PhD students are like startup founders. You're working for yourself, building something, racing against competitors, trying to get a foothold so that you can have the lifestyle you want in the future. That's why I think both groups get tempted to work a lot.
Are PhD students really racing against competitors? I can imagine that in some instances they might, e.g., in the race to discover the structure of DNA, but I don't think it's true in most cases.
This is actually not such a good example I think. A better example would be that you have a small improvement to an existing algorithm, but you know that another group across the globe is also working on/with that algorithm and maybe they also found out the same improvement as you did. If that is the case, then it will be a race on who publishes a paper about that first.
If you're not first, there's no reason to publish a paper anymore. And you just lost a good paper... (And paper count is really almost all there is in academia, at least that is what it seems to me sometimes)
Yes, always.
Consider when Pokémon Go came out, which for video game HCI researchers, was revolutionary from a social perspective. To give an idea: people standing on the street corner staring at their phone are now fellow trainers you can talk to.
Colleagues and I started data collection. Submitted a paper and were rejected as another group beat us at a prior conference because they could move just a bit faster. Sucks when a reviewer says "Unoriginal. Already saw this last month."
> Sucks when a reviewer says "Unoriginal. Already saw this last month."
That’s a systemic problem where we end up chasing the latest fad not because it’s interesting, but merely hoping to end up before someone else. Luckily, not all fields have that short-term outlook.
If most of science is normal science where you have a tough idea what you’re doing yes. You may not know the solution to your problem but you know the problem you’re trying to solve and mostly your trying to solve it with the same tools the rest of your discipline is using. Even when that’s not true you’re trying to finish a PhD faster than other PhD students or to publish more or better work so you can get a better postdoc. There’s a lot of competition.
I don't know, a lot of major discoveries happened independently in a very short time span. The discovery of dark energy was made by two independent groups. Einstein was racing against Hilbert to formulate general relativity. The quark model was developed independently by Gell-Mann and Ne'eman. Haumea was discovered independently by Mike Brown's group and Luis Ortiz Moreno's group. The Higgs mechanism was developed by three independent groups in 1964.
At a more prosaic level, during my own Ph.D. another group published a paper very similar to one I had been working on. (By a strange coincidence the lead author on that paper had a very similar name to my own.)
There is something to the idea that it's a good thing if you have to worry about other groups scooping your work. It means that you're probably on the right track --- it means your work is relevant and that you have a reasonable path to solving an important problem.
Yes. In some fields it is known that there are research groups that are dedicated to mining the conferences for ideas that can be quickly finished and published. The grad student who gets "scooped" is at a disadvantage because they have a variety of duties, including teaching, plus they are still getting up to speed on their technique.
Students are advised to withhold their best ideas from conferences.
My spouse is in a field like this, and "so-and-so got scooped today" is a not unfamiliar dinner table conversation.
For my PhD, I was lucky that my project required a huge pile of expensive and rather unique equipment, plus I really did have a skill advantage for the work I was doing. Two other labs replicated my work using the plans that I supplied to them, but I made sure that I got there first.
It varies a lot depending on the field. I was never particularly worried about being scooped during my PhD. I think that's because there were/are few people working on the same problem.
To give an example: One idea I've been working on lately is in a very narrow field where almost no one has the background to do it. (I consider myself barely qualified.) Someone publishes something similar perhaps a few times each decade. I'm really not worried about being scooped on this. I've thought before that whatever paper I write on this is basically written for someone a decade or more in the future, as possibly no contemporary researchers will care.
New ideas are worth everything in science, especially when doing experiments.
When I started my PhD in physics our group leader told us repeatedly not to talk to people from MIT or Caltech about our work too early, because there will be a postdoc who will do what will take you about 3 years in the space of 6 months.
> When I started my PhD in physics our group leader told us repeatedly not to talk to people from MIT or Caltech about our work too early, because there will be a postdoc who will do what will take you about 3 years in the space of 6 months.
Because of smarts, or because of their access to resources?
I would say experience instead of smarts, but also yes resources. Mainly the experience though: they know how to align a cavity, they know the theory, they can understand more quickly, etc.
The following high-paying or high status (or both) careers seem to necessitate a very heavy workload: law, finance, academia, business or strategy consulting, VC-funded start-ups, medicine, certain media jobs, certain software development jobs. Although the author talks about competition, it’s hard for me to escape the conclusion this is the major factor in the long hours work culture. If you want a high paying or high status job, it’s not clear to me how you escape the requirement for very hard work.
If you want a high paying or high status job, it’s not clear to me how you escape the requirement for very hard work.
Just tell people you work really hard but don't actually do it. You'd be surprised how many seemingly hard-working and successful people use this strategy. Some of them even believe themselves.
The work knows. That's actually the problem. Your work is constantly decaying. You have to beat the rate of decay.
Decay can take the form of materials and equipment (often homemade) deteriorating, forgetting what you've read or done (your lab notes are never good enough). Also, you're paying an opportunity cost for taking more time to finish, and you face a constant risk of getting scooped, your professor dying, getting knocked off your horse by personal / health issues, etc.
Klausewitz wrote that the only rational strategy in war is to make the maximum possible use of force. Your graduate research project is like that. Of course you also have to be cognizant of wearing out your mental equipment.
This makes sense if you are working on a problem in relative isolation, that's relatively well known, and who's research has low optionality to other work. However, there are many problems that are either utterly unknown, impractical to work on outside of a major collaboration, or option well into multiple papers.
It would seem that in academia, focusing on the latter will yield better long term career results. If you're a PhD student in a large collaboration, then you can't get scooped. If you are working on a problem no one knows/cares about, then getting scooped is a low risk, if you're working in an area where scooping is common - but you have good equipment/infrastructure - then you can re-use the work for something else.
In hindsight, almost all of my professors work was either part of a collaboration or focused on "high option" research. The few theoreticians and others working on areas which could be scooped seemed to either pick up more teaching/administrative work - or focus on "conservative" research.
There are many kinds of research. Sometimes you need long hours of mindless grinding to get the job done. Sometimes taking a long walk and letting your mind wander is a more productive use for your time. And sometimes you need to return to the work for 10-15 minutes at random hours when the computer finishes whatever it was doing.
If your work is cognitively demanding, there are maybe 20-30 productive hours in a week. If you try working harder, the chances are the marginal value of the extra hours becomes negative in the long term. You can get more productive hours by also doing less demanding work, but then there is a risk that the low-value work will take priority and reduce the time you can spend on demanding tasks. Especially when there are external deadlines involved.
Getting scooped is something many people are afraid of but which rarely happens in practice. Except maybe if you are working on some ultra-fashionable topics. I can't remember a single instance of it happening to anyone in my 15 years of string algorithms, space-efficient data structures, and their applications in bioinformatics. If someone manages to solve a problem another person was working on using the same approach as that person, it's typically years after the original person gave up.
I'm thinking about biochemistry in particular. There was also little or no scooping in my own field, experimental atomic spectroscopy. Maybe some fields are more cutthroat than others.
> The work knows. That's actually the problem. Your work is constantly decaying. You have to beat the rate of decay.
It's my experience that while not easy this is relatively straightforward for many valuable types of technical work.
The key I've found is that most people are incapable of making themselves seriously work through a textbook. Work through 10-15 good sized textbooks in your field and you are pretty much immediately near the top of your field in the professional world. Most of the churn in skills is happening at much lower levels of abstraction. If you have the fundamentals to understand the higher abstractions backing these changes making a switch is much easier.
As a concrete example, I'm at a big tech company using Typescript (TS). While I have less experience with TS than most of my coworkers many already consider me "one of the local experts". This is because:
1. I was previously aware of the challenges Racket trying to add types back in... I think the early 2000s? Forever ago. Anyway I could guess at the limitations TS would have as a result (e.g. when a type inference might fail).
2. I read a book about TS
3. I read the TS docs and changelogs
(2) and (3) to roughly 12 hours total.
I've had plenty of conversations with people who are much more "guru-ish" than I am and they report similar results e.g. a database expert who has spent 20 years almost exclusively leveraging knowledge discovered in the 80s despite keeping up to date with the field.
I don't think this holds in Academia, but most of us aren't there.
Accounting is the same, at least at the big auditing and consulting firms for partners and employees on the partnership track. There are huge volumes of work to grind through, and partners don't want to dilute their profits by hiring more associates.
Partners do like to hire and fire lots of associates so they don't have to pay as many senior associates and management. They sure as hell froze pay and fired a ton of people in 2020 and made everyone else double up. The pay when accounting for inflation in accounting has stayed flat or fallen until just recently. Now they're having a "talent shortage." Go figure.
The partners will happily make you work 50-90 hours during busy season which may be once or twice per year. And, don't get me started on the industry wide billing for time and lying on timesheets to meet the budget.
Imo, having worked in academia and industry (and having worked with academics transitioning into industry), phd students and profs have it way easier. (Versus the other careers the parent mentions)
Not only are the pace and pressure higher in industry, you're, to a large extent, doing what someone else wants you to do, which makes it much harder to push through. Academy is much more about doing the interesting stuff, even if you're working for a prof. Industry, someone has to do all the shitty, repetitive or uninteresting stuff, and that's mostly what the long hours are made of. Academics (including me) when they transition to industry often come in seeing all that stuff beneath them and think they are there because of their brain.
Industry work can be more interesting because you're working on high value problems, but it's way more work.
I think the variance wrt expected work load is way larger in industry though. Plenty of my friends working at FAANG companies have a really chill lifestyle, which is true of very few of my grad school friends. The notion that profs have way less work than industry in particular is crazy to me, their workloads are simply insane (at least for more junior faculty, but often also for senior ones).
I’ve worked in industry as a software developer, run a consulting business, founded a startup with seed funding, and been a TT professor. Only the last of these jobs required me to edit a document while my wife was in labor (a thing I have yet to live down.) The challenge in academia is the essentially unbounded nature of the work, and the fact that you’re constantly competing with what appear to be major stakes.
Also: my friends in industry have hobbies that don’t involve their work. Even the ones in FAANGs. Seriously I don’t even understand the concept: does going to the gym count?
> required me to edit a document while my wife was in labor
Not to be a jerk but, the only thing that required that was you.
