Michael Stebbins
(Science Advisors)
Leslie McIntosh
(Digital Science)
Stuart Buck
Format:
Pre-conference virtual symposium
The Good Metascience Project: Goal setting and emerging technologies to make science funders more impactful .
Virtual Symposium VS08 at conference Metascience 2025.
This symposium will propose research agencies set impact goals closely aligned to their missions, and the development AI tools using an integrated database of US federal research projects, including research applications, to accelerated discovery and forecast the future of science.
Description
This 90 minute panel and audience discussion presents highlights from the Good Science Project R&D Reform recommendations. After a brief introduction by Stuart Buck about this policy effort, Neil Thakur will outline the concept of setting agency wide funding goals aligned with the impact portion of the a funder’s mission (e.g. “improve health, extend life, and reduce illness and disability” for NIH, “Advance the national health, prosperity and welfare [and] secure the national defense” for NSF) and discuss how they can lead to investment changes within specific portfolios, across multiple areas of science, and across multiple time scales.
Leslie McIntosh will describe the broad array of new portfolio selection, management, and evaluation tools that could be enabled by applying AI techniques to an integrated database of all federal research projects. She will outline the benefits of this data integration, and how the enabled analytic capacity could drive benefits including accelerated discovery and collaboration, increased efficiency and accountability, and reduced waste.
Michael Stebbins and Neil Thakur will present a more specific use of AI tools by applying them to federal grant applications. They will describe how these data can predict the future of scientific and technological breakthroughs, enhance peer review, and encourage better research investments by both the public and the private sector. All presentations will be completed within 45 minutes, followed by audience discussion lead by Stuart Buck.
Leslie McIntosh (Digital Science)
Stuart Buck
Short Abstract
This symposium will propose research agencies set impact goals closely aligned to their missions, and the development AI tools using an integrated database of US federal research projects, including research applications, to accelerated discovery and forecast the future of science.
Description
This 90 minute panel and audience discussion presents highlights from the Good Science Project R&D Reform recommendations. After a brief introduction by Stuart Buck about this policy effort, Neil Thakur will outline the concept of setting agency wide funding goals aligned with the impact portion of the a funder’s mission (e.g. “improve health, extend life, and reduce illness and disability” for NIH, “Advance the national health, prosperity and welfare [and] secure the national defense” for NSF) and discuss how they can lead to investment changes within specific portfolios, across multiple areas of science, and across multiple time scales.
Leslie McIntosh will describe the broad array of new portfolio selection, management, and evaluation tools that could be enabled by applying AI techniques to an integrated database of all federal research projects. She will outline the benefits of this data integration, and how the enabled analytic capacity could drive benefits including accelerated discovery and collaboration, increased efficiency and accountability, and reduced waste.
Michael Stebbins and Neil Thakur will present a more specific use of AI tools by applying them to federal grant applications. They will describe how these data can predict the future of scientific and technological breakthroughs, enhance peer review, and encourage better research investments by both the public and the private sector. All presentations will be completed within 45 minutes, followed by audience discussion lead by Stuart Buck.