Saturday, March 10, 2012

Applying Asimov’s laws of robotics to our lawless computers

From North Jersey.com: Applying Asimov’s laws of robotics to our lawless computers
By GENE NEWMAN
COLUMNIST
Print | E-mail

Sci-fi writer Isaac Asimov wrote the "Three Laws of Robotics" in 1942 for one of his short stories. I think it's high time they were amended to deal with current conditions.

Briefly, the first law states a robot may not cause a human to be injured by the robot's action or inaction. The second calls for a robot's complete obedience unless that conflicts with the first law and the third orders the robot to protect itself unless that conflicts with first or second law.

Those laws were all very good in 1942 when there weren't many real robots around, but today I watched a video of the first handshake by a human and a robot in space when Space Station Commander Dan Burbank shook with NASA's Robonaut, a cartoon-like version of a helmeted astronaut. When that six-foot-tall android reached out his big realistic mitt, I hoped he had nothing else in mind but a friendly shake. Remember the murderous HAL 9000 in the old Stanley Kubrick film?

Are Asimov's laws being programmed into all robot designs? I'm sure they're not governing those annoying telephone robots we all have to deal with. They've damaged my nervous system and my shouted commands are never obeyed.

Computers should also be covered by the three laws and their amendments. They are, after all, robots without appendages. My particular PC is also missing a heart and a conscience. Early computers were much more obedient and cordial. Our company had one that was twice as big as a refrigerator and performed calculations in a few seconds that took a lowly engineer several hours to complete. His name was SAM which stood for Solid-state Accelerated Mainframe. SAM was user friendly and politely asked the operator's name at each startup. I soon realized he was rather naïve and I began to sign in as Albert Einstein.

"Good morning, Albert Einstein," would be SAM's cheerful greeting when I sat down at the keyboard. It was a nice ego trip.

That was long ago when computers knew their place. Now we have to strengthen Asimov's laws before these monsters take over completely. For the first law, we should add that humans can come to harm in other ways than physical injury. Frustration, exasperation and aggravation should all be banned.

For the second law, don't let that obedience loophole stand as is. If a computer argues that obeying a command could prove harmful, we should get a plain English explanation. I also think we should drop the third law altogether. If a computer needs protection, we can provide it – or not.

No comments:

Post a Comment