The big thing I learned in management consulting is that the sun always rises tomorrow no matter what you do. People attach urgency to things, but unless you're a doctor or deployed military or one of the other rare instances where life depends on what you are doing, it really doesn't matter if it's late or not done, and it's important to prioritize the things they do matter.
My wife is happy to tell me I’m a jerk (over that one), so don’t feel bad about that part. But in this instance I had a camera-ready due with a hard deadline and nobody else was available to submit it. I guess I could have put a year’s worth of research at risk, but that’s much harder than you think (for what it’s worth, labor lasted a long time and had a lot of ‘down time’. But it’s a choice I still regret.)
Add to the list of challenges: in industry you usually have a pretty deep bench to back you up, and in academic projects you may be one out of one or two contributors.
> The challenge in academia is the essentially unbounded nature of the work, and the fact that you’re constantly competing with what appear to be major stakes.
The thing is, the stakes really are not that high and as academics we are too often complicit in letting the situation fester (and administrators are guilty of putting insane pressure on academics and pushing them to overwork). We ought to take a step back and see what’s really important.
You're right. It's funny you mention law because my brother-in-law is a partner at a large law firm. He took family time: 3 hours away from work on a Sunday evening for an extended family dinner. The guy works crazy hard.
The odd thing is how John Maynard Keynes predicted that we'd have a 15 hour work week. If you translated everyone's salary at the dinner table from a 60 hour work week to a 15 hour work week, they'd be just fine (except for me, I'd need a 20 hour work week). But where are those jobs? It seems like you've got to work crazy hours for a crazy large amount of money.
It is always instructive to go back to the sources and look at what Keynes actually wrote:
"But beyond this, we shall endeavour to spread the bread thin on the butter-to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!"[1]
He is talking not only talking about avarage working hours per worker, but about avarage working hours per person able to work. So let's have a look at the UK (where Keynes lived) of 2021:
Gross domestic product: GBP ~2.2 tril.[2]
Working age population: 41 mil.[3]
Overall weekly hours worked: ~ 1 bil.[4]
This means 24.4 hours work per week earning GPB ~1,080 (GBP ~53,650 per year) for each one in the working population.
If they had worked only 15 hours with the same productivity, earnings would have been GBP 663 weekly (GBP ~35,500 per year), or a total of GBP ~1,35 tril, which is aprox. the number of 1994/1995. (Or a little less than Greece today.[5])
In other words: if work could be evenly distributed over the population able to work, hourly rates were identical and people were content with the living standard of the mid 1990s (and all else being equal), a 15 hour work week for everyone would be enough for the UK.
[3] https://fred.stlouisfed.org/series/LFWA64TTGBQ647S -- One might argue that this number includes a lot of people who are not able to work or still in education and thus are not available for the workforce. On the other side it excludes a lot of people in older age, who might still be able to work. For a lack of better statistics and to make calculation easy, I assume that the two roughly cancel each other out. If you disagree, just use your own estimation and see if the result of the calculation really differs much.
I was going off the exact quote that you quoted from "Economic Possibilities For Our Grandchildren." Keynes even had the caveat that this may take longer with a war. He was extrapolating out productivity gains and moving it to reduced hours instead of extra wealth.
My question is still where can I earn 1/3 my salary but only work 20 hours a week.
It's definitely not true that everyone works hard. I've seen some crazy hard workers in low paying jobs, and some extremely lazy workers, same as high paying jobs.
I think, though, that in some areas there are just fewer "places to hide" if you're going to fuck around.
I don't think my experiences follow this, I've had very cushy well-paying jobs both in VC-funded startups and FAANG.
I think there is this romantic notion that the software industry is this cut-throat free market utopia where the best ideas win, and I just don't think it's so. Most large companies have a sizable monopoly and most VC funded companies of certain size usually have years of runway. In both cases, people can and do coast.
Obviously the new economic realities will have something to say about this, but in my experience long hours is not required for high pay and status in software.
Kinda agree, kinda disagree. For VC-funded startups, in the early stages I do think a very high workload is the norm, depending on the job. At most startups, there is simply not enough redundancy early on, especially before product market fit. Note I don't think this applies to everyone, but I've worked in early stage startups, and up to about the size of, say, 20 developers, it was easy to point out 5-10 folks where the company would have really been in deep doodoo if they didn't deliver.
Of course, for startups where it's already pretty clear they are the "favorites" (think Stripe, AirBnB a couple years after their founding, etc.), I don't think that's true. And FAANGs are a whole different ball game. They are pretty famous for having a large subset of their dev teams that are more on the 10-4 schedule.
Having worked in both in MANGA and startup land, I can can say that not delivering is a big problem in both places. Culturally, startup land seemed to have a lot of younger people without families, and I suspect that is one important factor why hours are less at big corp. The other factor is you're in demand of you're in the MANGA club, so the boss doesn't have that much leverage.
One of the most painful, obvious, and annoying lessons I learned in my software career is that if you're team needs to grind hard - it's unlikely to be successful.
Software is a great business when its working, high margin, low maintenance, and great growth economics. When companies are successful at it, they often are so successful that there is very little internal pressure to produce. No successful company survives as a country club, so there is a high pressure for high quality delivery of some number of projects per year, or at least appreciable progress. There simply isn't any need for the company to push people beyond a 40 hour work week.
I disagree with this when it comes to early-stage startups, primarily because the world often throws you curveballs that even the best planning can't avoid, and at an early stage there just isn't enough spare capacity to deal with it without scrambling.
I worked at a startup that was very well managed IMO. We selected a critical partner early on. About 2 years later that critical partner was bought out by a larger corporation, and they decided to close the business line we depended on (lots of other customers were kicked off and pissed). They even tried to bring up BS contract issues to get out of their contract early. We had fortunately already started adding an additional partner to take advantage of more functionality and not have a single point of failure, but it was still a mad, stressful dash to get everything switched over.
Startups just fundamentally don't have the level of repeatability/predictability at an early stage that would make bulletproof planning possible.
> Startups just fundamentally don't have the level of repeatability/predictability at an early stage that would make bulletproof planning possible.
Agreed, startups are often low margin, chaotic businesses at the outset, with poor product/market fit. At the start, these are not "good" software businesses, many of these startups will fail/be bought early/struggle.
Early in my career I sought out teams who had strong "hustle", when I eventually went to work at a big tech - I was surprised at the slower pace.
An academic position at an average no-name university is not at all "high status", nor does it pay well.
But the larger argument I have: It wasn't too long ago that you could have a good work-life balance in academia. The requirements to get tenure continue to increase. A common refrain is "The work I did to get tenure will not get anyone tenure in my department today." Quite a few professors I know said they worked regular 40 hour weeks to get tenure.
Fully agree with the author, though: No one is enforcing this from outside. This is a problem created by the academics themselves, and they alone have the power to fix it.
PhDs really are not high paying or high status jobs. They are most of the time passion projects, however, which makes it very easy for supervisors and the academic system to abuse them and gaslight them into working longer. The best students I have seen rarely stop thinking about their PhD, but they definitely stop working when it’s time.
>> If you want a high paying or high status job, it’s not clear to me how you escape the requirement for very hard work.
This applies to almost everyone, so I agree. The counterexamples everyone brings up are people who are top 5th percentile in office politics / social skills or top 5th percentile in natural gifted ability, or simply luck. For the majority of what you're talking about, it's certainly accurate - very hard work at some point was required.
For some reason this is becoming very hard to accept in society, and popular discourse is trending towards "everything is mostly luck."
There's long hours, and hard work. Some of those jobs listed are mostly long hours, not so much hard work. Finance, consulting, and similar are good examples of such - there's A LOT of waiting as the work traverses up and down the hierarchy, combined with the fact that the work often fundamentally revolves around deals/projects with a finite (and quite short) window.
Actually doing long hours combined with hard work is not really sustainable. In the end you'll end up with burnouts, resignations, lower quality work, etc.
But with that said, lots of sectors are plagued with the mentality of "That is the only way it works, always has been", combined with "if I could/had to do it, others should too".
Who knows, maybe change is coming. IIRC, investment banking analysts have gone through their tribulations regarding this. Everyone joining IB knows the hours are long, and the competition is fierce - but still juniors are saying enough is enough.
> I started my master's degree in 2018 and my PhD in 2020, and one of the most important things I learned during my master's was how NOT to work on weekends. I learned how to structure my time, deadlines and classwork in such a way that everything actually fits in an eight hour workday. I was also quite proud of this achievement, and barring my master's thesis and other major deadlines, I was moderately successful in defending my freedom. Not working on weekends seemed like a graduation from the messy life of an undergrad into the more structured life of an adult.
The rest of the article talks about the various pressures leading the author away from the disciplined 5-day work week. Competition and a desire to get ahead is the obvious one. The competitive pressure is especially prescient for academic jobs, where universities are busy churning out huge numbers of grad students but the number of open academic positions can never accommodate more than a small percentage of them.
Anecdotally, I've noticed two types of weekend workers: The first type simply works all the time. Instead of going idle, they gravitate toward their task list and start working on the next thing. For whatever reason (drive, anxiety, perceived pressures, boredom) they are wired to return to work by default and the 40 hour work week doesn't contain them.
The second type of weekend worker is not actually producing or even "working" more than 40 hours per week, but they struggle to contain their work into the Monday-Friday bounds. Their weekend work isn't to get ahead or go the extra mile. They work weekends because they spent half of their weekdays doing fun things (meeting up with friends, exploring hobbies, exercising, messing around online) and the only way to accomplish their work is to repeat this half-focused schedule 7 days per week.
Much like the author, they could contain their work neatly within a M-F, 9-5 schedule if they made an effort, but at every juncture they choose to follow spontaneous whims or to relax or procrastinate instead. Some of them may even like the eclectic and flexible work schedule that allows them to do the things they want when they want and to get their work done in the boring gaps in between.
There's also a 3rd type of person who doesn't really work much on weekends, but will wait until Saturday or Sunday to send out important e-mails and type up a storm in Slack so that it looks like they're working hard on the weekend. Frankly, this is the type I see most frequently at tech companies in the past few years: Try to engage with their messages on a Saturday and you won't get a response until Monday, but they'll go to great lengths to look like they were working hard all weekend or tell you that they worked all weekend on something.
