Isn't the percentage of PhDs who go into academia nowadays lower in general because of fewer faculty openings than graduates? If that's true, then that's a huge confounding variable to consider, though perhaps that would be part of the "economic forces" aspect of it. And isn't it awfully similar, in basic substance, to an argument in the Unz Review (https://www.unz.com/article/are-women-destroying-academia-probably/) that having more women in academia is the culprit for this because on average women are more conscientious and more agreeable, which selects for, in his words, the "head girl" type (neurotypical "high-achiever"-coded) instead of the "brave truth-teller/lone eccentric genius" type (autistic nerd-coded)? It seems that all you did was to take the sexism out of it but say he was directionally correct. Do you think that he is? And if so, how would you suggest fixing it while also not reversing the progress that academia has made in terms of reducing the level of sexism that exists within it?
Another consideration here is that modern-day science is a very collaborative endeavor. I, like you, am in biomedical research (I just finished my PhD last year, in chem bio/med chem), so we both know how much being able to work with others matters in our world. Nobody can do biomedical research completely alone, and because of that, the ability to get along with colleagues, not create hostile/toxic working environments, being accommodating of those with different life situations/backgrounds, etc. are all extremely valuable in order to take advantage of all the talent in our world. Could it be that this, if true, is in some ways just academia selecting for what is more valuable in today's scientific environment, and thus isn't a bad thing at all? (I'm a "weird nerd" by your definition, albeit not nearly as successful of one, but I understand that just because something is against my interest doesn't mean it is against society's.)
Makes sense. Yeah, I think being agreeable might be more important in collaborative sciences *but* I’d also say super large teams are overrated in biology. If you read about the discovery of CRISPR and so on the teams were pretty small. And Charpentier at least seems like a weird nerd. Same as Kariko and drew weissman.
I think it’s entirely possible that the optimal level of weirdness for a biologist is lower than that of a physicist, but still, I think there’s a bit too much of careerist CNS style self promotion. So I think we’re not really at that optimum atm.
Yeah, the “top” scientists are at least as likely to be narcissistic self-promoters as they are to be weird autistic nerds, probably much more so. While of course narcissists are even more destructive to collaborative science than autistics are, they’re better at *appearing* to be more socially acceptable. I’d say focusing more on rooting out narcissists is a better use of effort in terms of improving collegiality than rooting out weird nerdy autistics.
Those appear to be the same thing at first, so the point stands! Socially adept is not the same as being genuinely good at being collaborative but appear to be at first. And autistic people can absolutely learn to be excellent team players who can work well with others - they more than anything just need some training and accommodation.
Has academia become less hospitable? Or have a lot of net-new attractive-to-Weird-Nerds jobs emerged on the market? The number of $200k+ TC SWE jobs has grown massively over the past decade, far outpacing the growth of academic STEM positions.
I believe that two other aspects are relevant:
- academia selects for personality types that may be negatively correlated with the Weird Nerd stereotype: a successful career in academia requires mentoring PhD students and teaching, which both may not be in the comfort zone of a lot of Weird Nerds.
- success ratio: academia is very much an up or out game. For those who don't exit early but fail to get tenure, the further career outlook is pretty bleak. Unless you collaborate really closely with industry during your academic career (like doing a couple of visiting researcher stints), switching tracks later might get really hard, so there's more than just an opportunity cost to it.
So maybe, for a lot of Weird Nerds, going into academia was merely the best shot they felt they had. But only a small fraction actually blossomed there, even before the rise of the Failed Corporatists.
Makes sense. I do think that the decreased ratio of science fair winners shows “something”. Does the extra completion actually drive away the best future potential academics (ofc many very good academics will not have participated in science fairs; but imo there is a correlation)
Oh, yeah, I wasn't saying "this is fine". While there are certainly examples of research conducted by private companies yielding tremendous benefits for society as a whole (e.g., Bell Labs, Xerox PARC), I can't imagine there *not* being a fallout from this brain drain out of academia.
Now, increased pressure on academia to become more attractive to top talent might in theory be a good thing. Unfortunately, short of kicking out all the bureaucrats, I don't have a great idea on how to do this.
I think (maybe I just hope) that it is slowly dawning on Western publics that the Academy that they have learned to revere as a magnet for the brightest and most independent-minded in society has in fact become a magnet for third-rate 'social justice' group-thinkers.
Decades ago, Camille Paglia pointed out that much of the really interesting academic work in the humanities was done by conservatives.
This was not because conservatives are inherently so much smarter and their arguments so much better, but because they had to go against the reigning academic climate. If their arguments and facts were not airtight, they'd be discredited, and by people looking for any reason to discredit them. One the other paw, even the laziest leftist scholarship was sure to be greeted with hosannas, publishing offers and tenure on offer to the brilliant minds researching queer performativity among dogs and similar burning issues crying out for serious study.
When I mentioned this to a feminist academic, she noted that the other source of interesting scholarship was gay men, mainly because other gay men were bitterly competitive, merciless critics, always looking for any flaw or reason to pick each other apart. By contrast, she said, lesbians rarely criticized each other and didn't want to hurt each others' feelings, so their academic output was correspondingly unrefined.
The best environment today for (highly capable) weird nerd types are imo in government-funded, especially military-funded research institutions (ex software/CS)or non-profit institution specializing in foundations (perimeter, aspen etc.), these places will really push state of the art research under a common cause (national defense) and will house incredible talents. Academia really isnt that friendly anymore the ambitious, as you‘ve noted and being partially replaced by well capitalized corporations (at least in some aspects)
I don't really agree. Once you involve government money there are too many prickly rules. Generally better off at the tech corporations or doing quantative finance.
Sure, it's not fundamental physics but AI research is also capable of changing the world, the pay is better and the environment often caters to people like you.
You might be right, its why i said excluding software/CS. The point is it's far easier in to do research in computer science and innovation can come from any place (academia, government, defense, industry), BUT industry will always only produce research favorable for their product/service, which can be detrimental in advancing the foundations of a field, even in CS. We need weird nerds kind of wandering, asking "what-if" questions in a variety of disciplines, hence creating such environments and nurture this is highly important. Imo you won't find this in industry, the short-term fixation will only lead to improving research which has already been done somewhere else in the past.
"... industry will always only produce research favorable for their product/service..."
So, go start an nonprofit entity devoting its goal towards being the difference. I swear to God, none of you have an ounce of experience in the market you speculate over; you really believe that you are the only ones who have seen this and then you think that no person has ever attempted a moral solution. It's just insane how ignorant you all sound.
we‘re not talking about profit/non-profit, or morality, the question is if and where upcoming researchers can explore impactful topics without being overly restricted by their environment.
It's funny you completely missed that I clearly stated why you can "explore impactful topics without being overly restricted by [your] environment" via nonprofit [self starter] projects. Researchers at foundations and such are not new; but you all would rather cry victim and for a mommy to hand you opportunities and funds instead of finding a way — including building the way yourselves. "Wahhhh, give me an easy, cushy office job where I get overpaid to undercontribute!" That's what you all sound like. If you don't see a place offering the kind of environment you want, go start the damn place.
Amazing, so the place where upcoming researchers/ "weird-nerd types" (universities) start are getting worse/don't welcome them like they used to, the places where they could do research are getting scarcer, and your solution is for them so simply self start? with what funds, equipment, facilities? do they have to fundraise now an entire organization, instead of, you know, exploring topics?
"Are we witnessing a radical restructuring of academia?"
No. We have witnessed "a radical restructuring of academia." It began in the 60s and was completed by the turn of the century. It's why I left graduate school in 1996 with only my master's thesis left to defend.
It happened in the Humanities and Education disciplines first, moved to the Social Science and primary and secondary schools, and now has come for STEM, so it's now glaringly obvious but was "slow-boiling a frog" until it was too late.
We're way past the tipping point, captured by the event horizon, on the backside of the tsunami being sucked out to sea already dead.
Can it be changed? Sure, but only in the same way that flora and fauna rise up from the nourishing nitric-rich soil after a forest is burnt to ash. A century? Hopefully that short. I'm not in despair. Nothing worth building can be done in a lifetime or by one.
