Shobita Parthasarathy

@shobitap

Science, tech, and innovation politics/policy incl patents. Author: next book on politics of inclusive innovation in India. #STS. Prof & Director, Science, Technology, and Public Policy Program, U of Michigan. Expert Voices columnist, Science. shobitap.org

152 Siguiendo5167 Seguidores1353 Publicaciones

Bluesky

¿Es tu perfil?Reclámalo y shobitap.dirty.social es tuyo.Reclamar perfil

Shobita Parthasarathy@shobitap· 18d

I spoke to STAT about the new Trump admin/OSTP/Kratsios strategy and why it's likely to harm science and public trust in scientific institutions.

Kratsios’ vision for science claims to make science more responsive to the needs of the public by making it more directly tied to industry. But critics argue this misses the value of publicly funded, curiosity-driven science — which can tackle lines of inquiry industry will not and open avenues of innovation that are difficult to predict.

“One of the big problems is that we have not adequately centered social needs, social priorities, and as a result, that has led scientific institutions to be out of step. … But this actually increases that problem,” said Shobita Parthasarathy, director of the science, technology, and public policy program at the University of Michigan. She pointed out that Kratsios’ report focuses on AI — despite a recent survey showing that 59% of adults don’t trust AI companies to develop and use the technology responsibly and 40% said they think AI will have a negative impact on society (16% think it will be positive).

Shobita Parthasarathy@shobitap· 19d

Having just returned from IAS, whose founding director imagined it “eschews the chase for the useful, the minds of its scholars will be liberated, they will be free to take advantage of surprises, and someday an unexpected discovery, apparently leading nowhere", Kratsios is doing the opposite.

Throughout history, Americans have developed various ways to support scientific progress that depart from the university-based, principal investigator-driven grant. These include Cold Spring Harbor Laboratory, a leading institution of biological research, founded in 1890, and the Institute for Advanced Study, founded in 1930, which gave luminaries like Einstein, Noether, Oppenheimer, and Gödel an opportunity to explore revolutionary ideas. Each provided a home for new kinds of science.

Shobita Parthasarathy@shobitap· 19d

This is such a deficient understanding of accountability, which also infects the proposed OMB rule. Science dictated by presidential administration & conducted mostly in (or in collaboration with) the private sector does not produce challenging discovery or solve the "entrenched interests" problem.

While repeating the mantra that science must be “independent” of politics, parts of academia have become highly dependent on government funding. But as Bush understood, federal support for science must be politically accountable. Accountability does not mean dictating how a research agenda is to be executed, nor turning away from the basic research that has long been the wellspring of American
prosperity. Quite the opposite. It means ensuring that the system serves the researchers who are its lifeblood, rather than the entrenched interests that have accreted around them. Only those parts of the national research enterprise directly responsive to the political process can prioritize among the many potential avenues of inquiry, fund those that best reflect national priorities, and be checked when the scientific process breaks down. In a self-governing nation of laws and citizens, the Federal Government elected by the people shall have the ability to
determine how to allocate public resources in the public interest.

Shobita Parthasarathy@shobitap· 19d

Ouch. He doesn't explicitly say it, but I think the NSF Broader Impacts criterion is not long for this world. That's so sad. While it is not the tool many had hoped, the point is to force investigators to think about the social impacts of their work (it existed long before "DEI"). That's important.

$Our up-and-coming scientist is further discouraged by the corrosion of the merit principle that once made American research the envy of the world. What began with the NSF’s “broader impacts” criterion, which was a reasonable effort to ensure taxpayer funds benefit society, has evolved into a sweeping distortion of the selection process that reduces mobility for the best researchers. Between 2021 and 2024, the share of new NSF grants focusing on diversity, equity, and inclusion initiatives surged from fractions of a percent to more than a quarter. Our scientist has learned the new rules from principal investigators she has worked for, and is tempted to dress up her otherwise excellent technical proposals with ideological language to survive review. Until recently, NASA required research proposals to include plans for furthering “inclusion goals,” which were to be reviewed by review panels one-half composed of “diversity, equity, and inclusion professionals.”$

Shobita Parthasarathy@shobitap· 19d

In broad strokes, I also think that we should be funding science of science research. But "metascience" as Kratsios defines it is EXTREMELY narrow, and it's mostly about whether the science efficiently/effectively leads to $$$. Rather, we should be thinking about societal outcomes and benefits.