Nicely said, I think this is a mostly accurate assessment. A lot of grad students are type A, a lot are type B, and plenty are both. Still, I do think a fourth kind exists (or a sub type of B) which is people who struggle to balance all the things in their lives and despite their best efforts end up with work on the weekends. Not everyone "choose[s] to follow spontaneous whims or to relax or procrastinate instead.", many people do their best to not procrastinate or to plan things out well, but particularly when there are deadlines it's simply hard to do.
I can also tell you from experience, although this is not regularly the case, when it comes time to try to submit a paper to a conference it's definitely possible that there is so much work that there is no way to contain the work to M-F. Which is of course how people get burned out, depressed, etc. And it's all too common in academia, with grads students having to balance teaching, research, classes, and more.
The competitive insentive is why social pressure not to work is important. It's a basic zero sum game: If you work more, I also have to work more, and the competitive advantage evens out.
If we can collectively decide (and enforce) not working weekends, we all benefit.
If the works not novel, there’s no PhD in it. So getting scooped can really suck…. And when doing wet work, access to the necessary machinery and getting enough time at the bench to make the desired white powder vs the undesired brown oil is critical. If you have to do x reactions to get good at doing reactions (ie you make what you set out to make), you can spent more time in a day or week running them and working them up. Or you can spend more months in the process.
I have yet to find the highly successful person who did it without working long hours. Sure you might not have to work weekends but the weekdays are very busy. Eventually you can coast on your past accomplishments but the path to get there requires a lot of effort, regardless of the field.
Yeah this is probably an unpopular opinion here. But weekends and 40 hour work weeks are for employees. If you want to spend your life as an employee, more power to you. I don't.
I don't know if I'm highly successful but I am probably a great PhD outcome who got a great postdoc position and I did it without working long hours. Extremely happy to be your counterexample :)
Huh. What I miss about being a grad student is working whenever I wanted and not working when it didn't make sense. Sometimes that meant working on a weekend, because I had an idea I was making progress on. Other times, even if it was Wednesday, I'd just not work.
The article seems to focus on working too much due to fear of competition. I would add that working in a fast-paced field (the author is doing ML at the Vector Institute) can feel like you're always in competition, but that it's important to remind yourself that you can choose your topic of research (if you have some horrible advisor that dictates what you do, I'm sorry, but recommend you find someone / someplace else). Even within ML, there are tons of pockets that nobody is really working on. That does make it harder to publish though, as your work will be "weird" or "not very hot".
tl;dr: people should choose topics and fields that fit their desired work profile. If you decide to compete in a heavily competitive subject, expect a competitive environment.
Honestly, it’s hard to imagine getting tenure without working on weekends. The demands are just too high. What finally broke me and forced me to draw a hard line was the pandemic Zoom shift, however: the stress of constant scheduling required me to set aside a little time just to be bored and recharge. I can do this because I’m post-tenure but I feel terrible for my more junior colleagues.
In my field, often times weekend work is needed to get actual work done, as most of my day is filled with meetings or other random things that require my attention — answering important emails, mentoring, managing the backlog, etc. Sure I have time in between these things, but rarely enough to focus uninterrupted for several consecutive hours. On a weekend I can knock out or catch up on things that would otherwise take me a full week or two under my normal week load. It’s just the way it is.
Numbers-wise, something like 3% of phds lead to professorships: it's not quite as competitive as becoming a professional athlete, but not far. Peak competition. Most professors watching their interview candidates nowadays freely admit they wouldn't have made the cut. That is unhealthy levels of competition bc at that point it is more of a lottery.
But there is something valuable going on here for everyone else. Most eventually realize they are not in the 3% bucket one way or the other, so then it becomes whether they want to invest in themselves or not. (Good) Phd programs are rare opportunities to work with the best from around the world and not worry about $ as much as results and other impact, and with like-minded people. For example, in my cohort, as folks switched for entertaining the idea of continuing in academics, they used the environment to do other equally hard things, such as startups with their peers. In both cases, only as limited as your work ethic. Others got cushy jobs where they could work hard till 5p but then clock out, but it was rarer, and even folks with families generally pulled the long hours: this was their time to grow.
For people happy to clock out and try to convince others to do the same, that devalues the experience for others who ARE there to work as hard & creatively as they can to do amazing things with like-minded individuals. A big part of picking a PhD program is the students you will be exploring with. Masters programs are great for learning technical skills and doing semi-structured technical projects, and a 9-5 view there makes sense. The PhD, which is researching new things, often starts after that, and a different environment. Think an artist colony where someone asks everyone to only do sports & gambling as soon as the sun sets.
I found myself agreeing with much of your comment, except: "For people happy to clock out and try to convince others to do the same, that devalues the experience for others who ARE there to work as hard & creatively as they can to do amazing things with like-minded individuals."
Is there possibly a separation between PhD students who work 9-5 with a mentality of hard work but strong time boundaries, versus PhD students who work 9-5 but don't plan to make the most of the opportunity?
I'm also instinctively wary of any romanticization of a PhD program, especially when working long hours. From anecdotal reports, national labs also provide the resources to do great research, with far more reasonable hours and less of a pressure to work longer hours. I've read of too many reports, which exist even if they are minority situations, where advisors take advantage of the power they have over students, such as in this HN thread: https://news.ycombinator.com/item?id=26367099
I believe that doing "amazing things with like-minded individuals" is worthwhile, but can also happen in the private sector where the pay is commensurate to the value of the provided work. In general, I'm very wary of the idea that working longer hours should ever be romanticized, because the potential downsides are severe and should not be underestimated (including physical/psychological health problems due to burnout from long hours and failed relationships/divorce).
Yes, weaker programs already do so. The faculty generally are top notch, and superstar students do seek out individual professors due to whatever alignment and succeed. So IMO it's more about the student, there are definitely places with different expectations already.
Likewise, in industry, it's not the hardest thing to get an r&d gig in say a hospital network or university as a research engineer supporting whatever lab(s). Basically a forever postdoc. Though salary may of course suffer.
One of the biggest things that comes out of a PhD is the ability to formulate and pursue open-ended problems. IMO that comes from (a) apprenticing and (b) repeated success/failure, both of which take time. So yes, you can stretch a PhD longer, or do a weak one and try to learn that later such as a postdoc. It's not hard to sniff out who has done the work / magically rocked it because they are magical / hasn't.
This also leads to a few bad outcomes for society at large.
- PhD programs can become abusive of their students/workers. Paying below subsistence or even charging money to study.
- PhD programs can become places where only those with no opportunity cost go.
- PhD students will stop caring about long term careers in their field - recognizing the impossibility.
- Science enters a hyper-competitive phase where scientists do not target ground breaking research. Instead focusing on small repeatable wins.
- Scientists adopt anti-competitive practices to secure their small slice of research funding including review kabals, and disuading research in conflicting views.
While I'm all for letting students take the risks they feel like taking, the current situation is the result of a systemic massive overproduction of PhDs through government funding. It's possible this overproduction of apprentices results in decreased productivity amongst those working in the field by introducing noise, and training costs to researchers.
If typical science PhD programs reduced their admittance by a factor of 3-10, you would still have a highly competitive field - but one where students, and in turn faculty have a career path should they produce results.
> If typical science PhD programs reduced their admittance by a factor of 3-10, you would still have a highly competitive field - but one where students, and in turn faculty have a career path should they produce results.
Then science would have a serious problem, because the only way to get things done is to abuse this cheap labour: it is much easier to get funding for a handful of PhD students or post-docs than for the equivalent in permanent staff. Large research universities would have a very hard time, and the way things work with funding agencies would have to change fundamentally.
I am not saying it would be bad, far from it. But in the current system, it’s easier for everybody to go along, so changing things will be hard. It does not depend on individual behaviours.
OTOH, I think that overall it is good to have PhDs going out and do stuff in unrelated fields. A PhD is still a wonderful way of broadening your horizons. We’d be better having more PhD politicians, for example, instead of lawyers. You can do a lot of things after a doctorate. I think the post-doc trap is much more insidious.
> If typical science PhD programs reduced their admittance by a factor of 3-10, you would still have a highly competitive field - but one where students, and in turn faculty have a career path should they produce results.
I don't know about that. A 3-10x reduction would mean grad classes of 1-2 students. In the grad classes I took, there was never "dead weight," everybody contributed in study sessions. Even those students who would go on to get tenure were learning from those who wouldn't. I'm not sure professors would even be motivated to teach such small courses. The price of graduate tuition would also go up by that 3-10x factor, so PhDs would be exclusively for the rich.
Several of my coworkers have PhDs, as do I. We all bring special skills, and working together, we combine those skills to great effect. None of us are faculty. That's not the only thing you can do with a PhD. Should ta/ra/sessional positions pay a living wage? Absolutely. But don't cut the rest of us out because you don't see our value.
If the latter is possible, then yes - however in a society with a declining population and modest productivity growth increasing the number of faculty positions 3-5x to account for current PhD program capacity is likely infeasible. Even a massive increase in science budgets at the federal level would likely only yield 2x the positions.
While Universities could cut administrative bloat to make way for faculty... they are unlikely to do so, and there are few levers to motivate them. With a looming student loan crises, they are likely to be forced to cut budgets significantly sometime in the next 10 years either through inflation or direct cuts.
The number of PhD candidates is partially based on the way the federal government cuts grants, at a roughly flat budget the only way to create more faculty positions is to cut PhD student grants. From a very rough perspective, the government could create roughly one career track research position for every ~1.5-2.5 PhD/PostDoc position grants they cut. If you doubled the science budget, then the 3-5x cut would work out as 1.5-3x, However due to the expectation on the number of students each professor has.. doubling the budget in the current system would simply yield the same problem with twice as many people involved.
You make some good points and I agree with your ballpark figures, but you seem very stuck in a status quo mindset. The entire NIH budget is about what we are giving Ukraine in aid. High-frequency trading alone saps billions from the economy. We could easily find money to triple our science expenditures if we wanted to.
And yes we'd need to evaluate faculty success differently so that we would no longer require each professor to run a lab of a half dozen grad students in order to have enough publications for tenure. But that's easily doable via a collective decision.
Fair, but we (hopefully) don’t need to fight the Ukraine war every year - and we have many more agencies than the NIH who need science funding. Choosing not to fund the military to support the NIH may directly translate to undesirable geopolitical outcomes.
