Originally shared by Daniel SuarezI'm glad you're sorry, Tom Simonite -- because you're wrong.I've seen these weak arguments in support of autonomous robotic weapons before (see article below). For example:
"Trying to ban them outright is probably a waste of time." I'll make the time.
Lethal autonomy was the topic of my TED Global talk back in 2013, and the focus of my third book,
Kill Decision, in 2012. I'm also on the executive board of Responsible Robotics, an organization dedicated to promoting the development of ethical machines. So I'm familiar with the issues involved, and pushing for a ban is not a waste of time.
Let's be clear:
a ban on autonomous robotic weapons does not mean there will be no autonomous robotic weapons in the world. It means that civilized nations will treat as a pariah those who develop or use such systems. Thus, only rogue nations, criminals, and terrorists will use them. Yes, they will still exist.
Robotic systems to detect and disable incoming killer robots, on the other hand, could be a huge industry for mainstream robotics companies. And by not killing people that industry would largely be viewed as protectors of civil society -- instead of engendering resistance.
The immediate counter-argument is usually: "But human beings make bad decisions in war." Yes. Undoubtedly. Perhaps we would do better to help human beings make better decisions -- unless, of course, you think we're already obsolete.
"The line between weapons controlled by humans and those that fire autonomously is blurry, and many nations—including the US—have begun the process of crossing it.Well, if they've already begun, why bother, right? This is a gross simplification of this process. Just because a matter is complex or already begun doesn't mean people should throw up their hands and give up. No issue of any significance is simple, and drawing lines on 'blurry' issues is what modern law is all about. In society lines are decided through debate, litigation, and precedent. This is nothing new. And history is littered with grand schemes that someone tried to rush through -- but which withered upon closer review.
If you believe the money to be made from autonomous weapons is just too enticing, then I invite the wealthy elites of the world to contemplate how much more hellish things could be if their enemies could select from a wide range of killer robots capable of anonymously assassinating them. And remember: successful tech is frequently pirated and knocked off in black market factories. In today's global economy there's LOTS of excess high tech manufacturing capacity. Far better to constrain the design and sale of killer robots. It won't be perfect, but it will make it that much easier to locate the perpetrators when illegal robotic weapons do kill.
"A report on artificial intelligence and war commissioned by the Office of the Director of National Intelligence concluded that the technology is set to massively magnify military power. "Precisely. And since the means by which human beings resolve conflict has historically shaped our sociopolitical landscape, magnification of military power could drastically alter society itself -- centralizing power into fewer, unaccountable hands and imperiling democracy.
Soft power -- that is, the power that comes with cultural and moral authority -- has been woefully underused in the past thirty years. A policy of banning robots that kill people would hold the moral high ground in defense of democratic principles and humanity. Would it be perfect? Of course not, but statements of principle matter more than most people think.
Ask yourself this: what exactly have we gained by increasing our use of robotics as weapons in war? Is there more or less conflict in the world? Are robotic weapons making us safer, or are they creating new enemies?
Very few would argue against defensive systems to guard against attacking robots from rogue nations, terrorists, and criminals -- but defense robots don't need to be traditional weapon systems. They can be robots that capture or cripple interloping robots, dragging them back to an explosive-proof enclosure for follow-up investigation by humans. That's the sort of system that a civilized society develops.
And one more thing: take a brief tour of sci-fi. In all the stories, movies, and TV shows, take note of the side that deploys the robotic killing machines. Are they the villains or the heroes? That illustrates the deep wellspring of disdain which exists in all cultures against those who refuse to own the consequences of their actions.
Autonomous robotic weapons will
create conflict, even where none existed before. Humanity will not go quietly into this darkness, and rather than apologizing for your casual dismissal of a critical issue, I urge you to think about the actual
consequences of autonomous robotic weapons -- not just whether it's practical or convenient to resist them. I assure you, avoiding unnecessary war and preserving democracy is eminently practical.
For those curious about my 2013 TED talk, it is predicated upon the idea that the means by which humans resolve conflict has always shaped our sociopolitical landscape -- and autonomous robotic weapons will drastically change that landscape. Here's the link:
https://www.ted.com/talks/daniel_suarez_the_kill_decision_shouldn_t_belong_to_a_robotYou can find Responsible Robotics here:
https://responsiblerobotics.org/https://www.wired.com/story/sorry-banning-killer-robots-just-isnt-practical
