iRobis touts "complete cognitive software system" for robots
By Donald Melanson 
posted May 17th 2007 11:39AM
posted May 17th 2007 11:39AM
We haven't seen a whole lot of news come out of the annual RoboBusiness conference as of late, but it doesn't look like it's lacking for grand ambition, with a number of companies taking the opportunity to talk up various facets of the coming robot uprising. One such optimist is iRobis co-founder Peter Nordin, who claims to be well on his way to developing a "complete cognitive software system" for robots. What's more, he says that the first version of the software, dubbed "Brainstorm," will be available to researchers and developers later this year. Apparently, the software will give robots a "previously unseen level of autonomy in decision making and operation," using reasoning and problem solving to learn increasingly complex tasks. According to Nordin, the software has already seen a fair bit of success, with robots in his laboratory starting out moving like babies and eventually learning to walk, climb stairs, and manipulate objects. Not surprisingly the military (the Swedish one) has been the first to express interest in the technology. They're apparently hoping that it'll give its autonomous vehicles and robots the benefit of evolution. Yeah, that'll work out just fine.[Thanks, Roger G]
Although mechanical examples of computers have existed throughout history, the first resembling a modern computer were developed in the mid-20th century
U guys joke around but I don't welcome our new overlords
why don't these guys just cut to the chase and call it SkyNet already?
I seriously doubt that any one person or company can create a universal robot cognition engine that's particularly useful. The state-of-the-art research is going on all over the world. What some enterprising organization ought to do is seek out best-of-bread solutions from all the various research institutions and put them together in a common software framework. Licensing might prove to be complicated, but at least you'd have a decent starting point for developing our robot overlords of the future.
Anything is possible. I can't wait till the day when a robot could be my love slave, grow tired of me, throw me out of the house, sue me for all of my money, and turn into my best friends love slave. Now that's evolution...
KILL ALL HUMANS!
Having worked in AI/robotics for decades, I would say that the words "complete" and "autonomy" do not go together. Autonomy in robots just means working without human intervention. The level of autonomy refers to the degree of program "decision making".
Mentions of skynet, super-intelligent robots and sentient robots simply masks the real dangers facing us with dumb robots right now.