Ex: 10X smaller enrollment. In top programs, most competitive undergraduate applicants have _already_ published top scientific papers. Making 10X more competitive will turn up the pressure even more.
Ex: More funding. Competing is hard. CS PhD programs, with assuming occasional well-paid CS internships, having slightly-below-median salaries. Definitely tough for those with families, but the rest can easily come out with savings even in the highest COL areas. So the real issue is opportunity cost when competing with FAANG etc (which grad students would be able to get into). I believe > 50% of US PhDs are foreign-born because american cs undergrads know they can work in cool tech envs for more $ and less work by going direct to the great companies courting them before graduation. Competing with FAANG salaries is hard - most professor salaries are below FAANG starting salaries, which is its own problem.
AFAIK the wages PhD stipends are pittances and make it very difficult for someone without money to survive, especially for non-STEM programs. If anything doing a PhD makes you worry more about money because you can't earn enough to survive
OP may have been referring to individuals who are independently wealthy or financially guaranteed by parents/relatives. Families may only be inclined towards generosity when someone is at least notionally working to better themselves.
Given a lack of career path through the PhD program, these programs are likely to attract individuals who don't need a career in the future.
Have you been in a PhD program yourself? It could be my field, but there are way more idealists than financially secure people. I was one, and my family was not wealthy.
To be fair, I went to a highly prestigious private institution where funding was guaranteed, and while pay was shit, it was enough for a single person to live reasonably comfortably (forget saving, of course). Some public schools I got offers from like Berkeley couldn’t even guarantee funding.
During my PhD (many years ago, so these figures are outdated) I was paid a stipend of £16k to teach undergraduate students who, upon graduation, were going to jobs paying £40k.
None of my PhD student peers had billionaire parents - but they almost all had parents who were reasonably well off. None of us had to rely on handouts from our parents, and none of us had loads of disposable income - but we all had a safety net.
The undergraduates who grew up in poverty and went to university to improve their lot in life? The £40k job offer represents financial security. It was rare that they'd forego that.
Yep, referring just to CS. Different even from other STEM fields.
In CS, stipends are below median income, but get close to median when students do the occasional R&D internship.
The main CS grad students I've seen struggle are those with families (ex: someone I respect had single income supporting his wife and ~3 kids!), big spenders, and those ignoring (many) corporate R&D internship offers. For everyone else, it's a nice life, and not hard to come out with savings. Likewise, if you're used to living a beyond-median-income lifestyle, e.g., software engineer returning to grad school, it may be hard to go back to living like the rest of America.
As others noted, much harder is the opportunity cost. Why would a student turn down an offer for 3X+ the US median income doing ads etc at some FAANG/VC co for a long, below-median, and relatively thankless & isolating phd experience? "The best minds of my generation are thinking about how to make people click ads".
> For people happy to clock out and try to convince others to do the same, that devalues the experience for others who ARE there to work as hard & creatively as they can to do amazing things with like-minded individuals.
“Clocking out” is not incompatible with “work as hard & creatively as they can to do amazing things with like-minded individuals”, far from it. Particularly in academia, where inspiration and reflection is important. Getting out for a walk, a dinner with friends, watching movies, etc are all opportunities to broaden your outlook on things and be more productive in the long run.
Far too often I have seen people “working hard” spending way too much time working on something because they did not really take the time to think about where they were going (including amongst the PhD students I work with in my lab). This leads to tunnel vision and abusive behaviour way too often.
> Numbers-wise, something like 3% of phds lead to professorships: it's not quite as competitive as becoming a professional athlete, but not far. Peak competition.
This is a false conclusion since less than 100% of PhDs have the ambission of professorship. It probably varies a lot by field but from anecdote most PhDs I know don't aim at staying longer that possibly postdoc, and even those are in minority - most are aiming for industry or research.
I have no idea if it's a significant fraction or not but IMO you'd need to to at least have some estimate to follow that assumption.
Agreed. Most of my labmates lost the desire to stay in academia by the end of PhD. Some of them didn't have it even in the beginning. Myself, I went from "I am not sure if academia is the right place for me" to "I am out of here and will never look back" in a span of two years.
> This is a false conclusion since less than 100% of PhDs have the ambission of professorship. It probably varies a lot by field but from anecdote most PhDs I know don't aim at staying longer that possibly postdoc, and even those are in minority - most are aiming for industry or research.
You can't say that without factoring in that by the end of their PhDs they realize that they cannot be professors. There aren't any positions for them. Competition tends to be nasty. The career wears them down.
If most PhDs had the option to stay, like say, 100 years ago, and conditions were better, the statistics would likely be very different!
Yep. Something like a masters is different, which explicitly is about not continuing in academic research. In contrast, the number of folks going on to a phd with one of their potential outcomes being a professorship is probably closer to 97% end or say 30-50% when the numbers get revealed, than the current 3% end.
There's another way to do a lot of work and get things done, but maybe it's not for everyone: it's basically three weeks on nonstop, one week off. However, this is more of a lab rat PhD approach, where you're working on ongoing experiments/procedures and your clock is kind of set by the experimental setup. Maybe it's more of a biotech thing, but I can see this being a fairly productive approach for compsci. You just have to make sure you take that week off. (*basically this means showing up in the morning, doing very little, leaving early and relaxing a bit and maybe reviewing the results a bit. If a PI or someone asks where you are, just let them know you're reviewing the results of your work binge).
It probably doesn't work as well for people with other commitments who want the stable 9-5 routine five days a week (I personally loathe that routine with a passion), but for some things this approach is really productive and allows you to focus without distractions. The ones who go nuts and burnout, though, they don't take that week off.
Although I agree that working weekends is common in academia, I believe that academia has more freedom in how to organise your time.
Are there any other jobs where someone can take a month off at a time to travel to conferences, meet with international colleagues, and go on long visits to other institutions? Compare this with any tech or high powered job that has much more sensitive deadlines.
The only real deadlines for academia are: PhD graduation (4-6 years), postdoc fellowship ends (2-4 years), and tenure (5-8 years).
Between those selection cutoffs you are absolutely free to structure your life however you want.
If it doesn’t work out, oh well, you go into industry and collect double pay.
> I believe that academia has more freedom in how to organise your time.
In hindsight this is definitely true. But that freedom will be worth a whole lot less, if you are not aware of it. This is definitely something that I wish I had known earlier and it wasn't entirely clear to me, until I read your post.
I feel this description overstates the freedom. Academia is full of rather hard deadlines, as the conference you're targeting a paper for might be the only fitting one for many months. Missing these deadlines can delay the whole "pipeline" of research, as you'd like to build on your existing work and properly cite it, but publishing subsequent work might make the original result harder to publish. Grant application deadlines are similarly hard.
In contrast, deadlines in engineering are often even not expected to be hit.
The "freedom" to travel for conferences is an integral part of your job as a researcher: either you network and sell your ideas, or you stay in obscurity. Of course travelling on someone else's dime can be fun, but the same is true of all business travel. It stops being fun the moment it becomes a chore and you'd rather be home putting time into your hobbies or family. Then it's just more work.
If writing grant applications is your hobby and you're married to your research, academia can be great, but the freedom doesn't include a balance with all the other parts of life. I know I'm not fully contradicting you, there is indeed a lot of freedom to choose what you work on. I just think it's important that people considering academia understand what the job actually consists of.
I agree with you. Academia gives freedom with how you structure your time, but you still need to put in the 60+ hours a week. Compared to other careers that require 60 hour weeks, however, academia is the most flexible.
Importantly though, within those 60 hours, you’re free to choose which conference to go, what kind of grant you want to write, and what kind of people you want to work with. I think this autonomy and self imposed goals is what leads in part to workaholism.
I'm a postdoc in Physics and I have very rarely worked on a weekend. Actually I would say that my workload even during my PhD has been far less than most people in the industry.
IMO it is a combination of luck and "work smarter, not harder". Most of the people I know who worked extremely hard produced about the same amount of work as me, the difference was that they spent literally hundreds of hours down strange corners that are not very helpful to their work. For example, one guy I know when writing his thesis did a huge amount of research into the theory of gravitational waves, because his work was related to gravitational wave detectors, but it wasn't directly related to his work. I also worked on gravitational wave detectors, but I just left the theory out, because it has been covered so many times. No one asked about it. You could make the case that it was interesting or important to learn the theory of gravitational waves, but he didn't really seem to be enjoying it; actually his life seemed quite stressful at that moment due to the high workload.
I would actually make the bold statement to say that, outside of cases of time consumption that are unavoidable, e.g. a lot of time in the lab/experimental difficulties, a huge amount of time is spent in academia because the student simply doesn't understand what they should spend their time on. This is because academia feels directionless or self-directed and the future of the research is uncertain, however you really should pause repeatedly and consider whether what you are doing is actually relevant to your research, and if it is "extra-curricular", you should ask yourself whether you are actually considering it with that mindset. Too many academics take their "research for fun" far too seriously.
I agree with this, I have read the theory for context sake but it has already been done. I'm building/using an instrument - that's where my focus is during the week, 9-5.
All those interesting corners are my free-time reading, because I enjoy it.
Yep. Scientists need to learn to differentiate between what needs to be done, and what they want to do. Many are woefully bad at estimating what needs to be done, and therefore they spend many hours doing what they want to do with the stressful mindset that it is something that needs to be done. It happened to me a few times, it happens to everyone to some degree, but the sooner you get a good compass for what you need to do the far smoother your life will be.
My personal example is when I was doing a theoretical project that had a very intense pure mathematical underpinning. I thought I needed to understand it, but my supervisor kept saying "just work through examples that you know the solutions to". I could have spent 100s or even 1000s of hours learning all this pure mathematics but actually if I would have done so it would not really have helped me today. Instead I just needed to do a few simple examples, which took maybe 20 to 40 hours? And then I was basically done with the research and could write the paper. Maybe if you ask some they will say "you didn't really do it, you didn't know the theory, you were just pretending!" but in the end I got the work done that I wanted to get done, and then went and had a beer.
> One point that was raised again and again was the fact of competition: (...) But at the same time, a lot of my friends in industry do not seem to feel the same guilt when taking the weekend off. (...) So there has to be something deeper going on here, since competition is not an exclusive feature of academia.
