Ride the Lightning

Killer Robots Authorized – Can Skynet Be Far Behind?

September 23, 2013

Hat tip to Tina Ayiotis.

Somehow, I managed to miss (perhaps because there was intentionally no fanfare) the fact that U.S. Deputy Defense Secretary Ashton Carter signed a directive in November of 2012 establishing policy for the “design, development, acquisition, testing, fielding, and … application of lethal or non-lethal, kinetic or non-kinetic, force by autonomous or semi-autonomous weapon systems.” Of course, there is no mention of killer robots, but that certainly is what the directive is talking about.

The Bulletin of the Atomic Scientists posted an excellent analysis of the policy and it is, in a word, sobering. It appears that the government believes it has adequately explored all the ethical, philosophical, technical and legal concerns involving autonomous weapons.

Some people interpret the directive as promising that a human will always decide when a robot kills another human. Even if that were true, I'm not sure I'd find it all that comforting. However, I agree with the article that there is no such assurance.

Members of the military have argued that warfare will become too fast for humans to direct. Great, let the machines make military decisions – how can they do worse than humans have? Then again . . .

Some have suggested that the machines have controlled levels of autonomy, which would inevitably be "dialed up" in the event of desperate combat.

For anyone who has forgotten the origins of Skynet in the cult classic film The Terminator, it is worth looking at the Wikipedia entry. I hope reality is not about to imitate art. This is one Hollywood ending I would like to avoid.

http://twitter.com/sharonnelson