The outgoing president of the US National Academy of Sciences, geophysicist Marcia McNutt, spoke out on several key issues during her final annual State of Science Address on 2 June. She talked about the current instability facing federal science agencies, the importance of overhauling the promotion and tenure system, and the benefits that artificial intelligence (AI) can bring.
This year has been ‘filled with turmoil’ on the science policy front, just like last year, including ‘a great deal of uncertainty in federal support for science,’ McNutt told the audience. She thanked Congress for standing up for science and rolling back the dramatic budget cuts to federal research agencies proposed by the Trump administration earlier this year. Her second five-year term as head of the NAS ends on 30 June and University of Chicago evolutionary biologist Neil Shubin will succeed her.
McNutt further lamented what she called ‘the abrupt downsizing’ of US science agencies, referencing estimates by the Partnership for Public Service and others that almost 100,000 employees at US science agencies have departed from late 2024 to mid-2026. She says this has left them ‘very lean and unable to fulfil their missions’.
The fact that US universities are having greater difficulty recruiting and retaining top science, technology, engineering and maths (Stem) talent is another concern McNutt raised. Recent National Science Board (NSB) analysis from May concluded that international student enrolment in science and engineering fields at US higher education institutions dropped 9% this past autumn, compared with the previous year. ‘We always were the country where Stem talent came to us, now we are exporting our science talent elsewhere,’ McNutt warned.
Rethinking tenure
To address many of these issues, she said it is critical to realign the academic reward system by revamping the traditional promotion and tenure process at US universities. McNutt pointed to several ongoing pilot projects, including one at Worcester Polytechnic Institute in Massachusetts that redesigned the school’s promotion criteria to move the focus from citations to the impact of scholarly contributions. Following the policy change, Worcester saw three times more faculty promoted overall and six times more women promoted, McNutt said.
During the panel discussion that followed, physical chemist Geri Richmond, who serves as a vice president of research and innovation at the University of Oregon and was formerly president of the American Association for the Advancement of Science, echoed her sentiments. ‘We have to think of ways to do things differently,’ Richmond stated. ‘We have to change the tenure and promotion process; it has stalled us for way too long,’ she continued. ‘We have to define what impact is and decide how we’re going to measure impact broader than h-indices and number of publications.’
Richmond served as under-secretary for science in the Department of Energy during the Biden administration and sat for a decade on the National Science Foundation’s governing body under President Barack Obama and during Trump’s first term. In April, the Trump administration fired all 22 members of that oversight board.
During her speech, McNutt also discussed the importance of reducing research regulations. She referenced analysis by the Council on Government Relations that shows 66% of regulations in this arena that have been issued since 1991 were enacted in the last decade, and pointed out that PIs at US research institutions are currently spending more than 40% of their time on regulatory work. ‘If these regulations increase, we’re talking about PIs not actually having any time for research; they will be filling out paperwork,’ McNutt stated.
Could AI help reduce paperwork overload?
But she suggested that AI can help address this issue by reducing duplicate paperwork that can hinder PIs from seeking funding from multiple agencies and by increasing innovation. ‘Imagine if you could recoup 44% of your time, how amazing you could be,’ McNutt stated, arguing that it would yield greater job satisfaction among academic researchers, increased regulatory compliance and facilitate greater discoveries.
‘I think we could increase the rate of innovation with the help of automation in shared facilities,’ McNutt stated. ‘My proposal here is to use AI and robotics to increase the rate of discovery.’ She noted that startup costs for individual labs are multi-millions of dollars and said graduate training also needs to be decoupled from lab staffing needs for the many students not pursuing academic careers.
To help make her case, McNutt pointed to a new robotics innovation centre at Carnegie Mellon University in Pennsylvania, which is a shared facility that combines autonomous experimentation, cloud data platforms and translational research.
She suggested that this type of shared infrastructure could reduce researcher bias, enhance morale among researchers and strengthen scientific outputs. ‘More samples will increase the reliability and trustworthiness of results,’ McNutt added. ‘Robots work 24-hours a day, your graduate students probably only work about 20!’
During the panel discussion after her talk, Ethan Klein, the Trump administration’s chief technology officer and associate director of the White House Office of Science and Technology Policy (OSTP), echoed McNutt’s AI enthusiasm.
Klein, who says he was trained as an undergraduate in physical chemistry and served in OSTP during the first Trump administration before pursuing a PhD in nuclear science at the Massachusetts Institute of Technology, stated that the Trump administration is ‘ushering in this new era of AI-enabled scientific discovery’.
He said the White House has been working to bring the various US government R&D funding agencies and departments together to identify long-term national science and technology challenges that can be solved with ‘AI-enabled workflows’ and ‘autonomous scientific experimentation’.
While Klein acknowledged that the federal government ‘plays an incredibly important role in funding foundational, basic science discovery research’, he suggested that ‘there are many places that you can do basic science today’. Klein also emphasised the importance of what he called ‘use-inspired’ basic research. ‘There are areas where we are seeing important innovations coming up … there can be some political direction with the technologies that we know are important for our society,’ he stated.
However, Klein did not directly address concerns, raised at the forum by McNutt and others, about a new White House proposal that would give political appointees power over federal research grant funding decisions.
No comments yet