Page 34 - FCW, October 2017

EN TE RP RISE
SPONSORED REPORT
OUTL K
Technology Modernization AI Comes of Age
Artificial Intelligence is helping drive government agency modernization efforts.
Artificial intelligence (AI) has been around in some form at least since the 1950s. Attempts to harness
its power into practical and useful applications, however, have stuttered. Only recently have improvements in the fundamental processing power of most computers risen to the level of being able to handle the sophisticated algorithms of AI. This has boosted the promise of practical AI and its close cousin machine-learning.
That promise is particularly relevant to the IT modernization efforts currently underway throughout government agencies. Driven by the expectations of an increasingly digital- savvy public, agencies have been straining to transform their sprawling, aged, costly, and resource-heavy IT environments into something that can quickly and efficiently deliver citizen- friendly digital services.
Some agencies have progressed faster with AI than others. The U.S. Postal Service, for example, uses handwriting recognition to have machines process up to 18,000 pieces of mail an hour, instead of relying on much slower and error prone human processing. The U.S. Army recently demonstrated how IBM’s Watson AI system could “personalize” repair and maintenance for some of its vehicles as an assist for its mechanics.
Even the Securities and Exchange Commission has made “recent and rapid advancements with analytic programs that harness the power of big data,” says Scott Baugess, acting
director and acting chief economist of the SEC’s Division of Economic and Risk Analysis (DERA), speaking at a recent financial industry forum.
“At this point in our risk assessment program, the power of machine learning is clearly evident,” he says. “We have utilized both machine learning and big data technologies to attract actionable insights from our massive datasets.”
Those agencies and others that have thus far engaged with AI and machine learning are the so-called “early adopters.” They
To that extent, according to a recent report by Deloitte’s Center for Government Insights, AI applications already being used in government is changing the nature of many jobs and “revolutionizing” parts of government operations. What agencies face, it states, is whether to use it to fully automate certain tasks or enhance tasks now done entirely by humans.
The potential rewards are huge. The report estimates that by simply automating tasks that computers already routinely do, AI-augmented government agencies could
“At this point in our risk assessment program, the power of machine learning is clearly evident.”
—Scott Baugess, acting director and acting chief economist of the SEC’s Division of Economic and Risk Analysis (DERA)
have the resources—researchers, software developers, and high-end computers—to manage resource intensive cutting-edge AI applications. The intelligence agencies, the Department of Defense, NASA, and the government laboratories are all leaders in the government’s use of AI.
Arguably, though, the most dramatic results will come when those agencies that don’t have these kinds of resources can still manage to successfully adopt AI and machine learning. Those smaller agencies
still need to modernize their IT to improve how they deliver on their individual missions.
“conservatively” free-up 96.7 million federal government working hours each year, with a cost savings of approximately $3.3 billion. That is a significant difference.
At the high-end, the report says, AI technology could free-up 1.2 billion working hours, for $41.1 billion in savings. Before that can happen, though, agencies need to assess
the range of AI technologies and determine which ones will be best suited to their missions. Then they can integrate those technologies into agency workflows. However, that transformation still requires funding and the appropriate technical skills.