Establish Metascience Capabilities. Agencies should establish metascience capabilities that evaluate what programs actually work and drive organization-wide reforms. Core responsibilities should include conducting research on how factors such as funding mechanisms, peer review, and publication practices affect scientific outcomes; piloting novel funding mechanisms and institutional models; and evaluating pilots and informing agency-wide portfolio management. These functions should be established at a sufficiently high level within agencies to effect real, cross-agency change.

Shobita Parthasarathy@shobitap· 19d

Back to nativism and DEI. Diversity enhances creativity, and a diverse scientific workforce asks questions that ensure that science serves *everyone*. One of my favorite (extremely depressing) studies: www.science.org/doi/10.1126/...

Identify Exceptional Talent Nationwide. S&T workforce programs should
identify and support all talented Americans across geographies, incomes, and demographics. Exceptional talent is defined by demonstrated technical ability, not background or identity. Agencies should therefore anchor selection processes in criteria predictive of STEM success, such as reasoning assessments, domain competitions, engineering portfolios, and technical work, rather than relying on self-selection, essays, institutional referrals, or polish and credentials. Agencies should leverage merit-based identification mechanisms that cover as many people as possible (e.g., SAT scores or other standardized quantitative assessments) to find overlooked talent.

Shobita Parthasarathy@shobitap· 19d

Honestly, this reminds me of the awful job that NSF and others have done over the years in defending why basic research is important beyond sparking tangible results. What he describes here as basic research is really just proto-innovation, The ARPAs are not known for discovery!

Support a Diverse Portfolio of Institutions. The Federal Government
should reflect a portfolio of institutions that collectively advance the core
objectives of the Nation’s S&T enterprise, including conducting a range of scientific work, training the next generation of scientists, and translating scientific discoveries into concrete benefits for Americans. Agencies should identify objectives that remain unaddressed because no existing institution is well-suited to pursue them. One notable gap is agile, mid-scale science: infrastructure-heavy, multidisciplinary basic research that requires coordinated teams of ten to a hundred people. Agencies have begun to address these gaps through new models like the U.S. National Science Foundation’s (NSF) X-Labs and certain Advanced Research Projects Agency programs. Agencies should consider these models and experiment with additional designs to enable new types of scientific pursuits.

Shobita Parthasarathy@shobitap· 19d

At a moment of real loss of trust in scientific institutions, and extreme public skepticism of AI, does this funding strategy really seem like a good idea? Especially as people are concerned that science and innovation are not benefiting them, or asking the questions that really concern them?

Many of the Administration’s strategic technology priorities, including AI,
quantum information science, semiconductors, advanced communications, robotics, advanced manufacturing, nuclear fission and fusion, and space systems, all rely on foundational research across the physical sciences, computer science, and engineering. However, the physical sciences and engineering have declined as a share of the Federal research portfolio over an extended period, even as the strategic importance of these fields has grown. Agencies are encouraged to prioritize both the absolute level and the relative share of funding directed within budget guidance levels to the physical sciences (physics, chemistry, materials science, space science, etc.), computer science, and supporting engineering and mathematical disciplines, especially within national security-relevant research portfolios. In addition, to support Administration priorities in biotechnology and biomanufacturing, agencies should prioritize foundational research in the biological sciences over the life sciences, a broader category not focused on foundational research.

Shobita Parthasarathy@shobitap· 19d

One of the unspoken contradictions in the document is between science and innovation (by which he largely means economic competitiveness). To benefit science, you need openness and international collaborations. To benefit markets, you need control, secrecy, and intellectual property. RIP science.

