Leveraging Machine Learning To Reduce Cost & Burden of Reviewing Research Proposals at S&T Agencies
Summary
With about $130 billion USD, the United States leads the world in federal research and development spending. Most of this spending is distributed by science and technology (S&T) agencies that use internal reviews to identify the best proposals submitted in response to competitive funding opportunities. As stewards of quality scientific research, part of each funding agency’s mission is to ensure fairness, transparency, and integrity in the proposal-review process. Manual proposal review is time-consuming and expensive, costing an estimated $300 million annually at the National Science Foundation alone. Yet at current proposal-success rates (between 5% and 20% for most funding opportunities), a substantial fraction of proposals reviewed are simply not competitive.
The next administration should initiate and execute a plan to advance machine learning to triage scientific proposals. This proposal presents a set of actions and a vision to diffuse machine-learning across science and technology agencies to ultimately become a standard component of proposal review, while improving the efficiency of the funding process without compromising the quality of funded research.
Science funding agencies are biased against risk, making transformative research difficult to fund. Forecast-based approaches to grantmaking could improve funding outcomes for high-risk, high-reward research.
With a PhD in materials science, a postdoc position at the National Institute of Standards and Technology, and a stint as a AAAS Fellow, Dr. Shawn Chen has had a range of roles in the research community.
Like climate change, the societal risks from AI will likely come from the cumulative impact of many different systems. Unilateral commitments are poor tools to address such risks.
Despite the uphill battle the country is facing, Dr. Schlaerth feels optimistic about the future possibilities of industrial decarbonization.