COVID-19 has shut down businesses and plunged the economy into a condition not seen since the Great Depression. Throughout the pandemic, public health experts have emphasized the necessity of social distancing and stay-at-home orders designed to flatten the infection curve and bring down hospitalizations and fatalities. But even with these directives, the pandemic’s effects on government, businesses and the general public have been dramatic in terms of public health, the economy and social wellbeing.
To meet this public health and economic challenge, we need to marshal all of our technological expertise to develop a national plan that supports innovation. Such a plan is vital to American leadership, security and competitiveness. If the United States does not lead, if it does not stimulate and catalyze innovation both for itself and the community of democracies, the results could be catastrophic, creating the opportunity for authoritarian and illiberal regimes to dominate key technologies in the 21st century.
The plan must have multiple elements that are adaptable over time, based on the competitive environment: innovative tools, infrastructure support, improved data access, and workforce development geared to the needs of a digital economy. Taken together, these ingredients lay the groundwork for continued leadership in technology innovation and economic prosperity.
The good news is we are well along the path towards developing innovative tools. In the health care area, for example, a Canadian startup used natural language processing to analyze hundreds of thousands of news reports and chart the spread of the virus. The analysis provided an early warning system for medical professionals beyond their own patient-based reports. Such innovative excellence could be enormously valuable to a global disease surveillance system.
In addition, artificial intelligence (AI) can speed up the research and testing process, qualities that expedite drug and vaccine discovery. Rather than have humans laboriously read scientific studies and deduce which ones have promising properties, natural language processing software can scan research studies, molecular databases and conference proceedings to identify possible new drug candidates.
Yet we need much more than new tools that expedite digital services and product development. We require a plan that invests in novel approaches to help academics and the private sector design applications and bring them to scale. That means infrastructure support and access to large-scale data storage and analysis facilities. One of the problems facing university researchers today is inadequate access to the super-computers and cloud computing required for AI and emerging technologies. Large digital companies have major cloud storage and processing facilities and can offer top researchers high salaries and supercomputer access, making it hard for universities to compete for top talent.
A way to deal with that problem is an idea initiated at Stanford University for a national research cloud that provides computing access to technical experts and academic investigators. In keeping with this idea, the National Science Foundation has started a CloudBank program for agency-funded researchers. It is a way for experts at universities and nonprofits to obtain the same, high-quality access to supercomputers and cloud platforms that is available in the private sector.
The United States also needs to improve access to fair and unbiased data. Having representative data that enable evidence-based algorithms is crucial for the development of the health care area and beyond. Right now, there are no uniform standards for data access or sharing, and this creates particular challenges for AI. Much of the data that go into algorithms are proprietary in nature and not shared with the research community, thereby limiting innovation. Artificial intelligence requires large data sets to test and improve its learning capacities.
The Center for Data Innovation has proposed that federal agencies develop “shared pools of high quality, application-specific training and validation data in key areas of public interest, such as agriculture, education, health care, public safety and law enforcement, and transportation.” That type of information would facilitate pilot testing and help researchers refine AI models. The center cites other countries that require “the private sector to share certain data sets in select circumstances, when it does not threaten a firms’ business and relates to key public interests such as health and safety.”
Finally, we need policies that emphasize workforce development and job retraining specifically geared to a digital economy. We need greater financial support for training data analysts, engineers, machine learning specialists, computer designers and ethicists. Right now, most Americans do not see universities as doing enough to build an AI-based workforce. In a national survey by Northeastern University and Gallup, only 25 percent of respondents believe higher education is providing adequate help in these areas. People say university programs are too expensive, course schedules are inflexible, and admissions too difficult to navigate. Instead, individuals look to employers or industry associations for retraining needs.
Responding to these concerns, Reps. Eddie Bernice Johnson (D-Texas) and Frank Lucas (R-Okla.) have introduced the National Artificial Intelligence Initiative Act designed to boost training resources. The bill would increase government funding by $6.5 billion and provide money for research, training and student scholarships. It proposes more spending for the National Science Foundation, Department of Energy, and the National Institute of Standards and Technology, among other agencies. To take full advantage of emerging technologies, we need a coordinated, comprehensive plan that invests meaningful resources and helps people prepare for the future.
In a country ravaged by the coronavirus, it is vital to have a national plan that positions the United States for long-term leadership. Such a roadmap is crucial to putting the country in a stronger position to fight pandemics, spur innovation and meet international challenges. Without a carefully thought out approach, we are weakening our capacity to deal with COVID-19 and to rebuild the economy. We need more investment, more sophisticated infrastructure, and better trained workers to mitigate the next pandemic and compete effectively in the future.
John R. Allen is the president and Darrell M. West the vice president of governance studies at Brookings Institution. They authored a Brookings Press book, “Turning Point: Policymaking in the Era of Artificial Intelligence.”