Japan agrees to rein in killer robots

It’s just a shame that Stephen Hawking didn’t live to see this day, but at least Elon Musk will be able to take some comfort in it. Both men were among some of the earlier experts to send out warning signals about the advent of Artificial Intelligence and particularly the dangers of adapting such technology to autonomous weapons. Referred to as Lethal Autonomous Weapons Systems (or LAWS), these are mobile weapons, in some cases actual robots, that have the ability to eliminate a target, including people, without the intervention of a human controller. If adding dodgy Artificial Intelligence to a weapons platform like that doesn’t worry you, I suggest you seek counseling.

In any event, Japan had been urging “caution” in introducing such restrictions because they feared it might stimy the development process. But at an upcoming convention, they will come out in support of such restraints. (Japan Today)

Japan is planning to give its backing to international efforts to regulate the development of lethal weapons controlled by artificial intelligence at a U.N. conference in Geneva later this month.

It would mark a departure from Japan’s current policy. The government was already opposed to the development of so-called killer robots that could kill without human involvement. But it had called for careful discussions when it comes to rules so as to make sure that commercial development of AI would not be hampered…

At the conference of the U.N.’s Convention on Certain Conventional Weapons, Japan will submit documents summarizing its views on LAWS and call for the establishment of a panel of technical experts in the field of AI, according to the sources.

In the documents, Japan will say it cannot overlook the development of such autonomous weapons and does not rule out the possibility of working toward a global ban on them, they said.

I don’t know if someone in the Japanese government finally sat down and watched the entire Terminator series of films or not, but this sounds like good news. We have at least somewhat better oversight of what’s happening in both the robotics labs and Artificial Intelligence development centers in the United States and Europe. But what goes on in China and Japan is quite often a mystery. (And there’s a lot going on.)

We already have some remotely operated robots being used by the military, as well as fire and police departments back home. But those generally require real-time piloting by a human being. One of the most independent killers out there is the unmanned drone aircraft with weapons onboard. Some of them can already be shifted into a mode where they can detect a target and take it out after they lose contact with command, and that’s frightening enough.

But the day may be coming sooner than we think when significant parts of the fighting taking place in a war zone will be handled by robots. (At least for the countries that can afford them.) I’m pretty sure we already expect the AI to wake up one of these days and decide we’ve had enough time ruling the Earth. We don’t need to make the job even easier for them.

On that subject, have you seen the Spot Mini from Boston Dynamics yet? Unlike its larger, freerunning cousins, the Spot Mini is more compact. It knows how to crouch down and crawl under low openings. And they’ve already trained to clean up after parties. If it gets knocked over, it can right itself and be back on its feet in no time. Here’s a video of the Mini in action. Now just imagine giving it a couple of weapons and a bit larger brain. Sleep tight, kids.