At a time when the government is scrambling to adopt artificial intelligence (AI) and other technologies, two top federal health officials offered some cautionary notes on Nov. 12, saying agencies should be careful in pursuing “shiny” objects and that human judgment remains vital.
Dr. Henry Walke, director of the Office of Readiness and Response at the Centers for Disease Control and Prevention (CDC), said the CDC is integrating new technologies in its communications with state and local health departments about potential health emergencies.
“When we think about these great tools and the shiny objects, that’s wonderful,” Walke said during a panel discussion on “health security in the homeland” at the GDIT Emerge: Protect the Homeland event in Washington, D.C. “But at the same time, we really need to continue to invest in our workforce.”
If emergencies occur, Walke said, “We talk a lot about how you envision a few people sitting in front of 1,000 computer screens and picking up that signal that’s going to save us all. But sometimes, it’s really just we see a signal, we pick up the phone, or a clinician calls us and says, ‘I have someone here in my emergency room who is traveling to Central Africa who has signs and symptoms of Ebola,’ and then we have a consult around that, and then we have quick testing.”
“There is a need to have these systems integrated together and (a need) for all of the technologies,” Walke added. “But we also need to communicate with our partners, especially the clinicians who are actually on the front lines.”
Saum Ghosh, director-Immediate Office in the Office of Mission IT Services at the Food and Drug Administration (FDA), said federal officials “need to be very careful about running after the next shiny object” as they integrate new technologies into agency missions.
“There are tools and algorithms and technologies available, which absolutely do a fine job in certain situations,” he said. While newer technologies offer greater speed that can be valuable, he added, “we need to be extremely comfortable and confident that the technology that we are deploying is going to meet the needs of the situation.”
Their comments came amid a federal mandate to quicken the adoption of AI and other technologies to improve government operations and national security.
In an executive order in June, President Donald Trump said the administration’s policy is to work with the private sector “to modernize government and private sector information systems” and to “cultivate America’s advanced AI-enabled capabilities.”
Trump has forcefully pushed for rapid AI adoption, repeatedly rejecting recent calls for additional AI regulation and arguing that restrictions could hinder U.S. technological development.
Among federal agencies, the scope and pace of AI adoption “accelerated significantly over the past three years,” according to a recent Brookings Institution report.
A third panelist at the discussion at the GDIT event, Deena Disraelly of the Department of Health and Human Services (HHS), said systems integration is a “huge part” of how HHS responds to potential emergencies, such as a series of recent chemical fires and incidents in states including Florida and Mississippi.
Assessing incidents as they develop and determining whether to “wake somebody up at the state and local level in the middle of the night” still rests primarily on human judgment, she said.
“That’s based on experience,” said Disraelly, deputy director of the Center for Preparedness in HHS’s Administration for Strategic Preparedness and Response. “That’s based on people who are trained to really look at the information they’re seeing.”