Special Opinion Article: A Modest Proposal
To Catalyse A New Golden Age of Science By Preventing the Postdocs of America from Being a Burden on Parents, Institutions, & GDP, and Making Them Beneficial to the Semiconductor Industry
by Dr. Jonathan Swift, Ph.D., M.B.A.
‘Science - A New Golden Age’ has diagnosed the problem. Here, we offer a solution.
Read the original Swift essay here.
It is a melancholy object to those who walk through the corridors of our great research universities, or visit the basements of our nation’s teaching hospitals, when they see the hallways and shared offices crowded with postdoctoral fellows of both sexes, followed by their fourth unfunded grant applications, all dressed in conference T-shirts they acquired during their graduate training, importuning every passing professor for a letter of recommendation. These researchers, instead of being able to work for their honest scientific livelihood, are forced to employ their time applying for positions that do not exist, resubmitting proposals that will not be funded, and updating LinkedIn profiles that will not be viewed, while their student loans accrue interest with the reliability of a well-controlled experiment. I myself have often reflected, while stepping over a postdoc sleeping in a departmental common room, that something ought to be done.
I think it is agreed by all parties—university administrators, taxpayers, and the many distinguished commentators who have written eloquently on this subject—that the PhD training system is, in the words of one influential 2014 PNAS article, in a state of “perpetual disequilibrium.” We have been warned that we are producing far more scientists than the system can absorb. We have been told that the biomedical workforce is “unsustainable” and “broken.” Four separate blue-ribbon panels have described our young scientists as “a generation at risk.” The Boston Globe has characterized the postdoctoral position as a “trap.” Nature has published no fewer than seventeen editorials describing the system as “in crisis.” And now the White House Office of Science and Technology Policy, in a report of some hundred pages , informs us that the academy is "a long and difficult road, leading to few stable positions." Surely—surely—if the situation is as dire as these eminent authorities suggest, we ought to act with corresponding urgency. And yet our response has been to convene additional panels, produce additional reports, and use phrases like “a generation at risk” while doing nothing whatsoever about the generation in question.
I shall now therefore humbly propose my own thoughts, which I hope will not be liable to the least objection. I have been assured by a very knowing acquaintance in the semiconductor industry—a fellow who has twice been featured on the cover of Wired magazine and whose stock options have appreciated more reliably than any graduate student’s career prospects—that a young, healthy scientist, well trained in computational methods, is at the age of thirty-five an ideal candidate for conversion into an organic neural processing unit. Recent advances in brain-computer interface technology, combined with the desperate shortage of advanced semiconductors that has brought our automotive and consumer electronics industries to their knees, present an unprecedented opportunity for what I shall term “workforce-to-wafer integration.” The biological neural networks of a postdoctoral fellow, having been optimized over eight to twelve years of intensive training, represent precisely the kind of specialized computational architecture that NVIDIA and its competitors have struggled to replicate in silicon. I am told by my acquaintance that Jensen Huang himself has expressed interest, though naturally through intermediaries.
I do therefore humbly offer it to public consideration that of the approximately sixty thousand postdoctoral researchers currently employed in American institutions, twenty thousand may be reserved for eventual faculty positions—a figure that represents, I am aware, a triumph of hope over arithmetic, but one must leave room for optimism. This leaves forty thousand surplus scientists per year, whose neural architectures could be harvested and converted into approximately 1.2 billion processing units of organic computation, assuming three hundred units per researcher based on current interface electrode density. At market rates of roughly forty-seven dollars per unit—a figure derived from recent NVIDIA earnings calls and adjusted for biological variability—the annual value to the semiconductor industry would exceed fifty-six billion dollars. This represents more than the current NIH budget and would, I am confident, please shareholders considerably more than any number of published papers on fruit fly genetics. The researchers themselves, having been converted to processing units, would at last achieve the productivity metrics that eluded them in life.
The advantages of this proposal are obvious and many. First, it would instantly address the “misaligned” workforce pipeline that so concerns the Executive Branch. If we truly believe—as the authors of this report apparently believe—that we are training too many of our current scientists for non-existent academic positions, then we have a moral obligation to do something with the surplus. One cannot in good conscience describe a situation as a “crisis” and then propose “fast-track grants” or “pre-competitive consortia.” These are not crisis responses. Converting excess researchers to semiconductor components—that is a crisis response. Second, the proposal would relieve universities of the administrative burden of managing postdoctoral benefits, visa applications, and the increasingly elaborate mental health services required to maintain researcher morale. Dead scientists require no counseling, no parking permits, and no awkward conversations about when, exactly, they plan to move on.
