Introduction
Dr. Scott Atlas, a prominent figure affiliated with the Hoover Institution on War, Revolution, and Peace at Stanford University, has stirred considerable controversy with his recent suggestion to abolish the National Institutes of Health (NIH). This perspective raises important questions about the role of public funding in scientific research, a topic that invites a deeper examination.
Part the First: Beware of Resting on the Shoulders of Atlas. Dr. Atlas is a radiologist who has found a home in the Hoover Institution, yet one can question the genuine commitment of this institution to the concepts of revolution and peace as they resonate within popular understanding. His recent advocacy for dismantling the NIH has been engagingly critiqued by Dr. David Gorski at SBM in the article “Abolish the NIH”? Dr. Scott Atlas gives the antiscience game away.
It is essential to acknowledge that while NIH is indeed a sizable organization that could benefit from reforms, it has been pivotal in the evolution of nearly every available drug and medical device, supported by the NIH, the National Science Foundation, and analogous organizations globally. Despite this, Dr. Atlas argues that the NIH, burdened as it is by a “$48 billion government monopoly” should be replaced by market solutions. He cites economist Milton Friedman’s 1980 assertion that the NIH should be abolished and points to Friedman’s position that the National Science Foundation should also cease to exist, stating, “Nothing” should replace it.
In 1980, economist Milton Friedman said the National Institutes of Health should be abolished. Friedman said the same about another government research agency, the National Science Foundation. And when he was asked what the NSF should be replaced with, he replied: “Nothing.”
He further argues in a Washington Post article:
Abolishing the NIH would not create a funding vacuum. The private sector funds 78 percent of U.S. biomedical research and development. Venture capital has exploded. Investment focused on artificial intelligence in drug discovery alone surged nearly 36-fold between 2010 and 2024, funding the overwhelming majority of innovative medical devices and driving the AI revolution in biomedicine. The Howard Hughes Medical Institute, the Wellcome Trust, the Gates Foundation, and the Chan Zuckerberg Initiative collectively represent over $150 billion in endowment capital. Nokia Bell Labs, financed initially by AT&T and later Nokia Revenue, garnered 10 Nobel Prizes.
As Dr. Gorski insightfully points out, any fifth grader could dissect this argument. If the NIH is truly a monopoly, how could the private sector account for 78% of U.S. biomedical R&D? Additionally, Dr. Atlas confuses the immense endowments of institutions like the Howard Hughes Medical Institute with their actual funding capabilities, which are but small fractions of those endowments. While it’s correct that Bell Labs was initially supported by AT&T, it’s critical to recognize that AT&T, once a regulated monopoly, invested only a nominal amount in Bell Labs in the broader financial context. After AT&T’s breakup, research initiated under Bell Labs continued to earn Nobel Prizes, but it’s worth noting that the most infamous output from later Bell Labs, Jan Hendrik Schön, was embroiled in scientific misconduct.
The expectation that Big Pharma will independently drive foundational biomedical science through “blue sky” research is misguided. Without a robust basis in fundamental research, advancements cannot materialize. Consider two notable examples: The drug imatinib, heralded as a groundbreaking chemotherapeutic agent developed by Nicholas Lydon at Ciba-Geigy (now Novartis), has its conceptual roots tracing back to 1950s work funded by the NIH from pioneers like Edwin G. Krebs and Edmond H. Fischer, who researched reversible protein phosphorylation, which underpins cellular processes.
A more unlikely yet equally significant case involves the Gila monster, whose venom has been foundational for developing GLP-1 antagonists for obesity treatment, with original research by Raufman, Singh, and Eng occurring far removed from industry interests.
The conviction that essential research can be overlooked while still yielding solutions is unrealistic. Numerous experts at Stanford could elucidate the fundamental nature of this issue to Dr. Atlas in simple terms. He ought not to be taken seriously, particularly given his scant engagement with NIH structures—his only NIH grant dates back over 25 years, suggesting insufficient relevant experience to support his views.
Part the Second: The Science of Natural History as It Should Be Done. Growing up, I collected insects, delighted in learning their scientific classifications, and sifted through fossil sharks’ teeth. This ignited a belief that naming new species was the pinnacle of biological research. However, as highlighted in an article about The Wilderness Project, discovering new species extends beyond mere nomenclature.
A scientific expedition to Angola’s remote Lisima Plateau has unveiled numerous species previously unknown, including eight unrecorded dragonflies, three new grasshoppers, and about 60 newly identified moths and butterflies. Findings from the Cassai Life Atlas—a biodiversity survey conducted by The Wilderness Project in February 2026—provide an unprecedented view of a landscape that nourishes the sources of Africa’s four significant river systems: the Congo, the Okavango, the Zambezi, and the Cuanza.
The stunning images of these new species are captivating, and the accompanying video is extraordinary. If such research can take place in Angola, the future of natural history appears promising. As one scientist expresses, “We are born naturalists.” Indeed, it’s crucial to remember that “protecting our biodiversity is tantamount to protecting ourselves,” a sobering reminder as we often view the world as a detached recipient of our mistakes.
