Coffee Break: Counterfeit Scientific Papers, Deep Fakes, CDC on the Ropes, MAHA, and Hope from the Middle of the Country

Part the First: Paper Mills and the Corruption of Research.  No not Hammermill.  I don’t think I have actually known of someone buying a “scholarly” paper for publication, and I remember reading (a few paragraphs) only a few that seemed to be purpose built.  But following up on The Credibility Crisis in Science from earlier this week, when it comes to reading something that looks good, the warning remains “Let the buyer beware!” as shown Nature article How much for a fake authorship? Ad database reveals secrets of scientific fraud:

Researchers have amassed a data set of thousands of advertisements selling research-paper authorships online, shedding light on the global marketplace for academic fraud.

The collection — the largest of its kind — contains more than 18,700 adverts that were posted between March 2020 and early April 2026 by seven paper mills — businesses that produce fake or low-quality research and sell authorships. Together, the companies cater to academics in the Middle East, Central Asia, Eastern Europe and India.

An analysis of the advert data found that a first-author slot on an article sold by a paper mill costs a median value of nearly US$800, with prices ranging from $57 to more than $5,600. The work is described in a preprint submitted to arXiv this week.

Researchers, publishers and indexing services could use the list of adverts to screen their publications and audit which journals and research topics are most likely to be targeted, says study co-author Reese Richardson, a metascientist at Northwestern University in Evanston, Illinois.

“The preprint paints a valuable picture of the significant financial scale of these operations, underscoring the pressure put on researchers to publish in order to advance in their careers,” a spokesperson for the New Jersey-based publisher Wiley told Nature.

One wonders how many of these papers were written by an AI app.  Actually, there is nothing to wonder – most of them.  You might expect that these papers are all published in fake journals that are little more than a website with a link where the authors submitters pay the article processing fee of $1200 to $8000.  Not so:

Out of the papers that have not been retracted, four were published in Springer Nature journals and five in Wiley journals. Twenty-three papers appeared in conference proceedings published by the Institute of Electrical and Electronics Engineers (IEEE). Nature’s analysis also identified a further three papers, one each in journals published by Elsevier, Frontiers and Taylor & Francis. Some of the papers appeared in different publishers to those named in the advert.

Anyway, the Association of American Medical Colleges and its ~150 medical schools are seriously considering how AI fits into medical education and medical practice.  The short answer is that it doesn’t fit in either place, except perhaps for high-resolution image analysis that still must be confirmed by a real live person.  But that has been the case for a long time.  Taken together with the Conor’s link on cheating using AI in colleges in universities (I wonder what a current student would do in the classes I took on Early Modern Europe and the Renaissance, in which we read 8-10 books in a 10-week quarter, and then wrote in-class essays in response to open-ended questions). Dr. Chat is coming your way.  So, beware.  As for me, it is good to be old and to hope that my two grandchildren will one day appreciate the actual books their grandfather will leave to them in hopes they will read them.

One last thing.  The paper discussed in the Nature article is a preprint.  Caveat emptor.  In any case, my prose can be clunky at times, but I have never asked ChatGPT a question and I never will.  AI is not a calculator for words, and I cannot afford to lose any more connections in my noggin at this age.

Part the Second: Deep Fakes Will Take Over the World, if We Let Them.  But Henry Farid keeps fighting the good fight, as told in this article from Science last week, Reality Check.  And the Tomahawk missile that killed more than 150 people at the school in Minab at the beginning of our Ramadan War in West Asia?  Was it real?  Yes:

The clip showed a downward streak of black through a clear blue sky, the silhouette of a U.S. Tomahawk missile like a metallic bird of prey diving for the kill. Then the impact, a plume of black smoke rising over buildings, palm trees, and electrical wires. By the time the video arrived in Hany Farid’s inbox on a Sunday morning in March, experts had already confirmed the scene showed Minab, the city in southern Iran where a missile strike had killed more than 150 people at a girls’ elementary school a week earlier. The U.S. government had denied responsibility, claiming a rogue Iranian missile was to blame. But the video, released overnight by an Iranian news agency, told a different story. Journalists had emailed Farid’s company, GetReal Security, asking him to verify the footage.