It was Solomon's trap in his late years that he eventually rose above. He forgot a king and one lifetime are not enough. There is, indeed, nothing new under the sun and a time and purpose for everything under heaven.
"Cast your bread upon the waters for you will find it after many days."
I wrote a reply on my phone last night, but I guess I forgot to save.
First, I didn't go to grad school (History) until nearly a decade after my BA (PolySc). PS had required enough History, and the Social Sciences are close enough to the Humanities (but as I said, it happened in the Humanities first) that it was alike waking up Rip Van Winkle. It was a sea change from PS just a decade prior.
Second, my wife (of 36 years) and I married young and immediately started a family. By the time I started grad school, our oldest (of 5) was in kindergarten and our 2nd was in preschool. We had I third just before I quit. So, I was also seeing a seismic shift in primary school from what it had been in the early-mid 70s when I was there.
I saw no reason to throw good money after bad, nor did I want to read the Credentialing Cartel that academia had clearly become just so I could become, well, credentialed .
I did *not*, at that time, have an inkling it were ever eat STEM, so that wasn't conceptualized. It merely made it possible to easily identify the stomach contents. Well, that's enough of an idea for a comment. The short version?
Very interesting. Yeah I do biology so I thought it would be different but it’s very bureaucratised and so on plus the stuff you mention above that happened to Hum
Yes, the bureaucracy. One thing I didn't mention is the students, which is related to both the internal bureaucracy of academics and the larger world in this way:
Of the ~50 grad students who started the History master's program when I did only one other student besides me hoped to get a PhD. The university I went to didn't offer one, but I had hoped to do well enough to get into the best PhD program I could afford. About 3 or 4 others hoped to work in museums. One was a high school history teacher who wanted to up his game.
The rest were there to tick the MA box in order to get a promotion or pay raise at there jobs and, since it could be any MA, History is easier than Biology, right? (It is for most people most of the time.) The standards were much lower, and the grading scale was a joke by the time I went back, but not because of the professors. It's hard to keep the administrators paid without students. The number of admin v professor/teaching jobs at universities in recent years is an ridiculous ratio I no longer remember.
I'm bright but not brilliant. I have no doubt I could have received an MA in History from, say Yale, in the 1940s, but I've have worked a lot harder and would not have made as good of grades. I never once made less than an A in a +- system on a single assignment or exam, and I was done with coursework. I can't tell you the number of times I heard a student talk to a professor after class asking for an extension of a paper or even take a test later, and it was always granted.
The professors didn't like this. In the mid-90s my thesis advisor had a PhD from the University of Chicago and a B.S. in Physics from a top(ish) school I can't remember. He could barely keep his frustration and sense of impotence out of his voice, and insinuated in private even more. The youngest of my professors is still very active in the field and last year had a his newest book in History Book Club and is now considered one of the top in the field of Medieval Military History.
The second factor created a feedback loop with the first: the demand for MAs to get promotions and pay raises by the private sector. I started calling universities Credentialing Cartels shortly after leaving grad school in 1996, because it was clear to me that is what they had become. In the last 5 years I've probably read the term 5-6 times and wondered "How did it take you so long? Why are more people not recognizing it?"
Gradually, and inevitably (without early intervention), the ideologues drove out the Weird Nerd by first taking over admin and then the faculty as the indoctrinated students were fed into them by high schools and economic forces demanded more and more students. Add to lax standards cell phones then add AI. Both their ideology and their livelihood demand an increasing flow of demand. A lot more people recognize universities as an Economic Bubble (which they are) than Credentialing Cartels, but they depend on each other.
Edit: I should add my thesis advisor was in his early 60s in 1996 and also spoke and read several languages as well as Medieval German and Latin handwriting.
People tend to blame administrators for this trend, but I think faculty are equal culprits. Weird nerds want freedom to work on what they want to work on, but a lot of PIs (especially the younger ones, in my experience) insist on telling their students and post-docs what projects to pursue. I'm not sure whether this is an age effect (ie pre-tenure faculty playing it safe) or a cohort effect (ie academia losing its Weird Nerd culture). Another interesting data point is that math doesn't have this problem. One factor is that PhD students are funded by teaching not grants, so in their culture it's common for students to reject their advisor's projects. Another factor is that in math, PIs (not that they're even called that in math!) do not "automatically" get authorship; authorship is not given out so easily. So there is little incentive to build a big lab for citation-maxxing purposes.
I think the reason is that experimental science has become too complex to master for students right after a first degree: it is just not realistic anymore for a young researcher/student to do his own research.
We pretend that graduate students are independent researchers, but in reality they do not have enough experience to safely conduct research on their own.
A PI will have spent at least a decade building a lab containing equipment worth millions, he will have written many grant applications, including the one that actually funds the student. It is absolutely reasonable that he gets academic credit for the work, conceived by him and done in his lab.
It just doesn't make any sense to build a lab at great expense, and then have "independent researchers" come in, get all the credit for the work you envisioned, and then only mention you in the acknowledgements like "we would like to thank PI Joe for letting us use his tools".
There are also lots of constraints, samples are exchanged between groups, students share equipment, work together on complex projects. I have participated in research projects with 10-20 people, who were all doing a small part of the work. This is not an environment for weird nerds.
This is very different in math where students do their own research, funded by their own grants, and only mentored by their advisors.
I think we should be more honest about this. Either graduate students need to bring their own funding, i.e. conceive their own projects, apply for their own grants, pay for the overhead, pay for the use of the equipment, be responsible for the safety of their experiments, write the articles completely on their own, or they should be classified more like an apprentice, who works to gain mastery in the subject area. Then in turn they should have all of the employment protections given to regular employees.
For example, when Sheri Sangji burned to death because she wasn't trained or supervised properly to work with dangerous substances:
Not surprisingly, UCLA chemistry professor Patrick Harran and UCLA itself denied all responsibility and called her an independent researcher.
If a carpenter had his apprentice killed by letting him work unsupervised with deadly chemicals, he couldn't deflect the blame by calling the apprentice as an independent master carpenter, who just happened to use his shop.
At the moment, we have the worst of both worlds, the PI takes credit unless the student burns to death, then he was an independent researcher, who was just a guest in the PI's lab.
In fields where PhD students are not able to do independent research, then we should just be honest about this, and not structure and fund research programs around PhD students. If PhD students are really like experimental technicians / engineers in industry, they should be compensated like such people in industry, with more money but no PhD degree. Having basic research experimentalists be higher paid but not rewarded with PhDs, besides being more honest, would create better incentives to focus on increasing productivity in experimental sciences. As it is, PIs have no real incentive to increase productivity: labor is cheap and PI status is tied to the total number of PhDs awarded.
I doubt this is actually the case, though; Big Science has terrible ROI and should be mostly done away with, especially within academia. And I don't see why equipment cost should preclude independent research; my wife's PhD was in a field with equipment worth millions, but had a famous hands-off advisor who gave his students freedom. Senior researchers who aren't interested in granting this freedom should just go to industry or non-degree granting research institutions; they have no business being in academia.
> Either graduate students need to bring their own funding, i.e. conceive their own projects, apply for their own grants, pay for the overhead, pay for the use of the equipment, be responsible for the safety of their experiments, write the articles completely on their own
I agree, with the exception of the funding aspect. Acceptance into a PhD program could come with reimbursements for equipment usage etc. This means that each PhD student will be much more expensive, and the number of PhD students would necessarily go down dramatically (assuming fixed federal research funding). But at least it would mean there would be *some* people in their 20s - early 30s with guaranteed research freedom, as opposed to no one.
> In fields where PhD students are not able to do independent research, then we should just be honest about this, and not structure and fund research programs around PhD students. If PhD students are really like experimental technicians / engineers in industry, they should be compensated like such people in industry, with more money but no PhD degree.
I absolutely agree. I know several people who were just making samples for the team and the PI's buddies. No understanding, basically technician stuff, not even engineering. They were rewarded with an insane amount of publications and some got faculty positions. The really smart people spent years on hard problems with little to show, or even left the program. No surprise the weird nerds don't thrive under these conditions.
