National Academy of Sciences president advocates for a national research strategy to coordinate basic research

In a recent State of the Science address, National Academy of Sciences President Marcia McNutt called attention to what she sees as an alarming trend: the federal contribution to basic research is stagnating while industry funding is rising. The federal share of U.S. spending on basic research, she said, was down from about 60% in 2011 to 40% in 2021. The fall is mostly the result of a sharp increase in basic research funding from industry (see this SSTI article and this National Science Board report for more details on this topic). McNutt proposed that a national research strategy should strive to reverse this trend so that the federal government regains its leadership position in funding basic research.

McNutt’s reasoning is that ultimately, industry-funded basic research will not be the best way to maintain U.S. leadership in innovation. She cautioned that industry primarily engages in basic research that serves the company’s own interests and is not often shared with researchers outside of the company. She noted that while industry performs 75% of the R&D in the U.S., it only accounts for 6% of the publications, attributing the low number of publications to their need to protect proprietary information.

Of particular concern, she said, is the dominance technology companies have on artificial intelligence (AI) R&D. Contributing to the issue is the fierce competition for the available AI research talent: McNutt noted that industry’s ability to offer more attractive compensation packages has resulted in 65.7% of all AI researchers working in the private sector  (see this paper from Science and this article from The Wall Street Journal for more on AI brain drain from academia). AI research, she says, will be focused on developing products for industry while innovations that might benefit society go unaddressed.

There was some pushback to McNutt’s viewpoint from panelist James Manyika, senior vice president of Google and Alphabet, who oversees Google research. He noted that Google is doing basic research that benefits scientists, citing as an example their work with AlphaFold, an artificial intelligence system that 1.8 million biologists in 190 countries have accessed to help them to understand proteins. He gave a second example of work they’re doing on Connectomics (an emerging area of neuroscience research that focuses on mapping and analyzing the connections within the brain), which has been a decade long investment. Google research collaborates with the Howard Hughes Medical Institute, Harvard, and others on that project. Manyika also described Google Research’s work on quantum science. According to Manyika, Google-Alphabet is funding $100 million for that research over the next few years.

Panelist Dr. E. Albert Reece from the University of Maryland School of Medicine cited an unpublished National Academy of Medicine report that calls for a national research strategy that is more coordinated among government, industry, philanthropy and academia, saying that such an approach could fill gaps left behind from the research community’s narrow focus on such high-profile (and profitable) diseases such as cancer, heart disease, and HIV Aids.

While advocating for a national research strategy, McNutt cautioned not to “close out opportunities for serendipity” by eliminating discoveries that come out of basic science. She gave as an example the field of quantum physics, which was a “scientific curiosity” until it became the foundation of the information age.

A recording of the State of the Science Address is available here.