South Korea Is Building Legions of Robots -- Soc, Tech

*
Nearly 80% of the homes in South Korea have high-speed, always-on broadband connections to the Internet. What could they possibly use all that bandwidth for? Well, among other things, controlling the security and military robots they intend for widespread use by the 2010s.

As noted in this article in The Korea Times:

When completed, the outdoor security robots will be able to make their night watch rounds and even chase criminals, according to Lee.

The government also seeks to build combat robots. They will take the shape of a dog or a horse, with six or eight legs or wheels.

Toward that end, the Ministry of Information and Communication (MIC) and the Defense Ministry will combine to channel a total of 33.4 billion won ($33.9 million) through 2011.

``The robots will be directed by a remote control system or move autonomously via their own artificial intelligence systems,'' MIC project manager Oh Sang-rok said.

``The two sophisticated robots will be empowered by the country's state-of-the-art mobile network, thus enabling mass production at an affordable price,'' Oh noted.

Smart robots need three basic functions of sensing, processing and action. Thus far, robotics researchers have tried to cram the three into a single dummy, causing expenses to soar.

Instead, the planned robots will be receiving most sensing and processing capabilities via a Web connection. Only the ability of movement will be located in the robot.

All right, I've scoffed, however mildly, at the idea of 50s SF-style legions of robotic shock troops in the past. And I still don't think a corny B-movie AI takeover is in the offing. But given a very serious push in the Pentagon to replace pilots on dangerous, long-range bombing missions (and perhaps ultimately all strike missions), we could end up with some very formidable military forces that are controlled entirely by robots.

Assuming there are still humans securely in the decision-making loop, this still may not be a critical concern. After all, our most powerful weapons, ICBMs and SLBMs, have been controlled by robotic guidance systems for as long as they've existed. So long as we (and the Soviets, Chinese, British, French and Israelis) controlled those programs, it wasn't the robots we were worried about.

And honestly, it's still human decisions about warfare and terrorism that bother me, not what the coffee machine may be plotting. But it is curious to see how much of our life has been automated in recent years... even tasks many thought could never be done without humans.

And it's also important to remember, even if we're not going to be overthrown by random tankbots or bombsniffing security programs, we're still going to have to secure all this infrastructure and make sure we can keep control when hackers and saboteurs decide to disrupt it. Honestly, I'm not so much worried about robotic aircraft going berserk on us (see Stealth) as I am seeing nations becoming dependent on remote-controlled aircraft in an era of radio communications. Until you have quantum entanglement communications (or something else that beats light speed and can't be physically disrupted) you'll always be vulnerable to someone with a big enough jammer. And if jamming your signal is the only way to stop your invincible Air Force, believe me, the bad guys'll have a big enough jammer.

Still, the idea that more robots may end up dying in our next war than people isn't necessarily that disturbing. What is disturbing is the question: With the technology we're developing now, for how much longer will war between nations be feasible, or even survivable?


Future Imperative