RAND Corporation, 2019

From the report: Recommendations for Policymakers, Practitioners, and the Public — Consider the full range of ML competencies. One clear observation that emerged from our analysis is that the ML field pulls from a nuanced body of knowledge, one that was in development well before recent increases in attention to the issues of misinformation and disinformation. As a result, we recommend that practitioners and policymakers turning to ML as a response to these challenges consider not only the narrow aspects of ML that appear immediately relevant (e.g., fact-checking, searching online) but also the full body of evidence that exists about the relationships between individuals and the information ecosystem. Increasing the lines of communication between practitioners and researchers can contribute to this work. 

Engage diverse constituencies in scaling ML education. The scaling of ML efforts is both necessary and difficult. One way to scale ML efforts, it seems, might be to use other agents of ML in addition to teachers, and other forums for ML in addition to schools. This might mean enlisting community members—pastors or rabbis, librarians, parents, or professional mentors can serve as agents of ML in both formal and informal ways. Although we need more research on what works when it comes to improving ML, pursuing this decentralized approach to ML education might be one way to achieve the “herd immunity” that the experts we interviewed pointed toward. We note that although this recommendation is aimed first at policymakers and practitioners, it is also a recommendation with important implications for the general public. If scaling ML relies on the grassroots efforts of individuals, then parents, coaches, bosses, mentors, and friends could play an important and central role in expanding the reach and success of ML education. Read the full report here.