In comments filed last week, the American Medical Informatics Association (AMIA) strongly encouraged the Trump administration to leverage an existing strategy on Artificial Intelligence Research and Development (AI R&D), rather than begin from scratch. The nation’s leaders in health informatics and health data science also emphasized a need for federal investments in areas of strong societal importance, not aimed at consumer markets, especially related to human-AI collaboration and the ethical, legal, and societal implications of AI, including malicious AI.
In September, the Trump administration issued a Request for Information (RFI) to see whether stakeholders recommended updating the 2016 National Artificial Intelligence Research and Development Strategic Plan. This 2016 AI R&D Plan established several strategic categories and specific strategic aims for federal R&D investment. It also included two recommendations to develop an implementation framework for AI R&D, as well as study ways to bolster and sustain an AI R&D workforce. The Trump administration additionally established a Select Committee on AI under the National Science and Technology Council, which is tasked with coordinating federal efforts related to AI and ensured continued U.S. leadership in AI.
AMIA noted that much of the 2016 Strategic Plan remains relevant, including the need to make long-term investments in AI research and the need to develop better methods to measure and evaluate AI through standards and benchmarks. The group urged the Select Committee to deliver on the 2016 AI R&D Plan’s two primary recommendations to develop an implementation framework, and study what is needed to sustain a robust AI R&D workforce.
“The 2016 Plan articulated a vision that must be implemented,” the group noted in its comments, “and the Select Committee should have a framework to understand how federal investments in AI R&D have aligned with the 2016 Plan since its release, as well as how such investments should be modified moving forward.” AMIA noted the recently released NIH Data Science Strategic plan and a 2017 JASON report on the impact of AI in healthcare as evidence that the federal government needs to coordinate various Cabinet-level Department efforts in this space.
AMIA went on to discuss the national strategy through the prism of health and argued for federal investments looking at the interaction between humans and AI, as well as investments in ethical AI. “In medicine, we tend to frame AI as ‘augmented intelligence,’ given that there is surely no better example of a scientific discipline so enmeshed with and influenced by the human condition. Given this view, the art and science of medicine will surely be impacted greatly by AI,” and “questions regarding how clinicians interact with AI or how AI will influence clinical decision-making, represent daunting challenges for which federal R&D funding should be leveraged.”
Citing a recent House Subcommittee report on the subject, AMIA also commented that federal research should focus on bias and possible malicious use of AI. It further noted that the danger that AI may be used for nefarious purposes is no different from any other technology, but that “the possibility that a malicous goal may be built into an AI progam, leading it to make subtle undermining decisions is of a different nature and should remain a source of concern.”
Finally, AMIA encouraged the Select Committee to develop a workforce strategy that promotes general-purpose AI R&D experts, as well as AI R&D experts with more specialized knowledge in a specific area of application, such as health data science. “Domain-independent AI experts will be necessary to advance R&D, but as is the case with a strategy that goes unimplemented, research that goes unapplied is useless. We believe the best AI R&D work will come from those scientists who are knowledgeable in specific domains and are able to make appropriate decisions about how to apply their craft.”
###
AMIA, the leading professional association for informatics professionals, is the center of action for 5,400 informatics professionals from more than 65 countries. As the voice of the nation’s top biomedical and health informatics professionals, AMIA and its members play a leading role in assessing the effect of health innovations on health policy and advancing the field of informatics. AMIA actively supports five domains in informatics: translational bioinformatics, clinical research informatics, clinical informatics, consumer health informatics, and public health informatics.