Come on, maybe not "exclusive" in an absolute sense, but the level of competition isn't even remotely similar. As an academic, I'm directly competing for something literally dozens of times a year (in each grant application, promotion application, evaluation application, salary complement application, grant/project request, paper submission). Each of these has an acceptance rate, we are competing directly among other academics. And each of these directly affects our career. Even when we are tenured professors we can't get away from it, because we have postdocs and students whose own careers depend directly on us and we can't let them down.
My friends in industry don't have an even remotely similar experience, of course competition exists in industry but it's not a constant pressure that you feel every day from several different sources.
The post makes an effort to argue that competition is just one factor among many, but I think it explains like 90% of the overwork phenomenon.
> But stability and money on average is still better in academia than being a musician.
That's just not true.
The average person in academia is a grad student. Their stability is low, they generally depend on one person who can easily take advantage of this. And unfortunately this often happens. And their pay is extremely low, below a living wage.
The next most common position are itinerant teachers of various kinds, adjuncts, course support staff, etc. These positions are also paid very poorly and have no stability. Also if you're fired, for any reason, even if your department is abolished (which happens!), that's it. Now you need to move maybe across the country. So not only is your pay abysmal, at any moment you could lose your family, friends, etc.
Even if you are tenured. Yeah, you have a job for life nominally. But, what if one of your fellow tenured professors turns out to be a verbally or even sexually abusive creep? Well, your options for moving are slim, and again, it means losing all of your connections and local life, for some unknown location at a roll of the dice. If you're lucky. Also, at the average university, in say, the average humanities department, a tenured professor after 20 years makes less than an entry-level CS grad at Google.
No. Academia is neither stable nor well paid, with only some exceptions.
It's only guaranteed employment in certain countries.
I completed my PhD in the UK, and I got funding that just about covered living expenses, and this is normally only for 3 years (13.500 pounds a year). The 4th and last year, normally reserved for write up, you normally get nothing (I managed to get 1/2 year extra funding). Even worse, any month longer than the 4 years and you start paying fines to the university if you still want to finish.
Also, my status was that of a student, not employment. No pension or other benefits.
> (in each grant application, promotion application, evaluation application, salary complement application, grant/project request, paper submission). Each of these has an acceptance rate, we are competing directly among other academics
I mean they're often very high acceptance rates though. The funding agency we use has a 90% acceptance rate. Basically you have to have some idea of what you are doing and they will accept you. It isn't as extreme as you are making out. There are many opportunities.
The main problem is that the opportunities are geographically constrained. I don't think there's a huge amount of competition for a given role, but the fact that it often requires uprooting your life (I moved from UK to Germany) is really bad
> I mean they're often very high acceptance rates though. The funding agency we use has a 90% acceptance rate. Basically you have to have some idea of what you are doing and they will accept you. It isn't as extreme as you are making out. There are many opportunities.
Not at all. You definitely aren't in the US.
Success rates are low! The stated funding rates by the NSF are around 25%. Depending on the program, rates for NIH tend to be a lot lower. R01 success rates are below 20%. Keep in mind, you're competing with the best and most recognizable names every time. This skews success rates, as a new or average researcher (I mean average as in, the mean person, not as in "mediocre") your rates will be lower.
In my country the best grants tend to have acceptance rates in the 10%-20% range. There are other grants with around 50% rate but they are barely enough to not outright close a lab (you're lucky if you can get 30K/year from those, per research group, not per person). There are no grants that I know of with a 90% acceptance rate.
And even if you compete for a 50% acceptance rate, which is relatively "easy", you are competing with another academics and the majority of them work on weekends.
I worked at the weekends when I was in academia and I work at the weekends now that I'm a software engineer. Why? Because I chose both careers because they were my area of intellectual interest and I do them for pleasure, not just money. It's not that complicated, for me at least.
Yes, agreed. My work when colleagues are offline is more freeform, and usually much more efficient. It's much easier to get work done when you're not paying attention to Slack. The idiots who think that software engineers can be more productive in offices are very hard for me to understand.
Beyond the sheer competitive aspects of modern academia, it is also a hodgepodge of ancient tradition and the reappropriation of the universities as schooling for the masses.
I bought in on the romantic dream of “monastic” pursuit of wisdom, where my whole being was dedicated to the search for tryth. That fits well with working most of the time.
Then, next to me sat the career professional, she wanted good grades and as little impact on her personal life from the “research experience” as she knew McKinsey was waiting.
The came family and the monastic dreams turned into work :)
I think it depends to some degree on the compromises you're ready to accept. I have worked 12 years in academia, got my PhD but then realized it is a lottery to proceed further. Nonetheless, Postdoc positions are still available in Germany and often no one wants those because they are so unreliable. I enjoy my work, produce good science - and for this, I have to ignore deadlines and requirements. I have a small child and I need to stay healthy, if only because I am responsible, for my child, for my research.
In the corona situation, I switched to breaking almost all deadlines - what is not possible is not possible. No weekend work. I work from 7am to 2 PM. If that means loosing my job, then that is it. But the outcome was different: I did not loose my job - others quit, because they could not stand the pressure, but I kept healthy and produced stable, ongoing work. Slower than everyone else, but higher quality (this is subjective, of course). I reduced my contract to 75% time (30hrs), which also helped. My compromise: No security for long term employment and no money saved. We do everything by bike, no car, no travels outside the country, no restaurants etc. - really everything cut down to the bare minimum. This is fine and it works for us/me, but a lot of PhDs in our department have higher expectations and I think they are better to pursue their career outside science. Look at how scientists worked a century ago: Their life was often equally precarious, but still their motivations helped them sustain constant progress.
Difficult to say, I got my MA/Diploma in Landscape Architecture, then went into Environmental Planning & GIS, then Cartography & Data Science [1]. Yes, it probably is (still) a niche field and maybe it is only possible to do like I do because it is.
Out of my many friends who have gone into academia, the most successful one works the least by far: maybe 1-2 hours per day on average. He publishes more papers than the rest of his department put together. At 35 he had more citations than 90% of PhDs at the end of their career.
This individual chose his field of study, university, department, collaborators etc etc based on time leverage. Everything is done remotely and there are practically no meetings. Every paper written produces material for another three. Every sentence is recycled from publication to publication if possible. Automation (python scripts) performs most of the analytical work. Conferences are attended exclusively to find more collaborators to feed into the system. Tier 1 journals are pursued relentlessly. Extra-research commitments are brutally minimised. Most importantly his data collection methodologies are hundreds of times more time-effective than his peers. We're at the beach together most summer afternoons. Having read some of his work, it's much better science than most as well.
The fact is most people aren't tactical about their careers. They don't want to hear what it takes to be successful in the system and would rather treat their PhD as a hoity-toity vocation than admit they have more in common with a blue-collar miner than with Kant or Mendel. If you want to cosplay as Darwin then you need to get on a boat for 5 years on your father's money and cross your fingers you have a brilliant revelation. But if you see your education as a job and apply a hacker mentality academia is easy ground to till.
Your friend sounds like a remarkable person, but it's not really clear from your account whether he has hacked career success in the career field of academia, or has found more efficient ways to do great science. If you accept there's a difference between those things, and that the latter is more important to the world at large, it's concerning.
I wouldn't say he does 'great science' and I don't think he would either. He pokes at the edges of a highly specific field using math/programming tools that are incredibly basic by HN's standards but almost magical to his peers. It's all leverage. It's good, solid work done very efficiently.
Will the next Pasteur or Semmelweis work this way? Probably not. But let's not pretend 99.9999% of PhD's are on that route anyway. And no big deal. Science would do better optimising for an army of effective workers than the odd genius. Those tend to make themselves known anyway.
From the outside it seems academia is an end, not a mean. Which in itself could be the full explanation. Also explains why even in "industry", as the author calls it, there's plenty of people willing to work (or appear to) a lot more than expected/required: to acquire more money/status/position/power/etc.
When you don't care to have your name plastered on numerous publications or on some company report, saying "no" to peer/company pressure is much easier.
Bonus point: Propaganda. Movies and tv shows are full of hard working heroes with impossible schedules and deadlines (literal most of the time) which makes for interesting entertainment but terrible lives.
I worked as a software engineer for 20 years in a computer science department and there are several reasons it is often necessary to work on weekends:
* Paper deadlines. If everyone planned well there would be no need to work on weekends, but nobody plans well, so a lot of the work is done last minute. Also research is international and as a result many conference paper deadlines are set randomly so as not to always be convenient for USA and inconvenient for everyone else. Hence some deadlines are on weekends. Again, if the work was well planned there would be no need even then to work on a weekend, you could submit the paper before the weekend, but no one plans well.
* Students time schedule. Students have classes. They sleep late and work nights. They go to parties and drink. They have exams and assignments that take all their attention during the week and then they get back to the research on weekends. If you are collaborating you often have to be connected in real time, even if it is remote work. This means working nights, weekends, really any time of the day. It can be like being on call.
* Profs and advisors try to get you to cram as much as possible into the paper to make it better. This means working up to the absolute last minute continuing to make minor revisions. I've seen changes made to papers in the last seconds before a deadline, and changes continue to be made for the hours grace period some conferences offer after the deadline. As long as the paper submission portal is open changes continue to be made. It's a stupid way to work, but it is common.
That said, since I was an employee I had more freedom to restrict my hours. Grad students on the other hand are slaves and are expected to work 80+ hour weeks. They are essentially almost always working. So they will try to access other people any time they encounter an issue. It's difficult to not interface with them during "off" hours because you know it might mean they can't get anything done until they talk with you, which can mean wasting days of time for them. They'll mention this at meetings, which makes it look like YOU are blocking the project. And you are, but for reasonable reasons. But that doesn't look reasonable to the student. And profs used to be students and in many cases that's basically the only life they know so they tend to sympathize with the students and not with you.
Some things that would help:
Plan for papers, posters, meetings so it's not all last minute. The conferences happen around the same time every year. It's not like it's a big surprise when the date is set.
Get grad students to form unions so they are not so overworked.
Set rules on what profs can task grad students with. Getting coffee, creating lessons for the profs classes, getting the profs laundry, being volunteered for work, being assigned more than 40hrs work/week, should not be on the list. This is a large part of why grad students are overworked.
Require management courses for profs so they learn that people are not machines. There is a reason we have a 40hr work week, and while going beyond 40hrs is _sometimes_ required, working longer hours continuously just damages productivity and is actually making the work go slower and burning people out.
