I run student and resident research projects as a trauma surgeon. For years the emails trickled in. This year they do not. Students arrive with a question and a draft and ask to be put on a paper. Large language models made literature searches, coding, chart review, and a first manuscript cheaper. They noticed.
On August 20, Will Flanary, better known as Dr. Glaucomflecken, titled a podcast episode “Medical Student Research Should Be Banned.” He said he did research in medical school to check a box. He pointed to residency applicants matching into orthopedics and neurosurgery with more than a hundred research experiences. Four days later, Vinay Prasad wrote that he shared the sentiment and rejected a legal ban. Most student research, he argued, is slop. The failure is committees that still count it. The episode is being forwarded among attendings. The proposed remedy is backward.
This site has already described the incentive. Vijay Rajput argued last year that students are trading bedside time for publications, and that they are not to blame: They are responding to what residency programs measure. Claudia Rodriguez wrote this month that the same box-checking now starts before medical school, with gap years stacked for research hours. Kathleen Muldoon has made the related point about AI: When students outsource the task, they are optimizing for the system we built. I agree with all three on the diagnosis. I do not agree that the next move is to close the lab.
William Stanley Jevons described the relevant mechanism in 1865. In The Coal Question he wrote that it is “wholly a confusion of ideas to suppose that the economical use of fuel is equivalent to a diminished consumption. The very contrary is the truth.” When coal became cheaper to use, more coal was used, not less. The same pattern appears when a research task becomes cheaper. More students attempt it. Jevons also noted the labor analog: Machinery that saves labor can, after a delay, widen demand for the more efficient work.
Real physician voices, twice a week
Free, and one click to unsubscribe.
A different nineteenth-century response is the one now being recommended, in comic form, for student research. Between 1811 and 1816, English textile workers later called Luddites broke machines in Nottinghamshire, Yorkshire, and Lancashire. In Nottinghamshire the targets were wide knitting frames used to produce cheap fabric and to cut wages. In Yorkshire they were shearing frames. In Lancashire they were power looms. The name came from a probably mythical “Ned Ludd.” Parliament made frame-breaking a capital crime in 1812. Trials at York in 1813 ended in hangings and transportation. An estimated 1,000 stocking frames were destroyed. The machines remained.
The Luddites were not uniformly opposed to tools. They opposed machines used to bypass apprenticeship, lower piece rates, and flood the market with inferior work. That distinction is easy to lose. The popular reading is that they smashed productivity itself. Physicians who answer a rise in student emails by closing the lab are closer to that popular reading than to the narrower grievance. The grievance that would map is unsupervised volume: papers the student cannot explain, authorship without contribution, and template analyses of public datasets. Restricting all student research does not address that. It may shift the extra demand toward paper mills and unreviewed drafts.
A response that fits both Jevons and the actual Luddite complaint is narrower. I take the students. They have to pull the data, write the methods, and defend a single paper they could explain to a program director. Authorship follows that work. The models that lowered the cost of a draft do not lower the cost of judgment. That is the remaining bottleneck, and it is a teaching task.
The 2026 residency Match numbers that prompted Glaucomflecken’s episode are real. They describe an incentive to list items, not a reason to stop teaching methods. Prasad is right that a legal ban is impractical and that box-checking should not be rewarded. He is wrong if the implication is that the students who just became able to try should be turned away. Cheaper drafts could mean more students who can be taught, if the attending treats the request as an apprenticeship and not as a bid for a line on an application.
Joshua P. Parreco is a surgeon.
