Engineering Chair: Most People Should Not Go to College — A Companion Guide

Source: Engineering Chair: Most People Should Not Go to College by Dr. Brian Keating (Into the Impossible) with Darren Lipomi · used as the primary text for this companion

Dr. Brian Keating's conversation with Darren J. Lipomi, chair of chemical and sustainability engineering at the University of Rochester, opens with a provocation. Lipomi argues that only a minority of young people should attend a four-year institution, largely because of debt and the long odds of the academic research track. Keating presses him to defend the claim and to say what should replace the current model. The rest of the conversation walks through the system that produces major discoveries anyway: the H-index, the PhD apprenticeship, nanotechnology that already arrived, mentorship written in blue ink, and the question of whether an engineer should read Plato.

What this conversation is about

The title is clickbait with a point. Lipomi is not anti-education; he is the chair of a department that is actively recruiting undergraduates. His argument is that the four-year degree has become a default path for too many people, especially when the debt is non-dischargeable, the academic job market is saturated, and high-paying hands-on trades and technician roles are undervalued. Keating pushes back from the perspective of someone who has spent his life inside universities and wants to know whether the system can be reformed or replaced.

The discussion moves quickly from metrics to materials. Lipomi explains the H-index and citations as the imperfect currency of promotion; he describes the PhD as an apprenticeship for people who cannot imagine doing anything else; he argues that the nanotechnology revolution happened in catalysis, drug delivery, and OLEDs but was not called nanotechnology; and he talks about what it was like to be mentored by George Whitesides, one of the most cited living chemists. The conversation is a map of how modern research actually works, with all of its incentives and absurdities.

Chapter map

These are the YouTube chapter markers from the video, with direct links to each starting point.

1. 00:00He says don't go to college. He runs the department.
2. 01:10The H-index: imperfect, irreplaceable
3. 02:041 in 20 PhDs gets the job
4. 05:42The meteorite and the origin of life
5. 08:34Did life's asymmetry come from space?
6. 12:00Who is an organic materials chemist at 3am?
7. 13:52The collaboration that started at a coffee shop
8. 15:22The nanobot revolution already happened
9. 18:24Nanotechnology inside your body right now
10. 19:08The $70 billion invention Kodak gave away
11. 22:12Kodak, Xerox, Bausch + Lomb: what Rochester built
12. 26:06Working under Whitesides at Harvard
13. 28:54What he wrote on the outline in blue ink
14. 30:38Who you work with matters more than what you work on
15. 31:56Department chair: hostage negotiator, not boss
16. 34:36Why the department changed its name after 110 years
17. 36:00The headwinds have never been stronger
18. 38:44Should engineers read Plato?
19. 39:38The debt you can't discharge in bankruptcy
20. 40:06Do you buy Baumol's cost disease?
21. 43:14Why he had to write the book
22. 48:16The Fisher-Price camcorder experiment
23. 51:06Teaching in one sentence. Research in one.
24. 53:42Losing Darren to Rochester: the Fernando Tatis analogy

Metrics and the H-index

Lipomi starts with the problem of evaluating faculty for promotion and tenure. You need a metric that is reliable and at least somewhat standardized within a field. The tools are the H-index, citation count, and grant dollars brought in per square foot of lab space. The H-index, introduced by physicist Jorge Hirsch in 2005, is the largest number h such that the researcher has h papers each cited at least h times. It is an author-level metric meant to balance productivity and impact, and it correlates with prizes and fellowships because hiring and funding panels need a number they can compare.

The limitations are just as important. The H-index is field-dependent, database-dependent, and insensitive to a few massively cited outliers. A methodological paper or a review can distort it. A paper with one citation today may have a thousand in ten years. Lipomi's summary is unsentimental: the system is imperfect, but it is the one we have, and it reflects the human competitive instinct that capitalism already exploits. The real question is whether the metric measures a deep, lasting impact, and that is much harder to judge.

The PhD apprenticeship

The video's most quoted number is stark: "Only one out of every 20 PhDs becomes a tenure track professor." Lipomi uses this as shorthand to argue that the PhD is not a continuation of undergraduate school but a research apprenticeship that consumes most of your mental bandwidth for four to seven years. He compares the people who should do it to artists, musicians, and chefs: they go to grad school because they cannot imagine doing anything else.

The published picture is slightly less dire but still sobering. A 2014 systems analysis of the academic job market estimated that, in steady state, about 12.8% of engineering PhD graduates in the United States can attain tenure-track positions, and less than 17% of new science, engineering, and health PhDs find tenure-track positions within three years of graduation. Lipomi's one-in-twenty figure is more dramatic than these estimates, but both show that the academic path is a small funnel. The cost is also real: a typical graduate-student stipend is around $35,000 a year, retirement savings are deferred, and student-loan debt generally cannot be discharged in bankruptcy.

Nanotechnology already happened

One of the strongest technical arguments in the conversation is that nanotechnology is already all around us, just not in the form of Eric Drexler's mechanical assemblers. Lipomi points to chemical catalysis, where nanomaterials form bonds at industrial scale, and to self-propelled micro- and nanomotors that can move in the bloodstream. The Drexler-Smalley debate about "sticky fingers" and "fat fingers" was a real dispute over whether molecular assemblers could physically position atoms with enough precision; Lipomi thinks the argument missed the fact that enzymes and catalysts already do similar work.

He also traces the local history. The organic light-emitting diode (OLED) was invented by Ching W. Tang and Steven Van Slyke at Kodak Research Labs, a few miles from the University of Rochester. Today OLEDs are a ~$70 billion industry, but Kodak captured little of that value. The region still lives in the shadow of Kodak, Xerox, and Bausch & Lomb, though the economy has shifted from imaging to biomedical devices and optics. Richard Feynman's 1959 lecture There's Plenty of Room at the Bottom is the usual origin story for nanotechnology; Lipomi's twist is that the room is already being used, we just stopped calling it by that name.

