Robots warring (in a flooded world?)

Sometimes I wish science fiction would stay science fiction. I occasionally read or hear a news story that triggers a flashback to a hundred science fiction stories I've read on the same subject; and I have to remind myself that, no, this is here and now in the real world. The imagined future has arrived. It's not just an abstract problem anymore.

The story that did that to me today was about robots. P.W. Singer, an expert on modern warfare, and the author of books on child soldiers and mercenaries, has come out with a new book on the growing use of robots in war. There are thousands already in use, and many more in the design & development stage. You can read some articles about his book here. I heard an interview with him today. He points out some of the troubling new ethical and social issues this kind of technology creates. For example:

So, for example, if you are sending less and less Americans into harm’s way, does it make you more cavalier about the use of force? And one of the people that was fascinating that I interviewed was a former assistant secretary of Defense for Ronald Reagan, who actually had this to say about these systems. He worried that they would create more marketization of war, as he put it. We might have more shock and awe talk to defray discussion of the true costs of war.... I mean, the concern I have is that it takes certain trends that are already in play in our body politic. We don’t have declarations of war anymore. We don’t have a draft. We don’t buy war bonds anymore. We don’t pay higher taxes for war. And now you have the fact that you may be sending more and more machines instead people. And so, you may be taking the already lowering bars to war and dropping them to the ground.

He suggests that in some ways we have a parallel to 1942, when atomic weapons were in development. Wouldn't it have been better to debate the morality of nuclear bombs before they were used, rather than after?

This exchange was particularly interesting:

AMY GOODMAN: What happens if a robot commits a massacre?

P.W. SINGER: It’s a great question. You know, who do you hold responsible? Do you hold responsible the human operator? Do you hold responsible the commander who authorized them there? Do you hold responsible the software engineer who wrote it wrong? And how do you hold them accountable? We don’t have good answers.

And what was funny is, one person that I interviewed was a Pentagon robotic scientist. And he said, “You know what? You’re wrong. There’s no social or ethical or legal dimensions with robotics in war that we have to figure out.” He said, “That is, unless the machine kills the wrong person repeatedly.” Quote, “Then it’s just a product recall issue.” That isn’t the way I think we should be looking at the social, ethical and legal dimensions of all of this. And that’s why we need to launch a discussion about it. Otherwise, we’re going to make the very same mistake that a past generation did with atomic bombs, you know, not talking about them until Pandora’s box is already opened.

You can read the whole interview here.

I also came across this story, about a new study that shows a general warming trend in Antarctica. If the West Antarctic ice sheet were to collapse, the average global sea level would go up by six meters! (I think previously the estimate was five meters.) And apparently we need that ozone hole after all, since it keeps things colder. Anyways, it's not an imminent threat (ie, not in my lifetime) but it would endanger humanity if we keep going the way we have been.