Robot bans: Sex toys and killing machines
“Don’t bet against technology” is a quip I often make, especially when my anti-nuclear friends try to tell me something is not possible. It is not the case that technology can do everything, nor that it is somehow all powerful. But these crazy monkeys we live with are ingenious and inspired and will continually surprise us.
We do have some small amount of influence over the direction of technology and it makes good sense to think about the directions that technology is going and whether we are served by those directions. Two different complimentary efforts to control the development of machines have caught my eye. One is the proposed ban on robot sex toys. The other is a ban on killer robots.
I was at a party in NYC and was talking with a guy who is working on virtual reality machines. I knew that video games had been instrumental in pushing computer technology to better performance, especially in the area of graphics. I speculated that games would be the first big application for VR systems. “You are wrong,” he said, “it is porn.”
There are apparently already thousands of pre-orders for Roxxxy, the first sex toy robot at US $7,000 each. Roxxxy has yet to be released. And while we have not yet deeply explored the implication there us already a call to ban these devices and the infant technology. As an anarchist, I am unexcited about this ban. I am unsure what purpose it serves. Generally speaking, consenting adults (and their machines) should be able to do what they want.
Noam Chomsky, Elon Musk and Stephen Hawkin have joined with over 1000 other AI and robotics experts to demand an end to autonomous robotic weapons systems in an open letter. The UN started considering unpersoned weapons systems in 2014. The signatories of this letter warn of a new global arms race and the destabilizing affects of continuing the current trend towards rapid and uncontrolled development.
This is not drone technology. Drones have pilots, humans at the other end of the controls (even if the controls are thousands of miles away) who are making the call as to whether someone should be killed. The Obama administration broke new ground by killing US citizens with no due process with drones. But AI killing machines are far worse.
“Autonomous weapons are ideal for tasks such as assassinations, destabilizing nations, subduing populations and selectively killing a particular ethnic group,” Open Letter from AI and Robotics Researchers.
But aren’t these clever engineers being a bit stupid? Nothing can stand in the way of the march of progress, right? If we can invent it we will make it no matter how horrible it is.
Wrong. Human history is filled with all manner of agreements that have banned technologies. Perhaps the most famous ban of technology is the Montreal Protocol which banned ozone destroying CFCs. This 1987 international agreement phased out and stopped the production of CFCs which was already happening in dozens of countries around the world. As a direct function of this agreement, the ozone hole is starting to recover.
While not quite as successful the 1997 Chemical Weapons Convention similarly was able to obtain international acceptance. Leading to a dramatic reduction in these particularly nasty and difficult to control weapons. By September of 2015, 192 countries representing 98% of the worlds population were signatories. And while the original deadlines for elimination were often missed or extended, the treaty has largely been effective.
The control of dangerous technologies is ultimately a political choice. The research and mass production of autonomous killing machines is still a choice nations must make. Or we can just leave control of it to “market forces” because that always works so well.