird to develop a military robot that can monitor dangerous areas before soldiers move in. Humans will train the robots in advance by walking them through model buildings. At the Massachusetts Institute of Technology, engineer Victor Zue is developing life-like digital human images that work in airports to tell passengers if their flight is delayed.
From The Washington Post
View Full Article
Abstracts Copyright © 2010 Information Inc., Bethesda, Maryland, USA