Nurturing broad adoption of artificial intelligence (AI) and machine learning techniques to improve the nation’s defense preparedness is less a matter of technical capacity, several presenters said at a workshop hosted on July 31 by the LMI Research Institute (LRI). Rather, it’s overcoming ingrained organizational challenges, such as culture, processes, and people—fostering both a skilled workforce and leadership that appreciate the possibilities and consequences of the nascent AI arms race.
Adversaries, notably China and Russia, have been less inhibited by such barriers, said retired U.S. Army Maj. Gen. Bill Hix.
“The challenge we have is [the U.S. and its allies are] not inclined to adopt data- and AI-driven strategies in the way the Chinese are, in part because they have no legacy processes or heritage in the modern industrial complex,” he said. “Companies in Europe and the U.S. have been at the top of their game for decades or longer. It’s really hard to move on from what they have done because it’s been successful.”
Josh Wilson, LMI’s vice president for advanced analytics, added context to this theme, explaining that the open source community is rapidly advancing capabilities. “The math and the algorithms are not impediments,” he said. “It’s getting the data in order and establishing data policies” to ensure compliance with laws, privacy obligations, and personnel protocols.
Wilson also discussed how LMI helps government customers accelerate delivery of analytics-driven insights. “Time to insight is the only metric that matters,” he said. “We align data, modeling, and development operations to make that duration as short as possible. Metrics like metadata coverage and data quality can help diagnose issues, but time to insight is the truest measure of effectiveness.”