The United States conducts research openly, publishes freely, and shares methods transparently. These values are central to scientific progress. However, there is a critical difference between sharing information on one’s own terms and being exploited; the parts of a scientific enterprise that are not shared are often its comparative advantages. These are precisely what our competitors work hardest
to extract, leveraging human capital educated in our universities, harvesting data from our papers while restricting access to their own, and scaling breakthroughs first achieved in our laboratories on their factory floors.

Shobita Parthasarathy@shobitap· 19d

Wow, did not realize that Kratsios was such a comedian. Love strategic use of openness and credit when discussing AI. "Open" except for data and models, "granular credit" in terms of completely false citations (not to mention the fact that studies show AI seems shift folks away from creativity).

Experiment With AI-Native Scientific Institutions: Today’s funding
structures, publication systems, and credit mechanisms were built for a
world of human-paced discovery. Begin the transition to AI-native institutions, including through faster and more open forms of scientific publication, more granular credit attribution, and new market mechanisms that direct resources to problems where breakthroughs matter most.

Shobita Parthasarathy@shobitap· 19d

HAHAHAHA. Who's going to tell him that the Human Genome Project was, in fact, an *international* partnership? There are some great books he could read to learn about it, but those were written by 😱 social scientists and historians. Or he could check out that Clinton/Blair press conference.

Organize Pre-Competitive Consortia and Grand Challenges: The Apollo
Program and the Human Genome Project succeeded because the Federal Government marshaled scientific effort at a scale no single institution could match. Leverage grand challenges that pull breakthroughs forward, and create moonshot-scale missions for issues of national importance. Support industry consortia and use federal resources to break shared engineering bottlenecks in foundational areas, as Extreme Ultraviolet Limited Liability Company (EUV LLC) did for semiconductor lithography.

Shobita Parthasarathy@shobitap· 19d

Not surprisingly, this document is all about funding industry. There is no discussion at all about ensuring transparency and accountability in these relationships, or managing conflicts-of-interest (even as it mentions the "gold standard" EO). The lesson seems to be industry good, academia bad.

Strengthen Public-Private Partnerships and Talent Flows: The university
is no longer the only home of America’s most innovative scientific research. Expand agency-adjacent foundations, focus Small Business
Innovation Research (SBIR) and Small Business Technology Transfer
(STTR) programs to build strategic capabilities, and support joint centers
among industry, academia, and federal facilities. Scale industry Ph.D. and
postdoc fellowships that move talent fluidly between sectors, drawing on
America’s private sector strengths to bring industry-scale resources to our academic researchers.

Shobita Parthasarathy@shobitap· 19d

Given the audience, I shouldn't be surprised that there is quiet nativism across the document. There are no citations here, of course. Instead of treating foreign PhD students as a problem, how about understanding universities as magnets that bring foreigners who become very productive citizens?

A smaller proportion of American citizens now fill
post-graduate spots in science, technology, engineering, and mathematics (STEM) fields. Our competitors are channeling unprecedented resources into science and engineering, taking a whole-of-society approach to seize the high ground in strategic technologies. The AI revolution, meanwhile, is transforming the conduct of science, and legacy scientific institutions and infrastructure are not ready to take full advantage of this transformation.

Shobita Parthasarathy@shobitap· 19d

Seems he grabbed the wrong end of this particular stick. The fact that industry funding is huge does not mean that federal funding should add to it. Shouldn't it be evidence that the govt should fund what industry *won't*, to expand the possibility of discovery and innovation? Or fund across fields?

Government funding of research and development, especially basic science in the academy and national labs, has rightly grown in the eight decades since Bush’s report. But private industry has become by far the largest source of research and development (R&D) funding in the United States, with its share nearly doubling from the 1950s to today. American companies now deploy around $700 billion annually, more than triple the combined spending from government and higher education. This evolution has made the pie bigger for everybody and should be welcomed across the research ecosystem, but it demands a corresponding adjustment to the nature of the Federal Government’s contributions.