Whitesides and the blue ink

Lipomi's PhD advisor at Harvard was George M. Whitesides, one of the most cited living chemists and a recipient of the National Medal of Science, the Kyoto Prize, and the Priestley Medal, but not a Nobel laureate. Whitesides was famous enough that getting into his lab was not as hard as getting into Harvard in the first place, in part because he trusted attrition to filter people out.

His mentoring came in the form of written comments on outlines, in blue Pilot rollerball ink. The transcript records phrases like "this is a pig's breakfast" and "this is illiterate." Lipomi describes these notes as harsher-looking than they actually were, and as formative. The lesson he takes away is that who you work with and how they teach you to think can matter more than the exact problem you work on.

Department chair as hostage negotiator

Lipomi describes the department chair as less a boss than a negotiator. A chair has final say on teaching and service assignments, allocates lab space, and is the first responder on promotion and tenure cases, but the faculty are not employees in a normal sense. He quotes a colleague's line: the chair is the person who negotiates with the hostage takers on behalf of the kidnapped.

The context for that role has become harder. Lipomi points to a "demographic cliff" in which fewer high school students are available, immigration policy is less friendly to international students, and tuition keeps rising faster than inflation. At the same time, physical AI and robotics will require materials science and mechanical engineering innovation that cannot be solved from a laptop alone. The universities that survive, he argues, will be the ones that differentiate their product rather than simply credential more students.

Plato, Baumol, and the debt problem

The conversation asks whether engineering students should read Plato in a freshman seminar. Lipomi says yes: the traditional reason for a bachelor's degree is to be a literate, numerate citizen who knows something about civic society, music, art, and literature. Other countries, he notes, treat engineering and medicine as vocational tracks and skip the liberal-arts breadth. The tension is between the ideal of a broad education and the economic pressure to certify a skill quickly.

That pressure connects to Baumol's cost disease, the idea that labor-intensive sectors that cannot automate as fast as manufacturing (education, healthcare, live performance) become more expensive over time. Keating raises it; Lipomi says he is not familiar with the term in that form. The practical consequence is that tuition at private institutions has risen faster than inflation by something like a factor of three, the debt cannot be discharged in bankruptcy, and the consumer is asked to buy a product whose return is uncertain. Lipomi is not calling for the abolition of college; he is asking families to do the same due diligence they would do at an Apple Store before taking on that debt.

The intersection of three interests

Lipomi's career advice, near the end, is to find a skill at the intersection of three or more interests that no one else is working on. The Venn diagram is simple but sharp: being a decent organic chemist might put you at one in ten; adding mechanics makes you one in fifty; adding neuroscience makes you rare enough that you can own a small field. This is the practical recipe behind his own work, which sits at the intersection of organic electronics, mechanics, and human perception.

Mentions beyond the main pipeline

The conversation touches several topics that are not expanded at length but are worth naming. Homochirality and the origin of life come up through a meteorite Keating gives Lipomi; the idea is that polarized radiation in space might create an excess of one molecular mirror image, which could explain why life uses one enantiomer. Lipomi also mentions the Fisher-Price PXL2000 camcorder as his first video and audio production tool, a black-and-white camera that recorded onto audiotape. The department's recent name change to chemical and sustainability engineering reflects its move toward recyclable polymers, e-waste, solar cell manufacturing, and water purification. The closing Fernando Tatis analogy refers to Lipomi leaving UC San Diego for Rochester, with Keating framing it as a fan watching a favorite player move to a new team.

Companion glossary

1. H-index — An author-level citation metric: the largest number h such that the author has h papers each cited at least h times. Proposed by Jorge Hirsch in 2005.
2. Tenure-track professor — A faculty position on the path to permanent employment at a university, typically involving research, teaching, and service.
3. Baumol's cost disease — An economic explanation for why labor-intensive sectors with low productivity growth (such as education and healthcare) tend to rise in price relative to sectors that automate easily.
4. Homochirality — The property of living systems that they use one mirror-image form (enantiomer) of many biological molecules, such as left-handed amino acids. The origin of this asymmetry is an open question.
5. Enantiomeric excess — A measure of how much one enantiomer dominates a sample, often reported as a percentage. Some meteorites show an excess of one form.
6. Drexler-Smalley debate — A 2001-2003 dispute over whether molecular assemblers could position atoms and build nanoscale machines, or whether chemical constraints like "sticky fingers" and "fat fingers" made that vision impossible.
7. OLED — Organic light-emitting diode. Tang and Van Slyke invented the practical device at Kodak in 1987; it is now the basis of most premium smartphone and television displays.
8. Nanomotors — Self-propelled micro- or nanoscale devices, often for swimming in the bloodstream or targeting tumor cells.
9. R0 in academia — The mean number of new PhDs a tenure-track professor graduates during a career. If this exceeds the replacement rate for faculty slots, the academic job market becomes increasingly competitive.
10. Sheepskin effect — The labor-market return to completing a degree, above and beyond the skills the degree represents; the credential itself signals something to employers.

Caveats and limitations

This companion is built from the video's automatically generated captions and the author-supplied description. Machine transcripts make errors, especially on technical names, so the following caveats apply.

Related transmissions

The same tensions between credentials, craft, and research show up across the Frontier archive. If you are thinking about how to build tools without losing judgment, the pragmatic-epistemology post is a natural follow-up. If you are interested in how research systems scale and self-correct, the physics-engine blueprint treats some of the same questions in a formal C++20 setting. The building-a-brain and self-improving-instructions posts are the same problem at different scales.