> Senior researchers who aren't interested in granting this freedom should just go to industry or non-degree granting research institutions; they have no business being in academia.
We don't want these people in industry either. I've been in industry for 15 years in multiple positions, and always have been given the freedom (and been expected) to drive my projects on my own. Something that academia did not prepare me for.
What I would like to see personally is to split the PhD into two degrees. First degree should be maybe 1-2 years to prepare people to work as independent researchers and to make people employable without a PhD. In this case, the students should have the right to hands-on tutoring by their PI, which in many cases (e.g. the sample makers above, they don't get).
Then for the real PhD, people should be classified as independent researchers doing their own work. This heavily depends on the field, the area of my PhD was a very mature one, it took me many years to even grasp the basics, I would not have been able to do independent research after a bachelor degree, and currently the only way to gain that knowledge is to stumble through a PhD where you are treated as a technician (when the PI wants you to do work for him), or as an independent researcher (when you don't deliver any immediate benefit to the PI).
I don't have a good solution for the instrumentation-heavy fields (e.g. the people building a small part of a particle detector, which itself is a small part of an accelerator). In this case, people really do have to work as a team and be managed heavily for the whole project to succeed.
I wonder if there are many weird nerds in these fields. My suspicion is not.
Either way, I personally feel that in these fields PhD students should be replaced by professional researchers and engineers, who should also be compensated as professionals, not students.
" Big Science has terrible ROI and should be mostly done away with, especially within academia." why do you say this? it seems to me very short sighted
I'm perhaps overly prejudiced by an experience years ago working with ENCODE data. Perhaps other Big Science efforts have been better managed, and perhaps ENCODE data has since become better organized.
There are the conceptual problems with the ENCODE Project that have been discussed by Lior Pachter and others, related to putting technicians in charge of science. But what I found striking is how bad a job they did from a data management and quality control perspective. To quote Grauer's paper "On the Immortality of Television Sets": "Did ENCODE generate massive amounts of reliable and easily accessible data? Judging by the computer memory it takes to store the data, ENCODE certainly delivered quantitatively. Unfortunately, the ENCODE data are neither easily accessible nor very useful—without ENCODE, researchers would have had to examine 3.5 billion nucleotides in search of function, with ENCODE, they would have to sift through 2.7 billion nucleotides. ENCODE’s biggest scientific sin was not being satisfied with its role as data provider; it assumed the small-science role of interpreter of the data, thereby performing a kind of textual hermeneutics on a 3.5-billion-long DNA text."
This lack of focus meant that it didn't (again, my experience might be outdated; perhaps these problems were fixed) even do a good job as a data generator and provider. One well-documented problem is how they confounded species with sequencing flowcells / lanes: "A reanalysis of mouse ENCODE comparative gene expression data" (Gilad & Mizrahi-Man, 2015). IMO, the problem is that within academia, the incentives and skills just don't exist for generating high-quality data at scale for *other* people to use. As a result, lab fiefdoms rushed to add their data with very little oversight and organization, in order to join the authorship list. So the ENCODE Consortium labs became interpreters of their own data, both due to incentives and out of necessity.
I think we should just make peace with this fact. Academia is a medieval-style institution, and that's okay for what it was created to do. But just as medieval institutions like the Holy Roman Empire were unable to succeed at nation-state-type tasks (eg efficient taxation and industrial-scale warfare), academia is unable to succeed at large-scale data generation tasks without transmogrifying into an industrial organization. And we don't want this to happen!
oh I see what you mean. Yeah I have heard about the ENCODE issues, ofc. I tend to agree with you- there are similar such efforts now in Science that I find very misguided. What are essentially non creative data generating processes that people do entire PhDs on and do not end up learning how to do proper science. But neither are these projects super well standardised so you get some weird in-between that ends up benefitting no one imo
I think the examination of startups is a bit misleading. Of my peers in mathematics PhD program and my friends from CalTech a huge fraction of them now work at companies like google. A few work at startups but many work at large tech firms of other large research enterprises.
I think this is a great thing. Sure, I worry about the future of academia but the fact that large companies now don't just reluctantly tolerate weird nerds but often cater to them by offering all sorts of benefits that let them avoid normal domestic duties is a huge benefit both to society and weird nerds.
I don’t know if that’s true. That it’s good for society. I link to a piece of mine where I discuss why academia does research that’s pretty crucial and irreplaceable long term…
I'm a social anthropologist, so it's possible my definition of a weird nerd (and I count myself as one) differs a little from yours, but my take is that the flight of the weird nerd from academia is due to two factors: 1) the corporatisation of academia itself, which means that corporate, entrepreneurial types who treat themselves as a brand they're selling now thrive there and 2) the fact that academic jobs are now heavily tied to external funding, and people doing highly applied research have a natural advantage. Sadly, the questions that weird nerds are interested in answering don't attract much in the way of grant funding - unless, of course, those questions involve the creation of machines-that-go-cha-ching (so, more like the weird nerds in 'Weird Science' than the ones in 'Revenge of the Nerds').
Academia had already become a less hospitable place for the Weird Nerd by the 1980s (and probably earlier), and the process has accelerated. When computer tech was mostly startups, there was plenty of room for the weird nerd in computer tech. Now that it has matured, there is much less room for nerds there, too.
I also want to object to the use of the winners of the Westinghouse science competition. Yes, competitions like that used to be won mainly by weird nerds but by the late 90s that was radically changing.
It's really damn hard to win those contests without lots of support nowdays, whether that is access to a lab or a mathematician advising you on your research. To some extent that's always been true it was just that was considered an odd thing that only extremely motivated weird nerds did.
The rise of programs that encourage youngsters, particularly women, to enter STEM fields and which provide access to these resources changes who ends up winning these prizes. You still have to be damn smart to win but the point is that it's shifted to require relatively more conscientiousness and organization rather than obsessiveness.
I don't necessarily know if that's a bad thing as long as we are aware of it and don't use it as a proxy for extreme brilliance or being a weird nerd.
I just realized something: While I think she was wrong to use Westinghouse/Intel winners as a proxy for weird nerds, as it turns out, using it sort of proved her point anyways, and in even stronger of a manner than she was originally going for. What the change in the life trajectories of Westinghouse/Intel winners shows is that weird nerds are being pushed out so early that it affected the ranks of *Westinghouse/Intel winners*. Not by *failed* corporate types mind you, but rather *successful* corporate types who were high achievers in school and did these competitions to get an edge for elite university admissions in the US, where extracurriculars are the main decisive factor as I've mentioned. That's why you see so many of them joining the ranks of more "corporate" professions nowadays, not because the corporate world has become more friendly to weird nerds.
I'm not sure if there are hard numbers for this, but based on what I know of the people who tend to do well at them (I know quite a few), they are usually more likely to be well-connected and are aiming for finance, medical school, big tech, etc. even as high schoolers as opposed to motivated by just intrinsic desire to do science. Their extraordinary conscientiousness and presentation skills, in combination with access to resources, tend to win the day. To put some specific names to this, they are more likely to be similar to Vivek Ramaswamy (politics aside and in terms of upbringing) than they are to be similar to someone like Einstein.
Hanfei Wang has a good point. From my personal experience: I was a high-school student in the 90s, on the spectrum and as weird a nerd as they get. Still, I never participated in any science competition.
The people competing mostly had very different personalities. They were the people who would socialize for the gains, who were politically savvy, who would suck up to the teachers for better grades etc. Many were also well-connected and had access to resources far beyond even what was available even to college students. None of these people had any strong interest in science. None of them had a lab at home, none of them did any experiments out of interest, none of them read science books.
One of them was a massive bully, who got a medal from the school for her extraordinary social engagement.
I do not want to downplay the skills of these people, though. They were extremely skilled self-promoters, who saw an opportunity, and executed perfectly. All of this in an area that they had no intrinsic interest in. Quite an achievement!
I felt completely out of place in this environment and just wanted to be left alone.
This made me leave academia at the very first opportunity and move to industry.
I had lots of sleepless nights before making this step, but I realized within the first week that most people in industry were smarter, more knowledgeable, more helpful, and also more tolerant of personal quirks.