Third, the program would provide closure for families who have long wondered when their children’s training would end. A mother who has spent fifteen years answering “she’s still in school” at dinner parties would at last have a definitive response. Fourth, it would eliminate the awkward Thanksgiving conversations during which young scientists must explain to uncles in the insurance industry why their seventeen years of higher education have not yet resulted in a salary exceeding that of a municipal bus driver. Fifth, the proposal would boost NVIDIA’s stock price substantially, thereby enriching the pension funds and 401(k) accounts of ordinary Americans. I calculate that the average schoolteacher’s retirement account would appreciate by approximately $4,700, an outcome that surely outweighs any number of postdoctoral research papers that nobody outside the field will read. Sixth, it would attract considerable interest from Mr. Sam Altman and the artificial intelligence community, who have long sought to combine human cognition with silicon infrastructure and would, I suspect, pay handsomely for early access. Mr. Altman has previously proposed that universal basic income might address technological displacement; how much more elegant to skip the middleman entirely.
Seventh, and most importantly, the program would demonstrate that American universities can produce something of tangible commercial value. For too long, critics have complained that our research institutions prioritize abstract knowledge over practical application, that our scientists pursue questions of no interest to anyone but themselves, that our federal investment in basic research yields insufficient economic returns. The conversion of researchers to semiconductors would answer these critics definitively. We would be taking the most expensively educated people in human history—individuals who represent roughly three-quarters of a million dollars each in taxpayer-funded training—and finally, at long last, extracting value from them.
I anticipate that some persons of a squeamish disposition may object to this proposal on the grounds that it involves, in some technical sense, the termination of human life. To these critics, I would pose a simple question: Is the current system not already accomplishing the same end, merely at a more leisurely pace? We have been told, by the authors of this report, that young scientists are “trapped” in “holding patterns,” their dreams “deferred,” their potential “wasted,” their careers “derailed.” If we are already “losing” this generation—if their careers are already being “terminated” by funding cuts and policy chaos—then my proposal merely accelerates the process while capturing the residual value. The current system imposes years of psychological torment before delivering its verdict. Conversion to neural processing units would be instantaneous and, I am told, painless. Which is the crueler fate?
Therefore let no man talk to me of other expedients: of actually funding the National Institutes of Health at levels commensurate with the growth of biomedical knowledge and the cost of conducting research; of restoring the success rate for first-time investigators to the thirty percent it enjoyed in 1998 rather than the eighteen percent it has collapsed to today; of establishing meaningful career paths for researchers who do not wish to run their own laboratories but who possess invaluable expertise in experimental techniques; of providing young scientists with the stability and independence they require to pursue genuinely innovative research rather than incremental studies designed primarily to minimize the risk of grant rejection; of treating the humans who conduct scientific research as though they possessed value independent of their capacity to generate publications and overhead; or of acknowledging, in any meaningful institutional way, that we are systematically driving our most talented young minds out of research entirely and into the welcoming arms of McKinsey, Goldman Sachs, and the biotechnology industries of competitor nations who have proven notably more willing to invest in their scientific workforce. Let no man speak to me of these solutions until he can explain why two decades of unanimous expert consensus, as well as the present hundred page report, have produced no meaningful policy change whatsoever. I have grown weary, after many years, of offering such vain, idle, visionary thoughts.
I profess, in the sincerity of my heart, that I have not the least personal interest in endeavoring to promote this necessary work, having no other motive than the public good of my country, by advancing our semiconductor industry, providing resolution for young scientists, relieving our universities of unsustainable burdens, and giving some appreciation to the stockholders. I myself hold no patents in brain-computer interface technology beyond a small provisional application, recently filed, covering certain methods of neural extraction that seemed promising during my consulting work for a firm I am not at liberty to name. My own postdoctoral researchers—of whom I employ seven, funded through three R01 grants that I have held continuously since 1987—would naturally be exempted from any such program, as their labor remains essential to my ongoing investigations into topics that I remain confident will prove significant. The youngest is forty-three years of age, and I am told she has nearly completed the data collection for her first independent publication; I have encouraged her to be patient, as I once was, and to trust in the system that has served me so well. I do not think it unreasonable that she should wait her turn.
Dr. Jonathan Swift, Ph.D., M.B.A. is the Endowed Chair of Workforce Optimization at Stanford University and author of the book “Grit and Grants: How I Built My Laboratory Through Hard Work and 100% Merit (Harvard University Press, $34.95). He receives consulting fees from NVIDIA totaling an amount he describes as "immaterial to someone of my stature" and has submitted a public comment in support of ‘Gold Standard Science.’
This is glorious