Looks as if researchers have made another step towards taking Skynet live: giving robots the groundwork for gloating. A Swiss team of misguided geniuses have developed learning algorithms that permit robot-kind to be told from human mistakes. Earthlings guide the robot through a flawed attempt at completing a role, akin to catapulting a ball right into a paper basket; the machine then extrapolates its goal, what went wrong within the human-guided example, and the way to succeed, via trial and error . Instead of presuming human demonstrations represent a role well done, this new algorithm assumes all human examples are failures, ultimately using their bad examples to assist the ‘bot one-up its creators. Thankfully, the hot algorithm is just getting used with a single hyper-learning appendage; heaven forbid it may ever the best way to use the robot-internet.
The Engadget Interview: BlackBerry PlayBook product manager Michael Clewley
Mozilla rumored to debut LG-made Boot to Gecko device at MWC