Today, when we meet with academics, you can easily keep us apart:
We are the people with beards, long hair, wearing t-shirts and jeans.
The academics are the ones with military-grade haircuts wearing suits.
I have never regretted my decision to leave academia.
Isn't modern science in general trending towards that? My PhD cohort, while being far more tolerant of weirdness than a typical social group would be, trended more towards well-balanced conscientious "high achiever" types, especially in the more biologically-inclined section of my cohort (we had a "molecular biology" section and a "biological chemistry" section of our cohort and I was in the latter - the ones in the "biological chemistry" section tend to be a bit quirkier, myself included). I wouldn't be surprised if the same is true for other programs of a similar nature - within the sciences, the closer to biology it is, the less weirdness/quirkiness (apparently this has been studied, using autism quotient scores as the measure: https://docs.autismresearchcentre.com/papers/2001_BCetal_AQ.pdf). Of course, higher-ranked programs than mine will select more for obsessiveness, but would select for conscientiousness even more because labs in those schools are often sweatshops where you work obscene hours for low pay.
I agree biology is less populated by weird nerds than other subjects but while modern physics requires far more equipment to do my sense is that it's still largely done by weird nerds (as is math but that's a bit different).
Yes, it's true that it requires more and more access to large equipment and it's always required some amount of effort but that's very different than getting access to that equipment at 15. That's more about being in the right programs, having parents who know the right people and will drive you to the right places etc etc..
My sense is that winning Westinghouse has slanted substantially toward the types who just want to overachieve and go to Harvard relative to the types who are weird nerds even apart from any differences between subjects.
I don't think that's necessarily a bad thing given that before it was mostly just the weird kids of scientists who could win but I just don't think it's a good weirdness proxy.
Oh I see what you mean now, and it’s likely true. But even that has a bit of an explanation: at the elite level, admissions has just gotten significantly more competitive between 1990 and now. While it doesn’t show up in standardized test scores as much, it absolutely will show up in the extracurricular arms race such that the time commitment, quantity, and “wow”-ness of your extracurriculars basically are the most decisive factors in those applications since academics are already topped out because our standardized exams are so easy. That will *massively* select for conscientious and more neurotypical high-achiever types than weird nerds, who weren’t in contention for those elite schools anyways because they can’t bring themselves to do ECs which don’t interest them (I played the elite college admissions game and got into Vanderbilt and got a full ride, but it was a miserable experience for me playing that game). Of course, the well-connected always had an advantage in science fairs, but the change you’re seeing reflects the change in state of elite college admissions imo - people who win it nowadays see it as a golden ticket to an elite school instead of an award for scientific achievement for its own sake.
Sure, I'm not arguing it's not explained or even necessarily a bad thing (the arms race is but the change in Westinghouse) just that it's not a good proxy for being a weird nerd so the change in what winners go on to do doesn't show what was suggested (plausibly is true regardless though).
What would be a good proxy then? I personally think the autism quotient is as good as any, but I’m open to ideas. The paper I cited earlier also mentioned that a sample of Math Olympiad UK national winners had an AQ average of 24.5 (n=16), which is much higher than the 16.4 that was found in a general population and 17.6 among Cambridge students. I’d be interested in seeing a study on how their life trajectories differ between then and now.
"That will *massively* select for conscientious and more neurotypical high-achiever types than weird nerds, who weren’t in contention for those elite schools anyways because they can’t bring themselves to do ECs which don’t interest them (I played the elite college admissions game and got into Vanderbilt and got a full ride, but it was a miserable experience for me playing that game)" Our society has granted a whole lot of influence to university admissions officers, without really considering who these people are and what kind of things they value. I doubt that many of them are themselves innovators, or have any real ability to choose innovators.
In business, there seems to be some movement away from having the decision made by the hiring manager (the person for whom the new employee will be reporting and who will be responsible for his results) and toward more diffused hiring responsibility. I'm fine with having the prospective new employee interviewed by various people with who he will be working and considering that in the decision--this is generally a good idea...and also considering any advice from HR people...but the final decision should lie with the hiring manager. I've heard stories recently about companies where ANY person interviewing the candidate can veto the hiring; this is a very bad idea.
In the past 40 years academia has been "feminized" (along with many other institutions). Feminized institutions place emphasis on conformity, correct beliefs, and socialization. Hanania wrote about this: https://www.richardhanania.com/p/womens-tears-win-in-the-marketplace
These trends hurt Nerds (of both sexes) hardest. They are the most focused on "truth" and the least able to understand why others want them to be quiet or actively lie. They are also generally worse on following social queues and have a harder time fighting oblique social battles (rather than upfront factual disputes).
"generally worse on following social queues"...I think you meant social *cues*, but I like it the other way, too...imagine a lot of people waiting in line to sign up to support whatever they think they're supposed to support.
In the last couple decades people started gaming the college admissions system way more, so science fairs became sought after by people who just want to get into a good college. Websites like College Confidential didn't exist in the 90s.
Peter Gerdes and I had a comment thread here talking about this exact same explanation! Highly ambitious/conscientious mostly neurotypical overachievers are mostly who wins science fairs now because they feel like they need to in order to get into an elite college, and increasingly, those are also who tend to be in actual elite STEM PhD programs as well. At elite institutions, lab sciences require a *lot* of conscientiousness, and science being more collaborative and team-based today means that social skills are much more selected for (though not as intensely as in the other professions that overachievers tend to go into).
You can be a nerd in Tech (and have a good career) as long as you don't touch any 3rd rails.
But you cannot succeed in academia without explicitly bowing to ideology. At least the places that are free of mandatory ideology are growing fewer and farther between.
The thesis is consistent with my experience as someone who quit a Linguistics PhD program at a top university, ABD, to go find himself at a quant hedge fund (no regrets!). But I thought the analysis by Sheltzer was kind of weak:
1. His first result basically says that after ~15 years, X% of science fair winners have entered academia, and after ~30 years, 2X% of science fair winners have entered academia, which is exactly what you would expect if all science fair winners enter academia with a fixed probability per unit of time. He's comparing two rates without accounting for the fact that they are measured over different "exposures" [1]. To make his point more convincingly, he would need to have granular data about the actual date (or year, more coarsely) on which each science fair winner entered academia (rather than just "are they now in academia?") and use that to test whether the two cohorts have different rates of entering academia per fixed unit of time.
2. Google/DeepMind didn't exist at any meaningful scale in the 1990s/early 2000s, and likewise this was a protean age for quant finance (aside from some famous pioneers like RenTech, LTCM, DESCO, etc.). There was no "funnel" from college graduation into these jobs like there is today, and so it's unsurprising that the cohort who came of age and made their initial career entry during this period entered these jobs at much lower rates.
Side note, for many years, the hiring motto of Two Sigma Investments (founded 2001) was that they were the home for "nice geeks." It was a competitive advantage of theirs that they had the right culture/vibe to recruit the Weird Nerd R&D talent who were turned off by academia but also turned off by traditional wall street culture and wouldn't ordinarily have considered going to a hedge fund.
Isn't the percentage of PhDs who go into academia nowadays lower in general because of fewer faculty openings than graduates? If that's true, then that's a huge confounding variable to consider, though perhaps that would be part of the "economic forces" aspect of it. And isn't it awfully similar, in basic substance, to an argument in the Unz Review (https://www.unz.com/article/are-women-destroying-academia-probably/) that having more women in academia is the culprit for this because on average women are more conscientious and more agreeable, which selects for, in his words, the "head girl" type (neurotypical "high-achiever"-coded) instead of the "brave truth-teller/lone eccentric genius" type (autistic nerd-coded)? It seems that all you did was to take the sexism out of it but say he was directionally correct. Do you think that he is? And if so, how would you suggest fixing it while also not reversing the progress that academia has made in terms of reducing the level of sexism that exists within it?
