The idea of machines revolting against their creators is the stuff of great science fiction: Skynet, HAL, and AI are a few terms that should be familiar and unsettling. Here at Botropolis the robot apocalypse is always at the back of our minds, but now scholars and manufacturers are also beginning to take the potential threat posed by robots into consideration.

“By 2015, personal robot sales in the U.S. will exceed $5 billion, more than quadrupling what they are now, according to ABI Research, which analyzes technology trends.” These robots of the next decade won’t just be Roomba-ing around the living room inhaling pet dander, they will be doing many different tasks, expanding their scope and visibility in general.

With the number of robots growing, so too will grow the number of robot-related accidents. If your toaster short circuits and starts a fire, the manufacturer seems liable. But what if your robotic toaster is programmed to start a fire and it torches your heavily insured establishment?

Legal scholars and the scientists building robots are beginning to think about what kind of ethics should be programmed into robots that interact with humans. Let’s not forget Asimov’s law that robots should not harm. That’s a pretty good starting point. If your robotic lawnmower chops off your leg is it your fault for not being more careful around a lunatic robot? Or should the robot be built so that every scenario for leg removal is avoided? As robots get more sophisticated, some of the old rules for product liability will probably have to be reevaluated.

I’m especially curious about the liability surrounding robots that were built to kill me. Which will be all of them, eventually.

[AP]

Everyone here should be familiar with Isaac Asimov’s three laws of robotics.  In the event that you may have recently been afflicted by amnesia, or if you just like reading them because it gives you a warm idealistic feeling about the future of robots, here they are again:

1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.

2. A robot must obey orders given to it by human beings, except where such orders would conflict with the First Law.

3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

So far, though, these laws have not been programmed into robots. Most industrial robots work apart from humans, in factories and on assembly lines. They work with tremendous speed and deadly force to accomplish specific tasks. Because of the hazard they pose to humans, robots can’t work closely to us. And programming a robot to be more wary of nearby humans presents increased complexity in the programming, thus making it less reliable.
“You have to give the robot a way of recognizing what a human is and a way to prevent itself from hurting a human.”
The European Union, however, is funding a project to develop robots that are simple and powerful, ones that will mimic human muscle action using new movement control techniques. New sensors will give the robots a sense of where their “limbs” are. The proposed robots would also weigh less and have limb actuators for lessening the force of a strike upon impact, so they don’t hit as hard. Having humans and robots work side by side is very desirable because industrial processes that involve aspects from both entities could be performed simultaneously, thereby streamlining those processes. Programming the three laws of robotics into robots is no longer science fiction, now it becomes a vital task in advancing robots in industry and, eventually, in society as a whole.

[io9]