Artificial intelligence is significant beyond the Defense Department, involving partners and allies, the acting undersecretary of defense for research and engineering said.
Michael J.K. Kratsios, also DOD’s chief technology officer, spoke during a fireside chat yesterday at the 2020 Department of Defense Artificial Intelligence Symposium and Exposition.
“[Over] the last three years, we’ve seen a tremendous emphasis across interagency and all the way up to the White House on ensuring American leadership in artificial intelligence very early on in,” the undersecretary said.
“In this administration, there was the desire to begin to emphasize AI and, particularly, [look at] what the things [are] that we need to do as a federal government to drive leadership in this particular critical domain as our adversaries are stepping up,” he added.
The first major step was the signing of an artificial intelligence executive order, and that laid out the national strategy for AI, Kratsios said.
The AI executive order laid out a thematic vision for the United States as a country to ensure its leadership, which is around research and development, he noted.
“It is absolutely critical that we are making [the] next great breakthroughs in AI and machine learning here in the United States, and that is something that requires a whole-of-government commitment to the endeavor,” he emphasized.
Other U.S. government agencies looking at AI need to come together and be part of the strategy, releasing a research and development strategic plan, he said, noting there are four core elements around it.
First, he said, there is a vision guiding document for the larger research and development enterprise across DOD to not only increase funding and support financially toward development, but actually coordinate it and make sure it’s done and used effectively.
The second line of effort is always, no matter what technology the DOD is working on, around the workforce.
“Broadly speaking, it’s very challenging to build a large-scale robust pipeline of talented AI scientists that want to continue to make breakthroughs and … figure out what the federal government has to prioritize, grants and fellowships and other types of funding opportunities,” Kratsios explained.
The third line of effort is around removing barriers to AI innovation, he said.
“I generally tend to see the world of emerging technologies as two buckets [that] are either technologies born in captivity or technologies that are born free,” he added.
Technologies that are born in captivity are often the technologies hardest to commercialize, Kratsios said, adding those are things like drones, autonomous vehicles or nuclear power.
In the fourth line of effort, “We recognize that our adversaries are pushing ahead with a very different view on the way AI should generally be used,” Kratsios saids.
Allies are critically important in the AI realm because of the way adversaries are thinking about it, Kratsios said.
“Adversaries try to influence the way AI is impacting citizens around the entire world, [so] we need to provide the right positive, Western-based alternative to that, and I think we can do that,” he said