Another consideration here is that modern-day science is a very collaborative endeavor. I, like you, am in biomedical research (I just finished my PhD last year, in chem bio/med chem), so we both know how much being able to work with others matters in our world. Nobody can do biomedical research completely alone, and because of that, the ability to get along with colleagues, not create hostile/toxic working environments, being accommodating of those with different life situations/backgrounds, etc. are all extremely valuable in order to take advantage of all the talent in our world. Could it be that this, if true, is in some ways just academia selecting for what is more valuable in today's scientific environment, and thus isn't a bad thing at all? (I'm a "weird nerd" by your definition, albeit not nearly as successful of one, but I understand that just because something is against my interest doesn't mean it is against society's.)
Makes sense. Yeah, I think being agreeable might be more important in collaborative sciences *but* I’d also say super large teams are overrated in biology. If you read about the discovery of CRISPR and so on the teams were pretty small. And Charpentier at least seems like a weird nerd. Same as Kariko and drew weissman.
I think it’s entirely possible that the optimal level of weirdness for a biologist is lower than that of a physicist, but still, I think there’s a bit too much of careerist CNS style self promotion. So I think we’re not really at that optimum atm.
Yeah, the “top” scientists are at least as likely to be narcissistic self-promoters as they are to be weird autistic nerds, probably much more so. While of course narcissists are even more destructive to collaborative science than autistics are, they’re better at *appearing* to be more socially acceptable. I’d say focusing more on rooting out narcissists is a better use of effort in terms of improving collegiality than rooting out weird nerdy autistics.
100%
I realise I switch between agreeable in the first part and then political. I mean agreeable more in the sense of socially adept not necessarily “good”
Those appear to be the same thing at first, so the point stands! Socially adept is not the same as being genuinely good at being collaborative but appear to be at first. And autistic people can absolutely learn to be excellent team players who can work well with others - they more than anything just need some training and accommodation.
Has academia become less hospitable? Or have a lot of net-new attractive-to-Weird-Nerds jobs emerged on the market? The number of $200k+ TC SWE jobs has grown massively over the past decade, far outpacing the growth of academic STEM positions.
I believe that two other aspects are relevant:
- academia selects for personality types that may be negatively correlated with the Weird Nerd stereotype: a successful career in academia requires mentoring PhD students and teaching, which both may not be in the comfort zone of a lot of Weird Nerds.
- success ratio: academia is very much an up or out game. For those who don't exit early but fail to get tenure, the further career outlook is pretty bleak. Unless you collaborate really closely with industry during your academic career (like doing a couple of visiting researcher stints), switching tracks later might get really hard, so there's more than just an opportunity cost to it.
So maybe, for a lot of Weird Nerds, going into academia was merely the best shot they felt they had. But only a small fraction actually blossomed there, even before the rise of the Failed Corporatists.
Makes sense. I do think that the decreased ratio of science fair winners shows “something”. Does the extra completion actually drive away the best future potential academics (ofc many very good academics will not have participated in science fairs; but imo there is a correlation)
Also see this
https://www.writingruxandrabio.com/p/can-the-markets-replace-academic
Oh, yeah, I wasn't saying "this is fine". While there are certainly examples of research conducted by private companies yielding tremendous benefits for society as a whole (e.g., Bell Labs, Xerox PARC), I can't imagine there *not* being a fallout from this brain drain out of academia.
Now, increased pressure on academia to become more attractive to top talent might in theory be a good thing. Unfortunately, short of kicking out all the bureaucrats, I don't have a great idea on how to do this.
making new institutions? see ARIA UK, Arc Institute etc
Academia increasingly is a haven for those who can Play The Game.
By definition, Weird nerds don't want to Play The Game and aren't very good at it.
I think (maybe I just hope) that it is slowly dawning on Western publics that the Academy that they have learned to revere as a magnet for the brightest and most independent-minded in society has in fact become a magnet for third-rate 'social justice' group-thinkers.
Decades ago, Camille Paglia pointed out that much of the really interesting academic work in the humanities was done by conservatives.
This was not because conservatives are inherently so much smarter and their arguments so much better, but because they had to go against the reigning academic climate. If their arguments and facts were not airtight, they'd be discredited, and by people looking for any reason to discredit them. One the other paw, even the laziest leftist scholarship was sure to be greeted with hosannas, publishing offers and tenure on offer to the brilliant minds researching queer performativity among dogs and similar burning issues crying out for serious study.
When I mentioned this to a feminist academic, she noted that the other source of interesting scholarship was gay men, mainly because other gay men were bitterly competitive, merciless critics, always looking for any flaw or reason to pick each other apart. By contrast, she said, lesbians rarely criticized each other and didn't want to hurt each others' feelings, so their academic output was correspondingly unrefined.
Yes a good point....conservative academics must be first class to have any chance. Hadn't thought about it that way before
💯
The best environment today for (highly capable) weird nerd types are imo in government-funded, especially military-funded research institutions (ex software/CS)or non-profit institution specializing in foundations (perimeter, aspen etc.), these places will really push state of the art research under a common cause (national defense) and will house incredible talents. Academia really isnt that friendly anymore the ambitious, as you‘ve noted and being partially replaced by well capitalized corporations (at least in some aspects)
I don't really agree. Once you involve government money there are too many prickly rules. Generally better off at the tech corporations or doing quantative finance.
Sure, it's not fundamental physics but AI research is also capable of changing the world, the pay is better and the environment often caters to people like you.
You might be right, its why i said excluding software/CS. The point is it's far easier in to do research in computer science and innovation can come from any place (academia, government, defense, industry), BUT industry will always only produce research favorable for their product/service, which can be detrimental in advancing the foundations of a field, even in CS. We need weird nerds kind of wandering, asking "what-if" questions in a variety of disciplines, hence creating such environments and nurture this is highly important. Imo you won't find this in industry, the short-term fixation will only lead to improving research which has already been done somewhere else in the past.
Ahh I didn't realize ex meant excluding...I thought you meant for example. My bad.
"... industry will always only produce research favorable for their product/service..."
So, go start an nonprofit entity devoting its goal towards being the difference. I swear to God, none of you have an ounce of experience in the market you speculate over; you really believe that you are the only ones who have seen this and then you think that no person has ever attempted a moral solution. It's just insane how ignorant you all sound.
What is your point Ligma?
we‘re not talking about profit/non-profit, or morality, the question is if and where upcoming researchers can explore impactful topics without being overly restricted by their environment.
It's funny you completely missed that I clearly stated why you can "explore impactful topics without being overly restricted by [your] environment" via nonprofit [self starter] projects. Researchers at foundations and such are not new; but you all would rather cry victim and for a mommy to hand you opportunities and funds instead of finding a way — including building the way yourselves. "Wahhhh, give me an easy, cushy office job where I get overpaid to undercontribute!" That's what you all sound like. If you don't see a place offering the kind of environment you want, go start the damn place.
Amazing, so the place where upcoming researchers/ "weird-nerd types" (universities) start are getting worse/don't welcome them like they used to, the places where they could do research are getting scarcer, and your solution is for them so simply self start? with what funds, equipment, facilities? do they have to fundraise now an entire organization, instead of, you know, exploring topics?
Yeah, although many areas I’d say don’t have this equivalents yet
"Are we witnessing a radical restructuring of academia?"
No. We have witnessed "a radical restructuring of academia." It began in the 60s and was completed by the turn of the century. It's why I left graduate school in 1996 with only my master's thesis left to defend.
It happened in the Humanities and Education disciplines first, moved to the Social Science and primary and secondary schools, and now has come for STEM, so it's now glaringly obvious but was "slow-boiling a frog" until it was too late.
We're way past the tipping point, captured by the event horizon, on the backside of the tsunami being sucked out to sea already dead.
Can it be changed? Sure, but only in the same way that flora and fauna rise up from the nourishing nitric-rich soil after a forest is burnt to ash. A century? Hopefully that short. I'm not in despair. Nothing worth building can be done in a lifetime or by one.
It was Solomon's trap in his late years that he eventually rose above. He forgot a king and one lifetime are not enough. There is, indeed, nothing new under the sun and a time and purpose for everything under heaven.
"Cast your bread upon the waters for you will find it after many days."
Very interesting. May I ask how you noticed this and how you conceptualised it back then?