112 comments
[ 3.1 ms ] story [ 189 ms ] threadIf you're not first, there's no reason to publish a paper anymore. And you just lost a good paper... (And paper count is really almost all there is in academia, at least that is what it seems to me sometimes)
Often the next things to explore in research are fairly obvious from what has just been tried.
Colleagues and I started data collection. Submitted a paper and were rejected as another group beat us at a prior conference because they could move just a bit faster. Sucks when a reviewer says "Unoriginal. Already saw this last month."
Ingress wasn't so popular that you could assume people in the same location as you were also playing.
At the height of its popularity, you were the odd one out on the street corner if you weren't playing Pokemon Go.
That’s a systemic problem where we end up chasing the latest fad not because it’s interesting, but merely hoping to end up before someone else. Luckily, not all fields have that short-term outlook.
At a more prosaic level, during my own Ph.D. another group published a paper very similar to one I had been working on. (By a strange coincidence the lead author on that paper had a very similar name to my own.)
There is something to the idea that it's a good thing if you have to worry about other groups scooping your work. It means that you're probably on the right track --- it means your work is relevant and that you have a reasonable path to solving an important problem.
Students are advised to withhold their best ideas from conferences.
My spouse is in a field like this, and "so-and-so got scooped today" is a not unfamiliar dinner table conversation.
For my PhD, I was lucky that my project required a huge pile of expensive and rather unique equipment, plus I really did have a skill advantage for the work I was doing. Two other labs replicated my work using the plans that I supplied to them, but I made sure that I got there first.
To give an example: One idea I've been working on lately is in a very narrow field where almost no one has the background to do it. (I consider myself barely qualified.) Someone publishes something similar perhaps a few times each decade. I'm really not worried about being scooped on this. I've thought before that whatever paper I write on this is basically written for someone a decade or more in the future, as possibly no contemporary researchers will care.
New ideas are worth everything in science, especially when doing experiments.
When I started my PhD in physics our group leader told us repeatedly not to talk to people from MIT or Caltech about our work too early, because there will be a postdoc who will do what will take you about 3 years in the space of 6 months.
Because of smarts, or because of their access to resources?
Just tell people you work really hard but don't actually do it. You'd be surprised how many seemingly hard-working and successful people use this strategy. Some of them even believe themselves.
Decay can take the form of materials and equipment (often homemade) deteriorating, forgetting what you've read or done (your lab notes are never good enough). Also, you're paying an opportunity cost for taking more time to finish, and you face a constant risk of getting scooped, your professor dying, getting knocked off your horse by personal / health issues, etc.
Klausewitz wrote that the only rational strategy in war is to make the maximum possible use of force. Your graduate research project is like that. Of course you also have to be cognizant of wearing out your mental equipment.
It would seem that in academia, focusing on the latter will yield better long term career results. If you're a PhD student in a large collaboration, then you can't get scooped. If you are working on a problem no one knows/cares about, then getting scooped is a low risk, if you're working in an area where scooping is common - but you have good equipment/infrastructure - then you can re-use the work for something else.
In hindsight, almost all of my professors work was either part of a collaboration or focused on "high option" research. The few theoreticians and others working on areas which could be scooped seemed to either pick up more teaching/administrative work - or focus on "conservative" research.
If your work is cognitively demanding, there are maybe 20-30 productive hours in a week. If you try working harder, the chances are the marginal value of the extra hours becomes negative in the long term. You can get more productive hours by also doing less demanding work, but then there is a risk that the low-value work will take priority and reduce the time you can spend on demanding tasks. Especially when there are external deadlines involved.
Getting scooped is something many people are afraid of but which rarely happens in practice. Except maybe if you are working on some ultra-fashionable topics. I can't remember a single instance of it happening to anyone in my 15 years of string algorithms, space-efficient data structures, and their applications in bioinformatics. If someone manages to solve a problem another person was working on using the same approach as that person, it's typically years after the original person gave up.
It's my experience that while not easy this is relatively straightforward for many valuable types of technical work.
The key I've found is that most people are incapable of making themselves seriously work through a textbook. Work through 10-15 good sized textbooks in your field and you are pretty much immediately near the top of your field in the professional world. Most of the churn in skills is happening at much lower levels of abstraction. If you have the fundamentals to understand the higher abstractions backing these changes making a switch is much easier.
As a concrete example, I'm at a big tech company using Typescript (TS). While I have less experience with TS than most of my coworkers many already consider me "one of the local experts". This is because:
1. I was previously aware of the challenges Racket trying to add types back in... I think the early 2000s? Forever ago. Anyway I could guess at the limitations TS would have as a result (e.g. when a type inference might fail).
2. I read a book about TS
3. I read the TS docs and changelogs
(2) and (3) to roughly 12 hours total.
I've had plenty of conversations with people who are much more "guru-ish" than I am and they report similar results e.g. a database expert who has spent 20 years almost exclusively leveraging knowledge discovered in the 80s despite keeping up to date with the field.
I don't think this holds in Academia, but most of us aren't there.
The partners will happily make you work 50-90 hours during busy season which may be once or twice per year. And, don't get me started on the industry wide billing for time and lying on timesheets to meet the budget.
Not only are the pace and pressure higher in industry, you're, to a large extent, doing what someone else wants you to do, which makes it much harder to push through. Academy is much more about doing the interesting stuff, even if you're working for a prof. Industry, someone has to do all the shitty, repetitive or uninteresting stuff, and that's mostly what the long hours are made of. Academics (including me) when they transition to industry often come in seeing all that stuff beneath them and think they are there because of their brain.
Industry work can be more interesting because you're working on high value problems, but it's way more work.
Also: my friends in industry have hobbies that don’t involve their work. Even the ones in FAANGs. Seriously I don’t even understand the concept: does going to the gym count?
Not to be a jerk but, the only thing that required that was you.
The big thing I learned in management consulting is that the sun always rises tomorrow no matter what you do. People attach urgency to things, but unless you're a doctor or deployed military or one of the other rare instances where life depends on what you are doing, it really doesn't matter if it's late or not done, and it's important to prioritize the things they do matter.
Add to the list of challenges: in industry you usually have a pretty deep bench to back you up, and in academic projects you may be one out of one or two contributors.
The thing is, the stakes really are not that high and as academics we are too often complicit in letting the situation fester (and administrators are guilty of putting insane pressure on academics and pushing them to overwork). We ought to take a step back and see what’s really important.
The odd thing is how John Maynard Keynes predicted that we'd have a 15 hour work week. If you translated everyone's salary at the dinner table from a 60 hour work week to a 15 hour work week, they'd be just fine (except for me, I'd need a 20 hour work week). But where are those jobs? It seems like you've got to work crazy hours for a crazy large amount of money.
"But beyond this, we shall endeavour to spread the bread thin on the butter-to make what work there is still to be done to be as widely shared as possible. Three-hour shifts or a fifteen-hour week may put off the problem for a great while. For three hours a day is quite enough to satisfy the old Adam in most of us!"[1]
He is talking not only talking about avarage working hours per worker, but about avarage working hours per person able to work. So let's have a look at the UK (where Keynes lived) of 2021:
This means 24.4 hours work per week earning GPB ~1,080 (GBP ~53,650 per year) for each one in the working population.If they had worked only 15 hours with the same productivity, earnings would have been GBP 663 weekly (GBP ~35,500 per year), or a total of GBP ~1,35 tril, which is aprox. the number of 1994/1995. (Or a little less than Greece today.[5])
In other words: if work could be evenly distributed over the population able to work, hourly rates were identical and people were content with the living standard of the mid 1990s (and all else being equal), a 15 hour work week for everyone would be enough for the UK.
[1] http://www.econ.yale.edu/smith/econ116a/keynes1.pdf
[2] https://www.statista.com/statistics/281744/gdp-of-the-united...
[3] https://fred.stlouisfed.org/series/LFWA64TTGBQ647S -- One might argue that this number includes a lot of people who are not able to work or still in education and thus are not available for the workforce. On the other side it excludes a lot of people in older age, who might still be able to work. For a lack of better statistics and to make calculation easy, I assume that the two roughly cancel each other out. If you disagree, just use your own estimation and see if the result of the calculation really differs much.
[4] Rough estimate from this table of weekly total hours: https://www.statista.com/statistics/280795/overall-weekly-ho...
[5] UK's GDP per capita * (15/24.4) ~= Greece's GDP, see: https://en.wikipedia.org/wiki/List_of_countries_by_GDP_(PPP)...
My question is still where can I earn 1/3 my salary but only work 20 hours a week.
Also, being a grad student is hardly "high status" tbf.
I think, though, that in some areas there are just fewer "places to hide" if you're going to fuck around.
I think there is this romantic notion that the software industry is this cut-throat free market utopia where the best ideas win, and I just don't think it's so. Most large companies have a sizable monopoly and most VC funded companies of certain size usually have years of runway. In both cases, people can and do coast.
Obviously the new economic realities will have something to say about this, but in my experience long hours is not required for high pay and status in software.
Of course, for startups where it's already pretty clear they are the "favorites" (think Stripe, AirBnB a couple years after their founding, etc.), I don't think that's true. And FAANGs are a whole different ball game. They are pretty famous for having a large subset of their dev teams that are more on the 10-4 schedule.
Software is a great business when its working, high margin, low maintenance, and great growth economics. When companies are successful at it, they often are so successful that there is very little internal pressure to produce. No successful company survives as a country club, so there is a high pressure for high quality delivery of some number of projects per year, or at least appreciable progress. There simply isn't any need for the company to push people beyond a 40 hour work week.
I worked at a startup that was very well managed IMO. We selected a critical partner early on. About 2 years later that critical partner was bought out by a larger corporation, and they decided to close the business line we depended on (lots of other customers were kicked off and pissed). They even tried to bring up BS contract issues to get out of their contract early. We had fortunately already started adding an additional partner to take advantage of more functionality and not have a single point of failure, but it was still a mad, stressful dash to get everything switched over.
Startups just fundamentally don't have the level of repeatability/predictability at an early stage that would make bulletproof planning possible.
Agreed, startups are often low margin, chaotic businesses at the outset, with poor product/market fit. At the start, these are not "good" software businesses, many of these startups will fail/be bought early/struggle.
