The Rise of Robot Warriors

Spectrum: In the Unmanned Aircraft Systems Roadmap, the Department of Defense said it hoped—and I think most people thought then it was a fantasy—to have around a total of 350 unmanned aerial vehicles in the military inventory by 2010. How may UAVs are in the U.S. inventory today?

Singer: Let’s pull back and not just talk about UAVs. Overall, the numbers are evidence of why we are living through such an important moment where science fiction really is becoming battlefield reality, and why this story is important, not only to war but maybe to humanity itself.

We go into Iraq with literally a handful of drones and the U.S. military inventory. All of Fifth Corps [which led the advance into Iraq during the invasion from the desert west of the Euphrates River] has one drone supporting it; we now have over 7000 in the U.S. military inventory. On the ground we go in with zero used in the invasion force; we now have over 12 000, and the growth curve appears to be continuing to be exponential. It’s zero in 2003, it’s 150 in 2004, and by 2005 it is 2400; I spoke with an Air Force three-star general who says we are soon going to be talking about ”tens of thousands” in our conflicts.

These numbers are remarkable, but you have to think about them in a couple of ways. One is that these are the earlier models; these are Model-T Fords or Wright Brothers fliers in comparison to what’s coming. It’s interesting that in comparison to World War I, the number of UGVs [unmanned ground vehicles] we have today is about the same number as the British had tanks at the end of the war.

The next part of this to remember is that the capability of these systems and the applications they are being used for is going up exponentially. If the idea of Moore’s Law continues to hold true, then 25 years from now these systems will be as much as 1 billion times as powerful as today’s in terms of their computing power. Now if that is not true, if they are only one-hundredth as much, then they will be ”merely” a million times as powerful.

That leads to the second thing, which is that technologies are revolutionary not only because of the incredible new capabilities they offer you but because of the incredible new questions they force you to ask—questions about what’s possible that was never possible before and also new questions about what’s proper, what’s right or wrong that you didn’t have to think about before. That’s really what the book is about.

It is about going around meeting all these people, from the robot scientist who wonders if he is equivalent to the nuclear physicist back in the 1940s, to the science fiction author who is now actually having a major impact on what’s being made and done in the real world, to the 19-year-old drone pilot sitting in Nevada who is now flying a system over Pakistan. All the experiences of war are fundamentally different for him than they were for every single generation of soldiers before. He’s experiencing war, but he is not experiencing any risk.

It goes to the officers who are commanding robots; it is largely different from commanding people. It may be silly to say that, it’s so evident, yet it’s no longer a science fiction question. It goes to politicians, where it is affecting when and where we decide go to war.

What Wired magazine calls the ”robot war in Pakistan” right now is a good illustration of that. We have carried out as many strikes in Pakistan last year as we did in the opening rounds of the Kosovo war. We didn’t debate about it in Congress like we did the Kosovo war, because I would argue it’s riskless, but just to us. But the other side of this is that it affects how people look at us: people in Pakistan, people in Lebanon, it affects how the insurgents look at us.

When you do have this act of force, but [you’re] not going in harm’s way, that raises all sorts of new questions about accountability and the laws of war. That’s what makes this revolutionary, what makes this equivalent to the printing press, the computer, the atomic bomb.

Spectrum: The IEEE just celebrated its official 125th anniversary. Given the direction of robotics, what advice would you give engineering students who are interested in robotics about what they need to be studying?

Singer: You need to have a really multidisciplinary background. Engineering can no longer operate in blissful isolation and in blissful ignorance of other fields. And this is not just crossing into other hard sciences such as engineering and biology, which is leading to the field of bioinspired robotics. It also crosses into the social sciences and, more broadly, the humanities. I think a major area needed is a better understanding of law and ethics.

The very definition of a robot is that it has sensors, processors, and effectors. A robot’s effectors create change in the world. They are the parts of the robot that, as one scientist put it, ”create drama.” Well, dramas can be enjoyable or heartrending. By their very definition, robots create change in the world.