I wrote a reply on my phone last night, but I guess I forgot to save.
First, I didn't go to grad school (History) until nearly a decade after my BA (PolySc). PS had required enough History, and the Social Sciences are close enough to the Humanities (but as I said, it happened in the Humanities first) that it was alike waking up Rip Van Winkle. It was a sea change from PS just a decade prior.
Second, my wife (of 36 years) and I married young and immediately started a family. By the time I started grad school, our oldest (of 5) was in kindergarten and our 2nd was in preschool. We had I third just before I quit. So, I was also seeing a seismic shift in primary school from what it had been in the early-mid 70s when I was there.
I saw no reason to throw good money after bad, nor did I want to read the Credentialing Cartel that academia had clearly become just so I could become, well, credentialed .
I did *not*, at that time, have an inkling it were ever eat STEM, so that wasn't conceptualized. It merely made it possible to easily identify the stomach contents. Well, that's enough of an idea for a comment. The short version?
Why I am a Weird Nerd, of course!
Very interesting. Yeah I do biology so I thought it would be different but it’s very bureaucratised and so on plus the stuff you mention above that happened to Hum
Yes, the bureaucracy. One thing I didn't mention is the students, which is related to both the internal bureaucracy of academics and the larger world in this way:
Of the ~50 grad students who started the History master's program when I did only one other student besides me hoped to get a PhD. The university I went to didn't offer one, but I had hoped to do well enough to get into the best PhD program I could afford. About 3 or 4 others hoped to work in museums. One was a high school history teacher who wanted to up his game.
The rest were there to tick the MA box in order to get a promotion or pay raise at there jobs and, since it could be any MA, History is easier than Biology, right? (It is for most people most of the time.) The standards were much lower, and the grading scale was a joke by the time I went back, but not because of the professors. It's hard to keep the administrators paid without students. The number of admin v professor/teaching jobs at universities in recent years is an ridiculous ratio I no longer remember.
I'm bright but not brilliant. I have no doubt I could have received an MA in History from, say Yale, in the 1940s, but I've have worked a lot harder and would not have made as good of grades. I never once made less than an A in a +- system on a single assignment or exam, and I was done with coursework. I can't tell you the number of times I heard a student talk to a professor after class asking for an extension of a paper or even take a test later, and it was always granted.
The professors didn't like this. In the mid-90s my thesis advisor had a PhD from the University of Chicago and a B.S. in Physics from a top(ish) school I can't remember. He could barely keep his frustration and sense of impotence out of his voice, and insinuated in private even more. The youngest of my professors is still very active in the field and last year had a his newest book in History Book Club and is now considered one of the top in the field of Medieval Military History.
The second factor created a feedback loop with the first: the demand for MAs to get promotions and pay raises by the private sector. I started calling universities Credentialing Cartels shortly after leaving grad school in 1996, because it was clear to me that is what they had become. In the last 5 years I've probably read the term 5-6 times and wondered "How did it take you so long? Why are more people not recognizing it?"
Gradually, and inevitably (without early intervention), the ideologues drove out the Weird Nerd by first taking over admin and then the faculty as the indoctrinated students were fed into them by high schools and economic forces demanded more and more students. Add to lax standards cell phones then add AI. Both their ideology and their livelihood demand an increasing flow of demand. A lot more people recognize universities as an Economic Bubble (which they are) than Credentialing Cartels, but they depend on each other.
Universities are inhospitable to Weird Nerds like Feynman: Can you image a story like this being written today: https://www.asc.ohio-state.edu/kilcup.1/262/feynman.html
Edit: I should add my thesis advisor was in his early 60s in 1996 and also spoke and read several languages as well as Medieval German and Latin handwriting.
Related: Welcome to the Conformity Gauntlet
https://greglukianoff.substack.com/p/welcome-to-the-conformity-gauntlet
Tyvm
People tend to blame administrators for this trend, but I think faculty are equal culprits. Weird nerds want freedom to work on what they want to work on, but a lot of PIs (especially the younger ones, in my experience) insist on telling their students and post-docs what projects to pursue. I'm not sure whether this is an age effect (ie pre-tenure faculty playing it safe) or a cohort effect (ie academia losing its Weird Nerd culture). Another interesting data point is that math doesn't have this problem. One factor is that PhD students are funded by teaching not grants, so in their culture it's common for students to reject their advisor's projects. Another factor is that in math, PIs (not that they're even called that in math!) do not "automatically" get authorship; authorship is not given out so easily. So there is little incentive to build a big lab for citation-maxxing purposes.
I don’t mainly blame admin as you can see in my post!
I think the reason is that experimental science has become too complex to master for students right after a first degree: it is just not realistic anymore for a young researcher/student to do his own research.
We pretend that graduate students are independent researchers, but in reality they do not have enough experience to safely conduct research on their own.
A PI will have spent at least a decade building a lab containing equipment worth millions, he will have written many grant applications, including the one that actually funds the student. It is absolutely reasonable that he gets academic credit for the work, conceived by him and done in his lab.
It just doesn't make any sense to build a lab at great expense, and then have "independent researchers" come in, get all the credit for the work you envisioned, and then only mention you in the acknowledgements like "we would like to thank PI Joe for letting us use his tools".
There are also lots of constraints, samples are exchanged between groups, students share equipment, work together on complex projects. I have participated in research projects with 10-20 people, who were all doing a small part of the work. This is not an environment for weird nerds.
This is very different in math where students do their own research, funded by their own grants, and only mentored by their advisors.
I think we should be more honest about this. Either graduate students need to bring their own funding, i.e. conceive their own projects, apply for their own grants, pay for the overhead, pay for the use of the equipment, be responsible for the safety of their experiments, write the articles completely on their own, or they should be classified more like an apprentice, who works to gain mastery in the subject area. Then in turn they should have all of the employment protections given to regular employees.
For example, when Sheri Sangji burned to death because she wasn't trained or supervised properly to work with dangerous substances:
https://cen.acs.org/safety/lab-safety/10-years-Sheri-Sangjis-death/97/i1
Not surprisingly, UCLA chemistry professor Patrick Harran and UCLA itself denied all responsibility and called her an independent researcher.
If a carpenter had his apprentice killed by letting him work unsupervised with deadly chemicals, he couldn't deflect the blame by calling the apprentice as an independent master carpenter, who just happened to use his shop.
At the moment, we have the worst of both worlds, the PI takes credit unless the student burns to death, then he was an independent researcher, who was just a guest in the PI's lab.
In fields where PhD students are not able to do independent research, then we should just be honest about this, and not structure and fund research programs around PhD students. If PhD students are really like experimental technicians / engineers in industry, they should be compensated like such people in industry, with more money but no PhD degree. Having basic research experimentalists be higher paid but not rewarded with PhDs, besides being more honest, would create better incentives to focus on increasing productivity in experimental sciences. As it is, PIs have no real incentive to increase productivity: labor is cheap and PI status is tied to the total number of PhDs awarded.
I doubt this is actually the case, though; Big Science has terrible ROI and should be mostly done away with, especially within academia. And I don't see why equipment cost should preclude independent research; my wife's PhD was in a field with equipment worth millions, but had a famous hands-off advisor who gave his students freedom. Senior researchers who aren't interested in granting this freedom should just go to industry or non-degree granting research institutions; they have no business being in academia.
> Either graduate students need to bring their own funding, i.e. conceive their own projects, apply for their own grants, pay for the overhead, pay for the use of the equipment, be responsible for the safety of their experiments, write the articles completely on their own
I agree, with the exception of the funding aspect. Acceptance into a PhD program could come with reimbursements for equipment usage etc. This means that each PhD student will be much more expensive, and the number of PhD students would necessarily go down dramatically (assuming fixed federal research funding). But at least it would mean there would be *some* people in their 20s - early 30s with guaranteed research freedom, as opposed to no one.
> In fields where PhD students are not able to do independent research, then we should just be honest about this, and not structure and fund research programs around PhD students. If PhD students are really like experimental technicians / engineers in industry, they should be compensated like such people in industry, with more money but no PhD degree.