Shobita Parthasarathy@shobitap· 19d

Despite what he says, this is a remarkably narrow scope. Obviously, no social scientists. But what about basic researchers? Evolutionary biologists? Theoretical physicists? Psychologists? Science philanthropies? Pharmaceutical companies? (Wow, I never thought I would be lobbying for pharma).

In preparing this report, I have consulted widely across the country’s scientific enterprise, from entrepreneurs driving world-changing breakthroughs in fusion technology, to researchers eager to pursue bold ideas in neuroscience at our universities, to venture capitalists funding revolutionary platforms for drug discovery. Their message is clear and inspiring: given the right conditions, American ingenuity will continue to achieve the impossible.

Shobita Parthasarathy@shobitap· 19d

It's very clear that Kratsios/Trump wants to destroy universities (and the indirect costs they receive), or, what he calls "legacy institutions". It seems that there is very little understanding of how current science works, particularly the massive infrastructure it needs. Who's gonna pay for that?

First, the U.S. research system should prioritize the individual scientist over legacy institutions. If we are serious about expanding what our scientists can do, we must invest directly in American researchers and the bold ideas that drive them. Too much of our research enterprise has come to serve itself rather than the scientists within it. Federal funding agencies should support a broader range of performers, including a new generation of mission-driven research organizations, to expand our scientific horizons. And these agencies should organize themselves
around the interdisciplinary frontiers of science today, rather than the academic silos of the last century.

Shobita Parthasarathy@shobitap· Jul 8

Ahem...what could possibly go wrong??? dl.acm.org/doi/pdf/10.1...

Abstract for: Patent Applications as Glimpses into the Sociotechnical
Imaginary: Ethical Speculation on the Imagined Futures of
Emotion AI for Mental Health Monitoring and Detection
NADIA KARIZAT, University of Michigan, USA
ALEXANDRA H. VINSON, University of Michigan, USA
SHOBITA PARTHASARATHY, University of Michigan, USA
NAZANIN ANDALIBI, University of Michigan, USA

Shobita Parthasarathy@shobitap· Jun 30

🚨New paper!🚨 Divya Ramesh, Nikola Banovic, and I argue that current discussions of AI accountability are too tech-solutionist. Using case studies from the global South, we advocate for a more grounded approach that takes social and political dynamics seriously. dl.acm.org/doi/10.1145/...

The Paradox of ‘Algorithmic’ Accountability and its Implications for
Design

DIVYA RAMESH∗, University of North Carolina at Charlotte, USA and MBZUAI, UAE
SHOBITA PARTHASARATHY, University of Michigan, USA
NIKOLA BANOVIC, University of Michigan, USA

Accountability calls for clearly defined relationships and responsibilities of different stakeholders to ensure benefits from AI and algorithmic systems are distributed equitably. While stakeholders across academia, industry, and civil society have championed mechanisms like audits and impact assessments to achieve accountability, such ideas have met limited success, suggesting that the accountability discourse is often captured by “solutionist” tendencies. Here, we treat algorithmic accountability as a discursive object to examine these tendencies. Employing a critical discourse analysis of 240 scholarly documents in the
ACM digital library, we find that while the discourse has varied meanings and aims that embody ideals of social change, it remains limited by computing-centric relations and interventions. By reflecting on two case studies from our Responsible AI research, we show that solutionism in the accountability discourse creates a narrow focus on AI and algorithms in three ways: it occludes multi-stakeholder and reciprocal dynamics (e.g., ignoring questions of accountability beyond the algorithm),
prioritizes postmaterial concerns (like bias and fairness) over the tangible and urgent requirements of communities, and flattens power relations that are deeply situated in cultural contexts. We propose a framework that shifts the focus to the politics of knowledge and expertise to help cultivate accountability sensibilities in the pursuit of AI innovation.

