Government will eventually integrate artificial intelligence into its operations as it did with innovations such as the personal computer and smart phones, but doing it effectively will require a great deal of management attention, according to a report from the National Academy of Public Administration.
“Based on past experience with the introduction of information technology tools, AI will change what an employee does, or how they do it, but it will not eliminate all positions. If introduced in the right way, AI can enable organizations to achieve more, and produce better and timelier outcomes for the citizens being served,” it says in the latest of a series of reports and projections on the potential impact of AI in the government workplace.
One major challenge is the concern of employees that AI applications will eliminate their jobs, a concern the paper calls “understandable.” But it said that AI “does not necessarily replace or eliminate jobs. Rather it can enhance and create opportunities for new jobs and better opportunities for workers.”
It said that tasks where AI is being introduced management must clearly define the program’s purpose and articulate the benefits. “Explain to your employees how AI can be used to free them up from work that is not directly related to their mission and allow them to focus on higher level work that is more impactful. A clear purpose statement is powerful because it shifts an individual’s focus from the tasks they perform to how all of the work done by a unit contributes toward a larger goal.”
During planning, management needs to: “assess the readiness of your workforce for AI,” including skills gaps and the need to train existing employees or add employees already having those skills and programs for those whose jobs may be eliminated or significantly changed; assess processes that could be candidates for AI deployment “and be ready to admit when a process is not viable”; “define and describe the future state of the business function”; and “include quality checks early in the process.”
During implementation, management needs to understand the importance of testing—including testing of user acceptance–and process improvements; and create a pilot project to conduct a trial run, defining roles for the team members and training them in advance.
“Once you have implemented your AI, make sure that your workforce is bought into and aware of the process for maintaining it. Maintenance activities may include scheduling of bots, real-time monitoring, communication with application owners, and infrastructure improvements,” it adds.