I absolutely agree. I know several people who were just making samples for the team and the PI's buddies. No understanding, basically technician stuff, not even engineering. They were rewarded with an insane amount of publications and some got faculty positions. The really smart people spent years on hard problems with little to show, or even left the program. No surprise the weird nerds don't thrive under these conditions.
> Senior researchers who aren't interested in granting this freedom should just go to industry or non-degree granting research institutions; they have no business being in academia.
We don't want these people in industry either. I've been in industry for 15 years in multiple positions, and always have been given the freedom (and been expected) to drive my projects on my own. Something that academia did not prepare me for.
What I would like to see personally is to split the PhD into two degrees. First degree should be maybe 1-2 years to prepare people to work as independent researchers and to make people employable without a PhD. In this case, the students should have the right to hands-on tutoring by their PI, which in many cases (e.g. the sample makers above, they don't get).
Then for the real PhD, people should be classified as independent researchers doing their own work. This heavily depends on the field, the area of my PhD was a very mature one, it took me many years to even grasp the basics, I would not have been able to do independent research after a bachelor degree, and currently the only way to gain that knowledge is to stumble through a PhD where you are treated as a technician (when the PI wants you to do work for him), or as an independent researcher (when you don't deliver any immediate benefit to the PI).
I don't have a good solution for the instrumentation-heavy fields (e.g. the people building a small part of a particle detector, which itself is a small part of an accelerator). In this case, people really do have to work as a team and be managed heavily for the whole project to succeed.
I wonder if there are many weird nerds in these fields. My suspicion is not.
Either way, I personally feel that in these fields PhD students should be replaced by professional researchers and engineers, who should also be compensated as professionals, not students.
" Big Science has terrible ROI and should be mostly done away with, especially within academia." why do you say this? it seems to me very short sighted
I'm perhaps overly prejudiced by an experience years ago working with ENCODE data. Perhaps other Big Science efforts have been better managed, and perhaps ENCODE data has since become better organized.
There are the conceptual problems with the ENCODE Project that have been discussed by Lior Pachter and others, related to putting technicians in charge of science. But what I found striking is how bad a job they did from a data management and quality control perspective. To quote Grauer's paper "On the Immortality of Television Sets": "Did ENCODE generate massive amounts of reliable and easily accessible data? Judging by the computer memory it takes to store the data, ENCODE certainly delivered quantitatively. Unfortunately, the ENCODE data are neither easily accessible nor very useful—without ENCODE, researchers would have had to examine 3.5 billion nucleotides in search of function, with ENCODE, they would have to sift through 2.7 billion nucleotides. ENCODE’s biggest scientific sin was not being satisfied with its role as data provider; it assumed the small-science role of interpreter of the data, thereby performing a kind of textual hermeneutics on a 3.5-billion-long DNA text."
This lack of focus meant that it didn't (again, my experience might be outdated; perhaps these problems were fixed) even do a good job as a data generator and provider. One well-documented problem is how they confounded species with sequencing flowcells / lanes: "A reanalysis of mouse ENCODE comparative gene expression data" (Gilad & Mizrahi-Man, 2015). IMO, the problem is that within academia, the incentives and skills just don't exist for generating high-quality data at scale for *other* people to use. As a result, lab fiefdoms rushed to add their data with very little oversight and organization, in order to join the authorship list. So the ENCODE Consortium labs became interpreters of their own data, both due to incentives and out of necessity.
I think we should just make peace with this fact. Academia is a medieval-style institution, and that's okay for what it was created to do. But just as medieval institutions like the Holy Roman Empire were unable to succeed at nation-state-type tasks (eg efficient taxation and industrial-scale warfare), academia is unable to succeed at large-scale data generation tasks without transmogrifying into an industrial organization. And we don't want this to happen!
oh I see what you mean. Yeah I have heard about the ENCODE issues, ofc. I tend to agree with you- there are similar such efforts now in Science that I find very misguided. What are essentially non creative data generating processes that people do entire PhDs on and do not end up learning how to do proper science. But neither are these projects super well standardised so you get some weird in-between that ends up benefitting no one imo
In defense of corporate research (in bio in this case!) Google did give us AlphaFold 2 and all the protein structures that it predicted.
Those giant tech companies have a lot of oligopoly profits they can afford to plow into research if they feel secure enough to do it
I link to one of my posts where I discuss why academia does a very specific type of research that can’t be easily replaced.
https://www.writingruxandrabio.com/p/can-the-markets-replace-academic
I think the examination of startups is a bit misleading. Of my peers in mathematics PhD program and my friends from CalTech a huge fraction of them now work at companies like google. A few work at startups but many work at large tech firms of other large research enterprises.
I think this is a great thing. Sure, I worry about the future of academia but the fact that large companies now don't just reluctantly tolerate weird nerds but often cater to them by offering all sorts of benefits that let them avoid normal domestic duties is a huge benefit both to society and weird nerds.
I don’t know if that’s true. That it’s good for society. I link to a piece of mine where I discuss why academia does research that’s pretty crucial and irreplaceable long term…
I'm a social anthropologist, so it's possible my definition of a weird nerd (and I count myself as one) differs a little from yours, but my take is that the flight of the weird nerd from academia is due to two factors: 1) the corporatisation of academia itself, which means that corporate, entrepreneurial types who treat themselves as a brand they're selling now thrive there and 2) the fact that academic jobs are now heavily tied to external funding, and people doing highly applied research have a natural advantage. Sadly, the questions that weird nerds are interested in answering don't attract much in the way of grant funding - unless, of course, those questions involve the creation of machines-that-go-cha-ching (so, more like the weird nerds in 'Weird Science' than the ones in 'Revenge of the Nerds').
I agree
Academia had already become a less hospitable place for the Weird Nerd by the 1980s (and probably earlier), and the process has accelerated. When computer tech was mostly startups, there was plenty of room for the weird nerd in computer tech. Now that it has matured, there is much less room for nerds there, too.
I also want to object to the use of the winners of the Westinghouse science competition. Yes, competitions like that used to be won mainly by weird nerds but by the late 90s that was radically changing.
It's really damn hard to win those contests without lots of support nowdays, whether that is access to a lab or a mathematician advising you on your research. To some extent that's always been true it was just that was considered an odd thing that only extremely motivated weird nerds did.
The rise of programs that encourage youngsters, particularly women, to enter STEM fields and which provide access to these resources changes who ends up winning these prizes. You still have to be damn smart to win but the point is that it's shifted to require relatively more conscientiousness and organization rather than obsessiveness.
I don't necessarily know if that's a bad thing as long as we are aware of it and don't use it as a proxy for extreme brilliance or being a weird nerd.
I just realized something: While I think she was wrong to use Westinghouse/Intel winners as a proxy for weird nerds, as it turns out, using it sort of proved her point anyways, and in even stronger of a manner than she was originally going for. What the change in the life trajectories of Westinghouse/Intel winners shows is that weird nerds are being pushed out so early that it affected the ranks of *Westinghouse/Intel winners*. Not by *failed* corporate types mind you, but rather *successful* corporate types who were high achievers in school and did these competitions to get an edge for elite university admissions in the US, where extracurriculars are the main decisive factor as I've mentioned. That's why you see so many of them joining the ranks of more "corporate" professions nowadays, not because the corporate world has become more friendly to weird nerds.
what makes you so sure these kids are not representatives of the "weird nerd"?
I'm not sure if there are hard numbers for this, but based on what I know of the people who tend to do well at them (I know quite a few), they are usually more likely to be well-connected and are aiming for finance, medical school, big tech, etc. even as high schoolers as opposed to motivated by just intrinsic desire to do science. Their extraordinary conscientiousness and presentation skills, in combination with access to resources, tend to win the day. To put some specific names to this, they are more likely to be similar to Vivek Ramaswamy (politics aside and in terms of upbringing) than they are to be similar to someone like Einstein.
Hanfei Wang has a good point. From my personal experience: I was a high-school student in the 90s, on the spectrum and as weird a nerd as they get. Still, I never participated in any science competition.