…

That Sunday morning, settling in front of a computer with his wife, Emily Cooper, in their home in the hills over Berkeley, Farid went to work. His first impulse was to be suspicious. The war in Iran had already produced a firehose of AI-generated images. Why had it taken a week for this video to become public? The low resolution of the footage did not help his confidence either.

The video was real, and the article goes into detail about how to spot a fake.  But the question is, should we have to do this, for everything?  No, but we do, one way or another.  When did this begin?  More on Henry Farid that goes back to a beginning that many of us lab rats remember:

In the late 1990s, as a postdoc at the Massachusetts Institute of Technology working on image analysis and computer vision, he found himself waiting in line at the library and started browsing a book from the return cart called Federal Rules of Evidence. He landed on a passage about introducing photographs in a court of law, which said that both a 35-millimeter negative and a print from that negative are considered “original.” “And then it said, there’s this new digital format, and we are going to treat it the same as a 35-millimeter negative,” Farid recalls. “I remember thinking: Wow, that seems like a bad idea.”

The history of manipulating photos is as old as photography itself. A famous portrait of Abraham Lincoln actually shows his head superimposed on the body of the politician John Calhoun, Stalin airbrushed enemies out of his photos, and two girls from Yorkshire in England, convinced legions of people—including Sherlock Holmes creator Arthur Conan Doyle—that their “photos” of fairies were real. But these fakes were rare, Farid says. It was clear to him that this was about to change. “I just started thinking about what happens when everything becomes digital and malleable?”

He was not the only one.  Back when one had to master a 35mm SLR camera attached to the microscope or on the stand over your blot that showed your DNA, RNA, or protein to be the right size at an expected density, great skill was required to produce a publishable image on the front end and the back end of an experiment.  The tools required to manipulate such an image existed only in the bowels of the NKVD, the CIA, and similar agencies the world over.  Back in those dark ages, William Summerlin had to use a black felt-tip pen to fake his mouse skin transplants in the absence of immunosuppression.  As you might expect, that didn’t get him very far.  I will never forget the first time I saw the image analysis tools integrated into the bright, shiny, and expensive phosphorimager.  The sales geek was nearly breathless while showing us how an image could be manipulated by flipping and inverting.  The nonlinear brightness/contrast gamma function was the most “useful.”  A few of us in the room asked no one in particular and everyone, “What happens when everything becomes digital and malleable?  Now we know the answer.  And ChatGPT has come along just in time to explain it all to us.  Oh, joy!

Part the Third: CDC Still on the Ropes.  The Morbidity and Mortality Weekly Report (MMWR) has been the most important public communication portal of CDC for 65 years.  MMWR is where we first learned of what became known as AIDS, and this knowledge was transmitted without delay because the scientists at CDC knew what they were doing.  Despite a few high-level bureaucratic foul-ups and technical goofs during the early days of COVID-19, they still do.  Or did until last year:

When Jay Bhattacharya took on a second job temporarily heading the U.S. Centers for Disease Control and Prevention in February, many CDC staff and public health leaders felt reassured by his avowed support for the battered agency. But last month, a cascade of events fed worries that Bhattacharya, who also directs the National Institutes of Health (NIH), is compromising CDC’s science. First, he pulled a routine study on COVID-19 vaccine effectiveness in press at CDC’s Morbidity and Mortality Weekly Report (MMWR). Then he questioned the peer-review process at MMWR—for 65 years a mainstay for CDC to convey urgent public health data. After saying MMWR needed revamping, Bhattacharya pivoted and called for a new, externally reviewed CDC journal.

The CDC-led study that touched off the uproar assessed how well COVID-19 vaccines work by confirming infections in people visiting clinics for respiratory illness. The Washington Post revealed in April that concerns about this methodology, known as test-negative, led Bhattacharya to delay publication of the study and later reject it. Although the CDC leader has the authority to decide what appears in MMWR, the move drew an outcry on social media.

…

“MMWR is not currently a peer-reviewed journal,” Bhattacharya replied. “But we are working on changing that.” (Bhattacharya, a fierce critic of CDC’s policies during the COVID-19 pandemic, is running the agency pending Senate confirmation of Erica Schwartz, former deputy surgeon general and the White House’s nominee to be director.)