Early in my career I sought out teams who had strong "hustle", when I eventually went to work at a big tech - I was surprised at the slower pace.
But the larger argument I have: It wasn't too long ago that you could have a good work-life balance in academia. The requirements to get tenure continue to increase. A common refrain is "The work I did to get tenure will not get anyone tenure in my department today." Quite a few professors I know said they worked regular 40 hour weeks to get tenure.
Fully agree with the author, though: No one is enforcing this from outside. This is a problem created by the academics themselves, and they alone have the power to fix it.
This applies to almost everyone, so I agree. The counterexamples everyone brings up are people who are top 5th percentile in office politics / social skills or top 5th percentile in natural gifted ability, or simply luck. For the majority of what you're talking about, it's certainly accurate - very hard work at some point was required.
For some reason this is becoming very hard to accept in society, and popular discourse is trending towards "everything is mostly luck."
Actually doing long hours combined with hard work is not really sustainable. In the end you'll end up with burnouts, resignations, lower quality work, etc.
But with that said, lots of sectors are plagued with the mentality of "That is the only way it works, always has been", combined with "if I could/had to do it, others should too".
Who knows, maybe change is coming. IIRC, investment banking analysts have gone through their tribulations regarding this. Everyone joining IB knows the hours are long, and the competition is fierce - but still juniors are saying enough is enough.
> I started my master's degree in 2018 and my PhD in 2020, and one of the most important things I learned during my master's was how NOT to work on weekends. I learned how to structure my time, deadlines and classwork in such a way that everything actually fits in an eight hour workday. I was also quite proud of this achievement, and barring my master's thesis and other major deadlines, I was moderately successful in defending my freedom. Not working on weekends seemed like a graduation from the messy life of an undergrad into the more structured life of an adult.
The rest of the article talks about the various pressures leading the author away from the disciplined 5-day work week. Competition and a desire to get ahead is the obvious one. The competitive pressure is especially prescient for academic jobs, where universities are busy churning out huge numbers of grad students but the number of open academic positions can never accommodate more than a small percentage of them.
Anecdotally, I've noticed two types of weekend workers: The first type simply works all the time. Instead of going idle, they gravitate toward their task list and start working on the next thing. For whatever reason (drive, anxiety, perceived pressures, boredom) they are wired to return to work by default and the 40 hour work week doesn't contain them.
The second type of weekend worker is not actually producing or even "working" more than 40 hours per week, but they struggle to contain their work into the Monday-Friday bounds. Their weekend work isn't to get ahead or go the extra mile. They work weekends because they spent half of their weekdays doing fun things (meeting up with friends, exploring hobbies, exercising, messing around online) and the only way to accomplish their work is to repeat this half-focused schedule 7 days per week.
Much like the author, they could contain their work neatly within a M-F, 9-5 schedule if they made an effort, but at every juncture they choose to follow spontaneous whims or to relax or procrastinate instead. Some of them may even like the eclectic and flexible work schedule that allows them to do the things they want when they want and to get their work done in the boring gaps in between.
There's also a 3rd type of person who doesn't really work much on weekends, but will wait until Saturday or Sunday to send out important e-mails and type up a storm in Slack so that it looks like they're working hard on the weekend. Frankly, this is the type I see most frequently at tech companies in the past few years: Try to engage with their messages on a Saturday and you won't get a response until Monday, but they'll go to great lengths to look like they were working hard all weekend or tell you that they worked all weekend on something.
I can also tell you from experience, although this is not regularly the case, when it comes time to try to submit a paper to a conference it's definitely possible that there is so much work that there is no way to contain the work to M-F. Which is of course how people get burned out, depressed, etc. And it's all too common in academia, with grads students having to balance teaching, research, classes, and more.
If we can collectively decide (and enforce) not working weekends, we all benefit.
either path can lead to good life and success
u r who u r, accept it
The article seems to focus on working too much due to fear of competition. I would add that working in a fast-paced field (the author is doing ML at the Vector Institute) can feel like you're always in competition, but that it's important to remind yourself that you can choose your topic of research (if you have some horrible advisor that dictates what you do, I'm sorry, but recommend you find someone / someplace else). Even within ML, there are tons of pockets that nobody is really working on. That does make it harder to publish though, as your work will be "weird" or "not very hot".
tl;dr: people should choose topics and fields that fit their desired work profile. If you decide to compete in a heavily competitive subject, expect a competitive environment.
But there is something valuable going on here for everyone else. Most eventually realize they are not in the 3% bucket one way or the other, so then it becomes whether they want to invest in themselves or not. (Good) Phd programs are rare opportunities to work with the best from around the world and not worry about $ as much as results and other impact, and with like-minded people. For example, in my cohort, as folks switched for entertaining the idea of continuing in academics, they used the environment to do other equally hard things, such as startups with their peers. In both cases, only as limited as your work ethic. Others got cushy jobs where they could work hard till 5p but then clock out, but it was rarer, and even folks with families generally pulled the long hours: this was their time to grow.
For people happy to clock out and try to convince others to do the same, that devalues the experience for others who ARE there to work as hard & creatively as they can to do amazing things with like-minded individuals. A big part of picking a PhD program is the students you will be exploring with. Masters programs are great for learning technical skills and doing semi-structured technical projects, and a 9-5 view there makes sense. The PhD, which is researching new things, often starts after that, and a different environment. Think an artist colony where someone asks everyone to only do sports & gambling as soon as the sun sets.
Is there possibly a separation between PhD students who work 9-5 with a mentality of hard work but strong time boundaries, versus PhD students who work 9-5 but don't plan to make the most of the opportunity?
I'm also instinctively wary of any romanticization of a PhD program, especially when working long hours. From anecdotal reports, national labs also provide the resources to do great research, with far more reasonable hours and less of a pressure to work longer hours. I've read of too many reports, which exist even if they are minority situations, where advisors take advantage of the power they have over students, such as in this HN thread: https://news.ycombinator.com/item?id=26367099
I believe that doing "amazing things with like-minded individuals" is worthwhile, but can also happen in the private sector where the pay is commensurate to the value of the provided work. In general, I'm very wary of the idea that working longer hours should ever be romanticized, because the potential downsides are severe and should not be underestimated (including physical/psychological health problems due to burnout from long hours and failed relationships/divorce).
Likewise, in industry, it's not the hardest thing to get an r&d gig in say a hospital network or university as a research engineer supporting whatever lab(s). Basically a forever postdoc. Though salary may of course suffer.
One of the biggest things that comes out of a PhD is the ability to formulate and pursue open-ended problems. IMO that comes from (a) apprenticing and (b) repeated success/failure, both of which take time. So yes, you can stretch a PhD longer, or do a weak one and try to learn that later such as a postdoc. It's not hard to sniff out who has done the work / magically rocked it because they are magical / hasn't.
- PhD programs can become abusive of their students/workers. Paying below subsistence or even charging money to study.
- PhD programs can become places where only those with no opportunity cost go.
- PhD students will stop caring about long term careers in their field - recognizing the impossibility.
- Science enters a hyper-competitive phase where scientists do not target ground breaking research. Instead focusing on small repeatable wins.
- Scientists adopt anti-competitive practices to secure their small slice of research funding including review kabals, and disuading research in conflicting views.
While I'm all for letting students take the risks they feel like taking, the current situation is the result of a systemic massive overproduction of PhDs through government funding. It's possible this overproduction of apprentices results in decreased productivity amongst those working in the field by introducing noise, and training costs to researchers.
If typical science PhD programs reduced their admittance by a factor of 3-10, you would still have a highly competitive field - but one where students, and in turn faculty have a career path should they produce results.
Then science would have a serious problem, because the only way to get things done is to abuse this cheap labour: it is much easier to get funding for a handful of PhD students or post-docs than for the equivalent in permanent staff. Large research universities would have a very hard time, and the way things work with funding agencies would have to change fundamentally.
I am not saying it would be bad, far from it. But in the current system, it’s easier for everybody to go along, so changing things will be hard. It does not depend on individual behaviours.
OTOH, I think that overall it is good to have PhDs going out and do stuff in unrelated fields. A PhD is still a wonderful way of broadening your horizons. We’d be better having more PhD politicians, for example, instead of lawyers. You can do a lot of things after a doctorate. I think the post-doc trap is much more insidious.
I don't know about that. A 3-10x reduction would mean grad classes of 1-2 students. In the grad classes I took, there was never "dead weight," everybody contributed in study sessions. Even those students who would go on to get tenure were learning from those who wouldn't. I'm not sure professors would even be motivated to teach such small courses. The price of graduate tuition would also go up by that 3-10x factor, so PhDs would be exclusively for the rich.
Several of my coworkers have PhDs, as do I. We all bring special skills, and working together, we combine those skills to great effect. None of us are faculty. That's not the only thing you can do with a PhD. Should ta/ra/sessional positions pay a living wage? Absolutely. But don't cut the rest of us out because you don't see our value.
Isn't research good for society?
While Universities could cut administrative bloat to make way for faculty... they are unlikely to do so, and there are few levers to motivate them. With a looming student loan crises, they are likely to be forced to cut budgets significantly sometime in the next 10 years either through inflation or direct cuts.
The number of PhD candidates is partially based on the way the federal government cuts grants, at a roughly flat budget the only way to create more faculty positions is to cut PhD student grants. From a very rough perspective, the government could create roughly one career track research position for every ~1.5-2.5 PhD/PostDoc position grants they cut. If you doubled the science budget, then the 3-5x cut would work out as 1.5-3x, However due to the expectation on the number of students each professor has.. doubling the budget in the current system would simply yield the same problem with twice as many people involved.
And yes we'd need to evaluate faculty success differently so that we would no longer require each professor to run a lab of a half dozen grad students in order to have enough publications for tenure. But that's easily doable via a collective decision.
Ex: 10X smaller enrollment. In top programs, most competitive undergraduate applicants have _already_ published top scientific papers. Making 10X more competitive will turn up the pressure even more.
Ex: More funding. Competing is hard. CS PhD programs, with assuming occasional well-paid CS internships, having slightly-below-median salaries. Definitely tough for those with families, but the rest can easily come out with savings even in the highest COL areas. So the real issue is opportunity cost when competing with FAANG etc (which grad students would be able to get into). I believe > 50% of US PhDs are foreign-born because american cs undergrads know they can work in cool tech envs for more $ and less work by going direct to the great companies courting them before graduation. Competing with FAANG salaries is hard - most professor salaries are below FAANG starting salaries, which is its own problem.