The people competing mostly had very different personalities. They were the people who would socialize for the gains, who were politically savvy, who would suck up to the teachers for better grades etc. Many were also well-connected and had access to resources far beyond even what was available even to college students. None of these people had any strong interest in science. None of them had a lab at home, none of them did any experiments out of interest, none of them read science books.
One of them was a massive bully, who got a medal from the school for her extraordinary social engagement.
I do not want to downplay the skills of these people, though. They were extremely skilled self-promoters, who saw an opportunity, and executed perfectly. All of this in an area that they had no intrinsic interest in. Quite an achievement!
I felt completely out of place in this environment and just wanted to be left alone.
This made me leave academia at the very first opportunity and move to industry.
I had lots of sleepless nights before making this step, but I realized within the first week that most people in industry were smarter, more knowledgeable, more helpful, and also more tolerant of personal quirks.
Today, when we meet with academics, you can easily keep us apart:
We are the people with beards, long hair, wearing t-shirts and jeans.
The academics are the ones with military-grade haircuts wearing suits.
I have never regretted my decision to leave academia.
Isn't modern science in general trending towards that? My PhD cohort, while being far more tolerant of weirdness than a typical social group would be, trended more towards well-balanced conscientious "high achiever" types, especially in the more biologically-inclined section of my cohort (we had a "molecular biology" section and a "biological chemistry" section of our cohort and I was in the latter - the ones in the "biological chemistry" section tend to be a bit quirkier, myself included). I wouldn't be surprised if the same is true for other programs of a similar nature - within the sciences, the closer to biology it is, the less weirdness/quirkiness (apparently this has been studied, using autism quotient scores as the measure: https://docs.autismresearchcentre.com/papers/2001_BCetal_AQ.pdf). Of course, higher-ranked programs than mine will select more for obsessiveness, but would select for conscientiousness even more because labs in those schools are often sweatshops where you work obscene hours for low pay.
I agree biology is less populated by weird nerds than other subjects but while modern physics requires far more equipment to do my sense is that it's still largely done by weird nerds (as is math but that's a bit different).
Yes, it's true that it requires more and more access to large equipment and it's always required some amount of effort but that's very different than getting access to that equipment at 15. That's more about being in the right programs, having parents who know the right people and will drive you to the right places etc etc..
My sense is that winning Westinghouse has slanted substantially toward the types who just want to overachieve and go to Harvard relative to the types who are weird nerds even apart from any differences between subjects.
I don't think that's necessarily a bad thing given that before it was mostly just the weird kids of scientists who could win but I just don't think it's a good weirdness proxy.
Oh I see what you mean now, and it’s likely true. But even that has a bit of an explanation: at the elite level, admissions has just gotten significantly more competitive between 1990 and now. While it doesn’t show up in standardized test scores as much, it absolutely will show up in the extracurricular arms race such that the time commitment, quantity, and “wow”-ness of your extracurriculars basically are the most decisive factors in those applications since academics are already topped out because our standardized exams are so easy. That will *massively* select for conscientious and more neurotypical high-achiever types than weird nerds, who weren’t in contention for those elite schools anyways because they can’t bring themselves to do ECs which don’t interest them (I played the elite college admissions game and got into Vanderbilt and got a full ride, but it was a miserable experience for me playing that game). Of course, the well-connected always had an advantage in science fairs, but the change you’re seeing reflects the change in state of elite college admissions imo - people who win it nowadays see it as a golden ticket to an elite school instead of an award for scientific achievement for its own sake.
Sure, I'm not arguing it's not explained or even necessarily a bad thing (the arms race is but the change in Westinghouse) just that it's not a good proxy for being a weird nerd so the change in what winners go on to do doesn't show what was suggested (plausibly is true regardless though).
What would be a good proxy then? I personally think the autism quotient is as good as any, but I’m open to ideas. The paper I cited earlier also mentioned that a sample of Math Olympiad UK national winners had an AQ average of 24.5 (n=16), which is much higher than the 16.4 that was found in a general population and 17.6 among Cambridge students. I’d be interested in seeing a study on how their life trajectories differ between then and now.
"That will *massively* select for conscientious and more neurotypical high-achiever types than weird nerds, who weren’t in contention for those elite schools anyways because they can’t bring themselves to do ECs which don’t interest them (I played the elite college admissions game and got into Vanderbilt and got a full ride, but it was a miserable experience for me playing that game)" Our society has granted a whole lot of influence to university admissions officers, without really considering who these people are and what kind of things they value. I doubt that many of them are themselves innovators, or have any real ability to choose innovators.
In business, there seems to be some movement away from having the decision made by the hiring manager (the person for whom the new employee will be reporting and who will be responsible for his results) and toward more diffused hiring responsibility. I'm fine with having the prospective new employee interviewed by various people with who he will be working and considering that in the decision--this is generally a good idea...and also considering any advice from HR people...but the final decision should lie with the hiring manager. I've heard stories recently about companies where ANY person interviewing the candidate can veto the hiring; this is a very bad idea.
In the past 40 years academia has been "feminized" (along with many other institutions). Feminized institutions place emphasis on conformity, correct beliefs, and socialization. Hanania wrote about this: https://www.richardhanania.com/p/womens-tears-win-in-the-marketplace
These trends hurt Nerds (of both sexes) hardest. They are the most focused on "truth" and the least able to understand why others want them to be quiet or actively lie. They are also generally worse on following social queues and have a harder time fighting oblique social battles (rather than upfront factual disputes).
"generally worse on following social queues"...I think you meant social *cues*, but I like it the other way, too...imagine a lot of people waiting in line to sign up to support whatever they think they're supposed to support.
In the last couple decades people started gaming the college admissions system way more, so science fairs became sought after by people who just want to get into a good college. Websites like College Confidential didn't exist in the 90s.
Peter Gerdes and I had a comment thread here talking about this exact same explanation! Highly ambitious/conscientious mostly neurotypical overachievers are mostly who wins science fairs now because they feel like they need to in order to get into an elite college, and increasingly, those are also who tend to be in actual elite STEM PhD programs as well. At elite institutions, lab sciences require a *lot* of conscientiousness, and science being more collaborative and team-based today means that social skills are much more selected for (though not as intensely as in the other professions that overachievers tend to go into).
I'm not sure that big tech is any more welcoming to these types than academia (see James Damore), although salaries are indeed higher.
You can be a nerd in Tech (and have a good career) as long as you don't touch any 3rd rails.
But you cannot succeed in academia without explicitly bowing to ideology. At least the places that are free of mandatory ideology are growing fewer and farther between.
What if you are a Weird Nerd who left academia and became a successful Weird Corporatist? I consult nationally from a shed in my yard.
The thesis is consistent with my experience as someone who quit a Linguistics PhD program at a top university, ABD, to go find himself at a quant hedge fund (no regrets!). But I thought the analysis by Sheltzer was kind of weak:
1. His first result basically says that after ~15 years, X% of science fair winners have entered academia, and after ~30 years, 2X% of science fair winners have entered academia, which is exactly what you would expect if all science fair winners enter academia with a fixed probability per unit of time. He's comparing two rates without accounting for the fact that they are measured over different "exposures" [1]. To make his point more convincingly, he would need to have granular data about the actual date (or year, more coarsely) on which each science fair winner entered academia (rather than just "are they now in academia?") and use that to test whether the two cohorts have different rates of entering academia per fixed unit of time.
2. Google/DeepMind didn't exist at any meaningful scale in the 1990s/early 2000s, and likewise this was a protean age for quant finance (aside from some famous pioneers like RenTech, LTCM, DESCO, etc.). There was no "funnel" from college graduation into these jobs like there is today, and so it's unsurprising that the cohort who came of age and made their initial career entry during this period entered these jobs at much lower rates.
[1] https://en.wikipedia.org/wiki/Poisson_regression#%22Exposure%22_and_offset
Side note, for many years, the hiring motto of Two Sigma Investments (founded 2001) was that they were the home for "nice geeks." It was a competitive advantage of theirs that they had the right culture/vibe to recruit the Weird Nerd R&D talent who were turned off by academia but also turned off by traditional wall street culture and wouldn't ordinarily have considered going to a hedge fund.