The Department of Defense (DoD) needs to prioritize problem solving over shiny new tools and hot technology trends, said Chris Lynch, director of the Defense Digital Service (DDS), at Defense One’s Tech Summit today.
DDS bills itself as “a team of nerds on a tour of duty at the Pentagon to improve technology within DoD,” and is the wing of the U.S. Digital Service deployed to DoD and headquartered at the White House. DDS consults with DoD leaders to transform mission delivery, and Lynch said sometimes those leaders get the wrong impression from emerging technology trends.
He said one of the previous technology waves to hit DoD was big data, and “recently it’s become a lot of artificial intelligence and machine learning.” But these aren’t a panacea to all of DoD’s woes, and Lynch said that when he consults with the department on how to best address mission, the thinking around the technology ought to change.
“A lot of times, the thing that I focus on is that these are simply tools. They are there to enable something to fix a problem,” he said. “So, I’m very big on, ‘Let’s identify the problems that we’re trying to solve and let’s come up with a way to go after solving that thing.’”
That mode of thinking, Lynch said, can help steer the discussion to much more productive routes–and even yield surprising findings.
“When I actually have most of those discussions, where I end up is, ‘Hey, you know what? Artificial intelligence, machine learning, that’s great, but actually, sometimes you’re not struggling with machine learning. You’re struggling with Excel.’”
Lynch said he recently provided a general officer with guidance, and noted the problem at hand stemmed from the entire mission being run on siloed spreadsheets.
In other situations, the need for emerging technology–and the right people to implement it–is more obvious, he said.
“This is a talent thing. The Department of Defense runs one of the most technical missions of the world,” Lynch said. He provided an example–the F-35 stealth fighter jet–calling it a “data center with a jet strapped on the back of it” and a place where advanced tools are properly suited.
“That is where you can have somebody say this could be solved with artificial intelligence or machine learning. Great! You can actually put that there,” he said. “If you’re doing image recognition, where you’re looking at vast amounts of data, there is no doubt that computer vision can help a lot there. It can do the job better than a human. And so, a lot of times what I want to do is put a technical person in the right place at the right time, so they can say, ‘I know exactly how that should be.’”