In a landmark move to bolster American technological and economic leadership, the National Academies of Sciences, Engineering, and Medicine (NASEM) have recently released a comprehensive consensus study report titled "Frontiers of Statistics in Science and Engineering: 2035 and Beyond."
The report, a yearlong effort of an interdisciplinary nationwide committee of academic, industry and government experts, was chaired by Katherine Ensor, Noah G. Harding Professor of Statistics at the George R. Brown School of Engineering and Computing at Rice University. Lance Waller from Emory University served as the Vice Chair and Brittany Segundo, program officer of the Board of Mathematical Sciences and Analytics, served as the study director.
After assessing the state of statistical and data sciences in the U.S., the study provides a strategic vision and practical recommendations for how the field must adapt and evolve in the next decade to ensure the U.S. remains at the forefront of scientific discovery and innovation. The report argues statistics is not just a support tool for science and technology development but the primary engine driving progress in fields as diverse as artificial intelligence, blockchain, precision agriculture, and public health.
More Than Just Data
Data are integral to progress in all fields of enquiry whether physical or social sciences, medicine or engineering. Statistics has long catalyzed and accelerated discovery and innovation by transforming data into ‘actionable wisdom’ in all fields and forms the foundation for artificial intelligence (AI) and machine learning.
In a world flooded with data, advances in the field provide the tools to: identify bias by recognizing when data is one-sided or misleading; quantify uncertainty by calculating exactly how sure we can be about a prediction, and bridge disciplines by acting as a melting pot where mathematics, physics, and computer science come together to solve real-world problems.
The report also makes a crucial point that as artificial intelligence and machine learning technologies become "interwoven with the fabric of society," statistics provides the reliable pathway to ensure they are:
- Transparent: Understanding why Al algorithms make certain decisions.
- Responsible: Ensuring algorithms don't cause unintended harm.
- Trustworthy: Creating systems that perform reliably in the real world, not just in a lab.
Five Key Pillars of the NASEM Report:
The report outlined the following five critical areas where the discipline must adapt and evolve to ensure America continues to remain competitive in science and technology.
- Advance foundational statistics research: As the research and innovation landscape evolves rapidly, there is increasing demand for new statistical methods and tools. U.S. competitiveness will depend on bolstering foundational statistics research through targeted cross-agency federal funding opportunities as well as increased institutional support in universities and industry. This support is essential to accelerate trustworthy, data-driven scientific discovery, keep pace with modern technological evolution, and maximize positive societal impact by enabling future breakthroughs in science and engineering at the frontiers of statistics.
- Strengthen interdisciplinary collaborations: Across almost every domain, massive amounts of interconnected, complex and multimodal data are now easily available. This requires statisticians to collaborate more closely with domain scientists to develop statistical methods and tools to keep pace with the changing landscape, and to solve complex, multi-dimensional problems of national interest.
- Develop nascent technologies: Statistical analyses of complex data involve iterative procedures requiring substantial computing and environmental resources, making it imperative to develop efficient, decentralized processing technologies and new statistical paradigms to reduce computational burden and ensure improved sustainability, privacy, security, and data sharing. The report recommends increased support from the federal government, as well as academic and industry partners, to help achieve these goals.
- Modernize education and workforce development: To create a pipeline of experts capable of handling "data-rich" applications across national security and healthcare, the report recommends that federal agencies invest in undergraduate and graduate training programs and develop innovative programs that vertically and cross-sectionally integrate students, research trainees, faculty and industry researchers, while also promoting inter-university collaborations.
- Establish public-private partnerships: The report recommends that federal funding agencies expand opportunities to advance science through the inclusion of statistics in current and emerging public–private partnerships. Historically, such partnerships have led to breakthroughs in medicine, health care, agriculture, energy and the discovery of new materials, as well as advances in finance, insurance and weather prediction. They have also supported the development of more resilient infrastructures and communities. Continued investment can drive future innovation and help ensure sustained U.S. competitiveness.
“As the global landscape is increasingly dominated by new and emerging data-driven technologies, we hope our study will provide a clear path forward for academic and industry leaders and policymakers to invest in the "science of uncertainty" to ensure a certain future for American science and innovation,” Ensor said.