Several former CDC officials who oversaw MMWR promptly disputed Bhattacharya’s claim. They argued the publication receives internal reviews more rigorous than ones done at most scientific journals. “Our scientists tend to be very enthusiastic and critique very thoroughly,” said Debra Houry, prior chief medical officer at CDC. Houry quit in August 2025 because of concerns about political interference in vaccine policy by Robert F. Kennedy Jr., secretary of the Department of Health and Human Services (HHS), which oversees CDC and NIH.

Actually, MMWR is a rigorously peer-reviewed journal.  It’s just that because the purpose of MMWR is to be the first warning of new and/or epidemic disease, the reviews are, or were, done by the committed and well qualified scientists at CDC.  Imagine 45 years ago, with the first reports of AIDS from New York and San Francisco being peer reviewed in the traditional way, including a 3- to 9-month delay between the writing of the report and publishing in a journal.  Ridiculous.

Anyway, Jay Bhattacharya, MD-PhD, has never been a physician or scientist, and it shows more and more every day.  He also apparently doesn’t know much about CDC, which already publishes the traditional peer-reviewed journals Emerging Infectious Diseases and Preventing Chronic Disease.  Hold on to your seats, the ride gets bumpier and bumpier. 

Related, for the latest on measles, this article from Science-Based Medicine covers the subject like a blanket.  The main thing to remember about measles, which spreads very efficiently, is that herd immunity requires a vaccination rate of 90% or better.  As the catastrophe that is Andrew Wakefield and his spurious connection between the MMR vaccine and autism continues to play itself out, vaccination rates get lower and lower.  And a very dangerous disease inexorably makes a comeback.  This animation of herd immunity shows how it works.  Measles starts at 1:00.

Part the Fourth: The MAHA Movement, What Does It Really Want?  The Kaiser Family Foundation has published a poll that asked that very question.  Turns out they want pretty much what everyone else wants: To not have to choose between health and bankruptcy:

When asked to identify their most important health priority for government to address, far more MAHA-supporting voters identify lowering the cost of health care (42%) than other issues more closely associated with the movement, such as restricting the use of chemical additives in the food supply (21%), reevaluating the safety of vaccines (10%), limiting corporate influence on food policy (8%), or restricting the use of pesticides in agriculture (8%).

The major difference between Republican MAHA voters and Democratic MAHA voters is vaccine policy, naturally.  But all voters put healthcare costs at the top of their list.  I was reminded of this earlier today when I noticed a coworker limping back to her office after a reception for our graduating medical students.  We are work friends so I asked her what was up.  I knew that she had had a hip replacement last year and that the outcome was better than she could ever have imagined.  She told me that it wasn’t so bad, but she was still paying off her part of the previous hip replacement and would have to wait until that $4,000 debt is retired before getting the second operation.  I was immediately reminded of the Aneurin (Nye) Bevan (1897-1960) quote that I make sure all of my medical students hear in our tutorials:

The field in which the claims of individual commercialism come into most immediate conflict with reputable notions of social values is that of health. From: In Place of Fear, 1952, back when Labour was useful

No civilized society makes someone wait for a hip replacement until the first one is paid off.  But this doesn’t really hit home until it, well, hits home.

Part the Fifth: Speaking of Civilized Society.  We have discussed Wendell Berry before, most recently when the subject was the book Feed the People: Why Industrial Food Is Good and How to Make It Even Better.  David Demeree has an article in Front Porch Republic on Wendell Berry’s return home to Kentucky from a conventional literary career in New York City.

But Demeree goes further and writes about the late Gene Logsdon, who did the same, from Philadelphia, and became a friend of Wendell Berry.  Logsdon’s books are also well worth the read.  I met him at the Kentucky Book Fair in Frankfort about twenty years ago and we spent a delightful hour talking about skills that I observed as a child but have largely been lost.  These are skills that our grandchildren will need to live a human life during the rest of the twenty-first century.  Gene Logsdon was The Contrary Farmer who also spoke to the people.  And if you find a copy of Good Spirits: A New Look at Ol’ Demon Alcohol, that is a very pleasant place to begin with Ol’ Gene – good history and much hilarity all in one place.

Thank you for reading!  See you next week from Scotland in the Kingdom of Fife.