Enable breadcrumbs token at /includes/pageheader.html.twig

Three Steps for AI Success

Artificial intelligence requires infrastructure, collaboration and training.

The steps required to get artificial intelligence efforts off the ground are clear and tangible—invest in the right infrastructure, develop strong industry and government partnerships and prioritize training and hiring for AI skillets. Credit: geralt/Pixabay

Over the next five years, artificial intelligence (AI) will redefine what the U.S. federal government can achieve with technology. AI will help ensure our nation stays competitive, effectively serves its citizens and maintains safety for Americans at home and abroad.

Potential applications include streamlining administrative operations to allow federal workers to focus on high-value tasks, thwarting cyber attacks through automated monitoring and mitigation, delivering actionable insights to better safeguard our troops on the battlefield and more efficiently providing critical services when they come home. AI could potentially save the federal government 96.7 million hours and $3.3 billion each year, enabling better support of these objectives.

Understanding this potential, the government is acting. We’ve recently seen the White House come out with an executive order focused on supporting AI, and the Department of Defense shared its AI strategy publicly. This has been coupled with a plethora of agency requests for information (RFI) issued over the past year. These RFIs are largely focused on understanding AI, not implementation or execution—indicating that AI is still a technology federal agencies are working to comprehend. However, if they work closely with industry, these agencies will finally be able to move from the drawing board to reality and fully realize AI’s benefits.

Three Steps to Prepare for AI Adoption

Step 1: Agencies can prepare a foundation for any AI application by building the right IT infrastructure. Strong industry and government alignment, prioritization of federal workforce training and recruitment and retention are also necessary areas of focus.

AI isn’t a new concept. In the 1960s and 70s, the government was already thinking about AI but lacked the infrastructure to support it. Today, the infrastructure is at our fingertips but isn’t widely used. A multicloud approach is well suited to AI and can provide a strong launching pad, offering an array of benefits over traditional cloud models.

Multicloud models can hold large volumes of data, provide the processing power necessary to deliver insights at the speed of the mission and lay the groundwork for an intelligent environment. Today, many government agencies are using public clouds to test the development of algorithms but are discovering those clouds need to run close to the data sets to get useful insights. Starting in a multicloud makes this a seamless transition.

Step 2: Industry-government collaboration is also critical to supporting AI, including working together to identify solutions to agency challenges and helping educate agency leaders and policymakers. Further, industry can invest in research and development that will ultimately benefit the government—and often allow new solutions to have an impact on a condensed timeline.

A great example is Kessel Run, a U.S. Air Force partnership with Dell Technologies’ Pivotal Labs built to help the government harness emerging technologies such as AI. When the Air Force needed a new solution for tanker refueling, this partnership resulted in the development of a modern, agile, software application in under 120 days. In that time, the application went from planning to deployment in an active theater of operations with minimal disruption. This solution is saving the U.S. government hours of manpower and millions of dollars a week in fuel costs.

Step 3: Finally, the federal workforce must be equipped with the necessary AI skills. While the government is closing the cyber talent gap, up to 77 percent of the federal workforce needs to be trained in emerging technologies, including AI and machine learning. The federal government must also commit to workforce recruitment and retention, focused on data scientists and coders.

The government is competing with industry to attract new talent. One of the things the government can use to its advantage is that it offers a mission. Forty-five percent of the tech-savvy Generation Z say that they want work that has meaning.

Executive-level recruitment from industry can also aid continued advancement. Agencies can tackle sophisticated problems on a short- or long-term assignment by recruiting individuals with practical experience in the commercial sector. The federal government could also mine academia to build a consistent and robust talent pipeline.

As the recent executive order and RFIs issued by agencies reflect, the possibilities of AI are endless. The steps required to get AI efforts off the ground are clear and tangible—invest in the right infrastructure, develop strong industry and government partnerships and prioritize training and hiring for AI skillsets. If we can work together to support this baseline, the intersection between government mission and technology will enable possibilities and support agency objectives in ways we never before thought possible.

Cameron Chehreh is a vice president, chief operating officer and chief technology officer, Dell EMC.