Our interview for the April -- June Critical Thinking with Paul Scharre--he's a Ranger vet who did tours in Iraq and Afghanistan who is now a robotics expert at the Center for a New American Security--covered so much good material, but we didn't want to leave it on the cutting room floor. Here's Scharre on how DOD should pick which advances in robotics and artificial intelligence (AI) to pursue.

ARMY AL&T: It's a high-stakes game of predicting the future, better testing and lower barriers to entry and innovation for the work of research labs and outside industry.

SCHARRE: It's just so hard to predict what technologies are going to matter. The most important thing, if you want to think about maintaining advantage, is to create bureaucratic structures and organizations that allow the Army to rapidly innovate--so units that go out in the field and do experiments, demonstration, processes of competitive doctrine development and concept of operations development. The more the Army can foster these various models of innovation--the more different, diverse processes you have, a diverse group of ideas and people together--the better in terms of connecting warfighters and engineers and really hard problems. And then shaking the box, and encouraging creativity and outside-the-box thinking and new solutions. That's what we're going to need if the United States is going to stay competitive in the long run.

So, for example, you look at the DARPA Grand Challenge [in 2007] with automobiles. [For more on historic military experimentation with driverless vehicles, go to "Nobody, Take the Wheel!"] It wasn't really good enough that you could transition any of it to military stuff, but initial competitors in the Grand Challenge went on to work in Google and places in the automobile industry where they're now maturing those technologies. Eventually they might be good enough that the military says, OK, I want to take those autonomous vehicles and apply them to military settings, which will be different. We won't have mapped the environment--we might be in a GPS-degraded or denied environment--and it's going to be more challenging. So we want to apply those [technologies].

But the problem is that DOD has erected barriers to that happening. There are barriers within DOD for a research lab transitioning into a program of record; there are even larger barriers going from commercial innovations into DOD. And a number of smart folks in the acquisition community are trying to find ways around those barriers.

ARMY AL&T: The Army could be doing more in terms of thinking long-term about how robotics might shape the battlefield, what would conflict look like if militaries moved heavily into robotics and automation and AI.

SCHARRE: That's a place where we're sort of sketching out the hazy contours of the future, but the more we can expand our primer and start thinking about new ways of use, the better. If you look at the interwar period, there were people in the U.S. Navy who dismissed the potential for aircraft to be used as a decisive advantage to sink battleships into Pearl Harbor. We don't want to be there. We don't want that to be our wake-up moment of wow, this technology is transformative, and then we have a Pearl Harbor moment.

You know what I want to see? I want to see a new Louisiana Maneuvers for robotics and automation and human-machine teaming: Do a series of exercises where we go into the field and we experiment and we break things and try new ideas, and we keep doing it. You know, the first DARPA Grand Challenge [in 2004] was a miserable thing. The robot didn't go anywhere. And they came back the next year, and then they were able to drive it across the desert. And it's only through that iterative process of really hard problems and trying to solve them that we can come up with solutions.

Read the full interview at http://usaasc.armyalt.com/#folio=90.