WASHINGTON, D.C. -- The private sector's pace of adoption of AI and automated software tools has been relatively sluggish. While companies are beginning to use AI tools and benefit from them, a lack of education about the technologies, a shortage of data scientists, and the unwillingness of employees to use the new technologies have slowed down use rates.
Still, the private sector is using these technologies faster and more efficiently than the United States federal government. The use of AI in government, according to experts, is lagging. Compared even to other governments, particularly China, the U.S. government is behind.
"The federal government is not ready for the new world of AI," said Bill Valdez, president of the Senior Executives Association, during a panel at the AI World Government conference.
Valdez cited a 2017 survey of the Senior Executive Service (SES), a class of executives specified by the U.S. government civil service regulations, that indicated the executives' awareness of AI technologies was low.
Fifty percent of SES executives surveyed said they were not ready for the future of technology, Valdez said. "They just had no clue as to what's coming."
In general, he added, "the federal government is not designed to embrace this big of a change."
Transition to the use of AI in government will have to be aided by the private sector, he noted.
Stimulating use of AI in government must be done one person at a time, said Margaret Williams, learning policy and program support lead at the United States Postal Service, during a panel.
"It will require a lot of thought, and I do believe it will require putting a lot of teams together," she said.
Good uses of AI
That's not to say the U.S. federal government doesn't use AI and automation. While the technology may not be available to the typical employee or manager, some agencies use advanced AI technologies extensively.
NASA, for example, has long developed and deployed AI tools, long before AI became a hot term.
Derrick Graham is an intern at NASA's Goddard Space Flight Center. In an interview, he said he attended the AI World Government conference in part to explore the AI, machine learning and automation tools available on the market.
At NASA, Graham helps an augmented and virtual reality team integrate point clouds into virtual reality, and assist with projects to achieve super-resolution, using deep learning and machine learning techniques to make terrain more visually clear.
Bill ValdezPresident, Senior Executives Association
Krishna Pokharel, an employee of the Federal Energy Regulatory Commission, an independent agency within the Department of Energy, said in an interview he hopes to find tools to assist with market monitoring or compliance issues.
It would be worthwhile looking at how we can automatically monitor "utility companies filing on time if they file the correct numbers, that kind of stuff," Pokharel said.
Both Graham and Pokharel said their comments expressed their own views and are not representative of their agencies.
In a keynote by Lindsey Sheppard, an associate fellow at the International Security Program at the Center for Strategic and International Studies in Washington, D.C., cited a recent survey that she said showed a lack of AI readiness in the U.S. government.
The Government Business Council, a private research group, surveyed 668 decision-makers in the federal government on AI readiness. It found that about 60% were far or extremely far behind other government entities on AI. About 32% said they were on par with other agencies, while a combined 8% said they were either ahead or very far ahead.
Another survey indicated the decision-makers largely have no plans to implement AI or have some type of program for the future.
The numbers, Sheppard said, showed a general lack of AI readiness in the U.S. government.
"Modernization isn't going to happen overnight," Sheppard said.