ROSE L. THAYER STARS AND STRIPES
Army Human Resources Command has introduced artificial intelligence into noncommissioned officer boards to help more quickly evaluate who has the qualifications to be considered for promotion. (Jermaine Branch/U.S. Army) Army Human Resources Command has introduced artificial intelligence into noncommissioned officer boards to help more quickly evaluate who has the qualifications to be considered for promotion. The driving force behind the Army’s Comprehensive Board Reform and Analysis program is to make boards more efficient, smarter and ultimately more transparent, Col. Tom Malejko, chief talent analytics officer at Army Human Resources Command, said last week during the Association of the U.S. Army’s annual meeting in Washington. “The first thing to understand is that we are not using it to replace humans,” he said. “We’re using it very broadly to augment their decision-making.” The command created a “naive” AI system that does not look at names, branches or ranks, but instead can screen out people who haven’t reached a certain command level or attended a specific school. “Can we screen out individuals that are not really competitive for the process upfront … helping our board members to focus their valuable time and resources on those individuals that are then most competitive for that selection?”
Malejko said. Maj. Gen. Hope Rampy, commander of Human Resources Command, said this could help in the sergeant first class evaluation boards, when every NCO of that rank is being evaluated. Some are not competitive for the promotion yet but must get a merit list score. Before acting on AI decisions, Malejko said a team of people reviews every step the AI model considered to ensure there is no bias. “Our approach is to work our way through our noncommissioned officer boards first, learn from them and then pilot from those,” he said. “Based on what we’ve learned, go back to Congress and ask for additional authorities so we can actually execute them within our officer boards, since Congress ultimately controls those responsibilities.” The Army has, however, used an AI-like algorithm for the past four years to determine which officers should be invited to a selection board, Rampy said. Each year their post-evaluation has helped programmers refine the algorithm to be more effective. In the beginning, more than 30 officers were overlooked by the system and had to be manually included, Rampy said. Now that’s down to about three to five. “Because you can retrain it, it got better every year,” she said. The command is now looking to make a similar program to search the Army’s personnel and pay system for specific skills for a specific mission or job. Because soldiers can add skills, such as languages or hobbies, to their talent profile, this could include skills outside the scope of their job in the Army.
No comments:
Post a Comment