Shobita Parthasarathy@shobitap· Jun 30

For those of you following the proposed OMB rule putting political appointees in charge of grant funding, note that NSF is *seeking comment* to change its procedures in alignment with this proposal (from FYI Science Policy). Please submit a comment; you can use our guide below to help you!!

Shobita Parthasarathy@shobitap· Jun 5

Apparently there's a new industry of AI slop books about AI! Alexander H. Blackwell has produced 15 books between May 14th and June 4th. And Quentin G. Dalton produced the same number even more quickly, between May 8th and 28th! If only!

List of books supposedly written by Alexander H. Blackwell, about AI and digital technologies.

Another list of books supposedly written by Alexander H. Blackwell, about AI and digital technologies.

Shobita Parthasarathy@shobitap· Jun 2

Hmmm. Looks like Rogers has Abdul envy.

Picture of US Senate candidate Abdul El-Sayed, who is muscular.

Shobita Parthasarathy@shobitap· Apr 20

Deeply moved by Didier Fassin’s final lecture at the Institute for Advanced Study, where he argued for the importance of the “right to search for truth” wherever it leads, courage in academia, and noted that critical thinking is crucial, and is a unifier across the sciences.

Shobita Parthasarathy@shobitap· Apr 19

In case it is useful, here is what I put in my letter.

Text of letter to Senator Gary Peters, opposing the proposed dissolution of the National Science Foundation's Social, Behavioral, and Economic Sciences Directorate.

Shobita Parthasarathy@shobitap· Mar 28

Ready to go in Philly!!

Shobita Parthasarathy@shobitap· Mar 6

Thinking about, or engaging in, politics and public policy with an #STS lens? Apply to present your work in the open panels (#66) I am organizing with Gwen Ottinger and Jason Delborne @4sweb.bsky.social in Toronto in October! (Paper abstracts due April 30th!) www.4sonline.org/accepted_ope...

Panel #66: Engaging Policy in STS Research and Practice

Abstract: Throughout its history, the STS community has engaged with the important policy questions of the day, showing how public issues are constructed in tandem with evidenceand expertise, and how technological infrastructures are inseparable from social systems.
STS scholars have worked to have their insights recognized by policy makers and have used their direct policy engagements to advance knowledge of policy processes. The past decade, however, has seen massive shifts in the policy landscape–including in the ways that
research is funded, evidence is valued, and expertise is constructed–as well as new substantive issues at the intersection of science, technology, and policy. This panel asks what STS scholars are bringing to, and taking away from, their engagements with evolving policy processes. What STS tools best help us understand policy issues, and where do new circumstances require new theories and analytical methods? How can STS analysis not only advance knowledge but shape effective political action? And how does engagement with policy function as an intellectual project in its own right? This open panel invites submissions from across sectors (e.g., environment, innovation, health, criminal justice), and with historical and global perspectives (particularly outside the United States and Europe).

Shobita Parthasarathy@shobitap· Feb 5

Great. Federal employees will no longer be able to provide expert assessments or advice without fearing for their jobs.

The Trump administration finalized a new policy on Thursday that would strip job protections from up to 50,000 federal workers, a move that would make it easier for President Trump to remove or discipline them, in his latest effort to dismantle the federal work force.

Until now, the roughly 4,000 people appointed by the president, known as political appointees, were the only federal workers who could be fired at will. The policy issued on Thursday allows the administration to expand that number to include career employees whom the administration considers to also have policy-related roles. For these employees, any whistle-blower complaints would be handled by their agencies internally and not the independent office of special counsel as it has been in the past.

Shobita Parthasarathy@shobitap· Jan 24

THIS is how democracies die. The point is that this administration *doesn't care* that we know they are lying. They no longer care about trust in institutions and legitimacy, and have moved to governance by brute force and fear.

Videos on social media that were verified by The New York Times contradict the Department of Homeland Security’s account of the fatal shooting of a man by federal agents in Minneapolis on Saturday morning.