Part the Third: There Goes Another One. Since January of last year, opportunities to pursue scientific research abroad have become increasingly appealing, as illustrated by Omar Yaghi‘s recent move from UC-Berkeley to Tsinghua University in Beijing. This shift means that the Nobel Prize winners at Berkeley will have less competition for coveted parking spaces outside LeConte Hall, reserved for laureates.
The move, first reported by the South China Morning Post, coincides with the Trump administration’s efforts to reduce U.S. science funding and limit international research collaborations. Countries like China are actively enticing American researchers with financial support and resources. Earlier in the year, France announced funding for numerous U.S. scientists relocating there, while China has introduced talent recruitment programs, offering sums and monthly stipends to researchers who settle in their regions.
Yaghi already held a connection with Tsinghua University, having earned an honorary professorship there in 2022, and he was formally welcomed as a full-time faculty member on July 3. He did not comment for this piece.
However, in a recent Scientific American interview, Yaghi expressed concerns about the current state of U.S. science, citing reductions in grants and diminished backing from science agencies critical to academic research. He urges U.S. researchers to engage with the emerging “AI revolution,” asserting that involvement with AI models is essential for the survival of advanced research in the United States.
If AI is to serve a constructive purpose, it will likely stem from the foundational research exemplified by Yaghi’s work. Ironically, it could be that the current administration’s response is dismissive. Moreover, it’s notable that Yaghi, a Palestinian born in Amman, achieved the American dream by arriving in the U.S. at 15. In science, merit prevails over background, and his contributions will indeed be missed.
Part the Fourth: Tell Me Once Again Why the People Are Losing Respect for Science and Scientists. What are the public to think when faced with headlines like this? Paper mill studies receive twice the citations as authentic research. Having engaged with biomedical literature for over half a century, this revelation is unsurprising. The issue lies in an alarming trend among many in my age group who are publishing in questionable, predatory journals, where quantity trumps quality.
Studies in cancer research displaying characteristics of paper mill production garner double the citations compared to legitimate papers, according to an analysis of thousands of publications.
In a study posted on the bioRxiv preprint server, the authors reveal that manuscripts likely generated by paper mills often cite or are cited by other suspect articles. Paper mills manufacture and sell low-quality manuscripts—often replete with falsified data—designed to mimic genuine research.
Adrian Barnett, a statistician from Queensland University of Technology in Brisbane, echoes the concern that citation manipulation is inflating impact metrics of journals within molecular oncology, often considered for academic positions and funding.
The fallout has infiltrated esteemed scientific journals from major legacy publishers, leading to an ongoing conundrum: how do we educate medical students to critically evaluate the literature in such an environment? It seems increasingly challenging, especially as AI tools generate superficially appealing but flawed outputs in response to prompts about new developments in cancer treatment. This situation portends a troubling future.
Part the Fifth: AI Detectors, Do They Work? The answer appears to be “probably not.” It has become evident that talented writers, particularly in scientific fields who don’t utilize AI, may still be misidentified as having submitted AI-generated work. As reported in a Nature article:
In November, Lauren Jager, an undergraduate chemistry student at Idaho State University, discovered that her application for PhD programs faced scrutiny. Some portals cautioned against using AI tools for personal statements and stated they would employ detectors to filter out applications suspected of containing AI-generated content. Although Jager hadn’t utilized AI, she nonetheless opted to ensure her application was safe, testing several online AI detectors.
“They all came back at almost 100% AI,” she reported, leading her to panic.
And rightly so. The pertinent question is whether these tools actually function effectively and whether they should be deployed if they risk unjustly branding innocent students like Jager as cheaters.
Various academics have examined this issue. A 2025 research paper assessed GPTZero, described as a widely-used AI detector. It found that while most fully AI-generated essays were detected with high confidence, GPTZero had a false-positive rate arising from misidentifying human-written texts as AI-generated, resting at about 16%. Thus, its “reliability in distinguishing human-authored texts is limited.”
A 2023 evaluation of multiple AI detection tools—OpenAI, Writer, Copyleaks, GPTZero, and CrossPlag—showed that the tools performed better in identifying texts generated by the LLM GPT-3.5 than by the more sophisticated GPT-4 model. When applied to human-authored submissions, the detectors yielded erratic results, including false positives and vague classifications.
User experiences on Reddit illustrated the challenges AI detectors face in categorizing human-written content. In several instances, the US Declaration of Independence was flagged as AI-generated. Nature tested sections of this 1776 document through ZeroGPT multiple times, with results indicating that it was 95% to 100% likely to be AI-generated.
Thomas Jefferson, an imperfection like the rest of us, nonetheless possessed an intelligence that was far from artificial. In a fair world, this outcome should effectively challenge the legitimacy of AI detectors. Indeed, while AI is likely here to stay, the implications of its integration into research remain uncertain and may prove problematic.
Conclusion
As we navigate these complex issues, from the debates surrounding public funding in biomedical research to the challenges presented by AI in academia, it’s critical to maintain an emphasis on sound scientific principles. The integrity of research and education relies on us recognizing the value of fundamental inquiry and remaining vigilant about the quality of our outputs. Thank you for reading! Until next week, let’s keep in mind that the impact of heat is palpable and its consequences are insidious.