The Department of Homeland Security said the episode began after a man approached Border Patrol agents with a handgun and they tried to disarm him. But footage from the scene shows the man was holding a phone in his hand, not a gun, when federal agents took him to the ground and shot him.

Shobita Parthasarathy@shobitap· Jan 13

Another data point against AI: apparently I have a Grokepedia entry, and it is riddled with errors (and contradicts itself) even though it was "fact-checked by Grok 3 weeks ago". 😱 (While my Wikipedia entry is a little out of date, none of it is flat wrong.)

Inaccurate Grokepdia entry that claims, for example, that my first cook compared US and Indian approaches to genetic testing for breast cancer (it does not).

Grokepedia text: Shobita Parthasarathy is an American professor of public policy and women's and gender studies at the University of Michigan's Gerald R. Ford School of Public Policy, where she serves as director and co-founder of the Science, Technology, and Public Policy Program.[1][2] Her research centers on the comparative politics of science and technology governance, including the social implications of biotechnology, intellectual property regimes, emerging technologies such as artificial intelligence and gene editing, and challenges related to equity, inequality, and public trust in scientific innovation.[1][3] Parthasarathy is the author of two books—Building Genetic Medicine: Breast Cancer, Technology, and the Comparative Politics of Health Care (MIT Press, 2012), which compares U.S. and Indian approaches to genetic testing for breast cancer, and Patent Politics: Life Forms, Markets, and the Public Interest in the United States and Europe (University of Chicago Press, 2017), which critiques the political influences shaping patent systems for biotechnological inventions.[2][4] She has contributed to policy discussions, including congressional testimony on science policy, and co-hosts the podcast The Received Wisdom, which examines governance strategies for science and technology to address global challenges like inequality and expert distrust.[5][6] Currently, her ongoing projects include books on failures in technological innovation and strategies to enhance public trust in science, alongside recent publications in journals such as Science critiquing the uncritical scaling of technologies.[1][3]

This text inaccurately says I co-founded and directed STPP in 2015, and was a lecturer at UChicago (I was not).

Text: In September 2015, Parthasarathy co-founded and became Director of the Science, Technology, and Public Policy (STPP) Program at the University of Michigan, a role she continues to hold.[7][2] Under her direction, the program has developed initiatives such as the Technology Assessment Project, which produced reports on topics including large language models.[2]

Prior to her Michigan appointment, Parthasarathy served as Lecturer in the Department of Political Science at the University of Chicago from September 2006 to August 2007.[7] She has held fellowships at the Max Planck Institute for the Study of Societies and the Woodrow Wilson International Center for Scholars, focusing on innovation equity.[8]

Shobita Parthasarathy@shobitap· Jan 13

While the US undoes its legacy, ANSES and other agencies around the world are making risk assessment better serve society: incorporating public/stakeholder knowledge and rethinking cost-benefit analysis to center social needs. This passage of the statement gives a sense of ANSES' novel approach:

ANSES demonstrates how such expertise can operate in practice. As France’s national agency for health and safety in food, environment, and work environment, it assesses risks ranging from pesticide exposure and air pollution to food contaminants and occupational hazards. Acting independently, it informs policymaking, anticipates emerging threats, and ensures that assessments are transparent and interdisciplinary.

ANSES is a world leader in integrating socioeconomic and social science perspectives into risk assessment, while maintaining close engagement with stakeholders, civil society, and institutions. It has advanced systemic approaches such as One Health, which recognizes interdependence between human, animal, plant and environmental health, and the exposome, which accounts for the lifetime history of exposures that affect health.

All these initiatives make scientific expertise socially robust: they encompass not only the technical dimensions of risk but also their determinants, their implications for society, and the governance challenges they raise. As such, they offer resources to identify and address the deeper reasons for mistrust in scientific expertise.

Shobita Parthasarathy@shobitap· Jan 2

Love it.