AFAIK the wages PhD stipends are pittances and make it very difficult for someone without money to survive, especially for non-STEM programs. If anything doing a PhD makes you worry more about money because you can't earn enough to survive
Given a lack of career path through the PhD program, these programs are likely to attract individuals who don't need a career in the future.
To be fair, I went to a highly prestigious private institution where funding was guaranteed, and while pay was shit, it was enough for a single person to live reasonably comfortably (forget saving, of course). Some public schools I got offers from like Berkeley couldn’t even guarantee funding.
None of my PhD student peers had billionaire parents - but they almost all had parents who were reasonably well off. None of us had to rely on handouts from our parents, and none of us had loads of disposable income - but we all had a safety net.
The undergraduates who grew up in poverty and went to university to improve their lot in life? The £40k job offer represents financial security. It was rare that they'd forego that.
In CS, stipends are below median income, but get close to median when students do the occasional R&D internship.
The main CS grad students I've seen struggle are those with families (ex: someone I respect had single income supporting his wife and ~3 kids!), big spenders, and those ignoring (many) corporate R&D internship offers. For everyone else, it's a nice life, and not hard to come out with savings. Likewise, if you're used to living a beyond-median-income lifestyle, e.g., software engineer returning to grad school, it may be hard to go back to living like the rest of America.
As others noted, much harder is the opportunity cost. Why would a student turn down an offer for 3X+ the US median income doing ads etc at some FAANG/VC co for a long, below-median, and relatively thankless & isolating phd experience? "The best minds of my generation are thinking about how to make people click ads".
“Clocking out” is not incompatible with “work as hard & creatively as they can to do amazing things with like-minded individuals”, far from it. Particularly in academia, where inspiration and reflection is important. Getting out for a walk, a dinner with friends, watching movies, etc are all opportunities to broaden your outlook on things and be more productive in the long run.
Far too often I have seen people “working hard” spending way too much time working on something because they did not really take the time to think about where they were going (including amongst the PhD students I work with in my lab). This leads to tunnel vision and abusive behaviour way too often.
This is a false conclusion since less than 100% of PhDs have the ambission of professorship. It probably varies a lot by field but from anecdote most PhDs I know don't aim at staying longer that possibly postdoc, and even those are in minority - most are aiming for industry or research.
I have no idea if it's a significant fraction or not but IMO you'd need to to at least have some estimate to follow that assumption.
You can't say that without factoring in that by the end of their PhDs they realize that they cannot be professors. There aren't any positions for them. Competition tends to be nasty. The career wears them down.
If most PhDs had the option to stay, like say, 100 years ago, and conditions were better, the statistics would likely be very different!
It probably doesn't work as well for people with other commitments who want the stable 9-5 routine five days a week (I personally loathe that routine with a passion), but for some things this approach is really productive and allows you to focus without distractions. The ones who go nuts and burnout, though, they don't take that week off.
Are there any other jobs where someone can take a month off at a time to travel to conferences, meet with international colleagues, and go on long visits to other institutions? Compare this with any tech or high powered job that has much more sensitive deadlines.
The only real deadlines for academia are: PhD graduation (4-6 years), postdoc fellowship ends (2-4 years), and tenure (5-8 years).
Between those selection cutoffs you are absolutely free to structure your life however you want.
If it doesn’t work out, oh well, you go into industry and collect double pay.
In hindsight this is definitely true. But that freedom will be worth a whole lot less, if you are not aware of it. This is definitely something that I wish I had known earlier and it wasn't entirely clear to me, until I read your post.
In contrast, deadlines in engineering are often even not expected to be hit.
The "freedom" to travel for conferences is an integral part of your job as a researcher: either you network and sell your ideas, or you stay in obscurity. Of course travelling on someone else's dime can be fun, but the same is true of all business travel. It stops being fun the moment it becomes a chore and you'd rather be home putting time into your hobbies or family. Then it's just more work.
If writing grant applications is your hobby and you're married to your research, academia can be great, but the freedom doesn't include a balance with all the other parts of life. I know I'm not fully contradicting you, there is indeed a lot of freedom to choose what you work on. I just think it's important that people considering academia understand what the job actually consists of.
Importantly though, within those 60 hours, you’re free to choose which conference to go, what kind of grant you want to write, and what kind of people you want to work with. I think this autonomy and self imposed goals is what leads in part to workaholism.
IMO it is a combination of luck and "work smarter, not harder". Most of the people I know who worked extremely hard produced about the same amount of work as me, the difference was that they spent literally hundreds of hours down strange corners that are not very helpful to their work. For example, one guy I know when writing his thesis did a huge amount of research into the theory of gravitational waves, because his work was related to gravitational wave detectors, but it wasn't directly related to his work. I also worked on gravitational wave detectors, but I just left the theory out, because it has been covered so many times. No one asked about it. You could make the case that it was interesting or important to learn the theory of gravitational waves, but he didn't really seem to be enjoying it; actually his life seemed quite stressful at that moment due to the high workload.
I would actually make the bold statement to say that, outside of cases of time consumption that are unavoidable, e.g. a lot of time in the lab/experimental difficulties, a huge amount of time is spent in academia because the student simply doesn't understand what they should spend their time on. This is because academia feels directionless or self-directed and the future of the research is uncertain, however you really should pause repeatedly and consider whether what you are doing is actually relevant to your research, and if it is "extra-curricular", you should ask yourself whether you are actually considering it with that mindset. Too many academics take their "research for fun" far too seriously.
All those interesting corners are my free-time reading, because I enjoy it.
My personal example is when I was doing a theoretical project that had a very intense pure mathematical underpinning. I thought I needed to understand it, but my supervisor kept saying "just work through examples that you know the solutions to". I could have spent 100s or even 1000s of hours learning all this pure mathematics but actually if I would have done so it would not really have helped me today. Instead I just needed to do a few simple examples, which took maybe 20 to 40 hours? And then I was basically done with the research and could write the paper. Maybe if you ask some they will say "you didn't really do it, you didn't know the theory, you were just pretending!" but in the end I got the work done that I wanted to get done, and then went and had a beer.
Come on, maybe not "exclusive" in an absolute sense, but the level of competition isn't even remotely similar. As an academic, I'm directly competing for something literally dozens of times a year (in each grant application, promotion application, evaluation application, salary complement application, grant/project request, paper submission). Each of these has an acceptance rate, we are competing directly among other academics. And each of these directly affects our career. Even when we are tenured professors we can't get away from it, because we have postdocs and students whose own careers depend directly on us and we can't let them down.
My friends in industry don't have an even remotely similar experience, of course competition exists in industry but it's not a constant pressure that you feel every day from several different sources.
The post makes an effort to argue that competition is just one factor among many, but I think it explains like 90% of the overwork phenomenon.
But stability and money on average is still better in academia than being a musician.
That's just not true.
The average person in academia is a grad student. Their stability is low, they generally depend on one person who can easily take advantage of this. And unfortunately this often happens. And their pay is extremely low, below a living wage.
The next most common position are itinerant teachers of various kinds, adjuncts, course support staff, etc. These positions are also paid very poorly and have no stability. Also if you're fired, for any reason, even if your department is abolished (which happens!), that's it. Now you need to move maybe across the country. So not only is your pay abysmal, at any moment you could lose your family, friends, etc.
Even if you are tenured. Yeah, you have a job for life nominally. But, what if one of your fellow tenured professors turns out to be a verbally or even sexually abusive creep? Well, your options for moving are slim, and again, it means losing all of your connections and local life, for some unknown location at a roll of the dice. If you're lucky. Also, at the average university, in say, the average humanities department, a tenured professor after 20 years makes less than an entry-level CS grad at Google.
No. Academia is neither stable nor well paid, with only some exceptions.
A PhD has 4-6 years of guaranteed employment. Most entry-level musicians would love to have 4-6 years of guaranteed income.
I completed my PhD in the UK, and I got funding that just about covered living expenses, and this is normally only for 3 years (13.500 pounds a year). The 4th and last year, normally reserved for write up, you normally get nothing (I managed to get 1/2 year extra funding). Even worse, any month longer than the 4 years and you start paying fines to the university if you still want to finish.
Also, my status was that of a student, not employment. No pension or other benefits.
I mean they're often very high acceptance rates though. The funding agency we use has a 90% acceptance rate. Basically you have to have some idea of what you are doing and they will accept you. It isn't as extreme as you are making out. There are many opportunities.
The main problem is that the opportunities are geographically constrained. I don't think there's a huge amount of competition for a given role, but the fact that it often requires uprooting your life (I moved from UK to Germany) is really bad
What? Wow. Please share?
Not at all. You definitely aren't in the US.
Success rates are low! The stated funding rates by the NSF are around 25%. Depending on the program, rates for NIH tend to be a lot lower. R01 success rates are below 20%. Keep in mind, you're competing with the best and most recognizable names every time. This skews success rates, as a new or average researcher (I mean average as in, the mean person, not as in "mediocre") your rates will be lower.
And even if you compete for a 50% acceptance rate, which is relatively "easy", you are competing with another academics and the majority of them work on weekends.
I bought in on the romantic dream of “monastic” pursuit of wisdom, where my whole being was dedicated to the search for tryth. That fits well with working most of the time. Then, next to me sat the career professional, she wanted good grades and as little impact on her personal life from the “research experience” as she knew McKinsey was waiting. The came family and the monastic dreams turned into work :)
In the corona situation, I switched to breaking almost all deadlines - what is not possible is not possible. No weekend work. I work from 7am to 2 PM. If that means loosing my job, then that is it. But the outcome was different: I did not loose my job - others quit, because they could not stand the pressure, but I kept healthy and produced stable, ongoing work. Slower than everyone else, but higher quality (this is subjective, of course). I reduced my contract to 75% time (30hrs), which also helped. My compromise: No security for long term employment and no money saved. We do everything by bike, no car, no travels outside the country, no restaurants etc. - really everything cut down to the bare minimum. This is fine and it works for us/me, but a lot of PhDs in our department have higher expectations and I think they are better to pursue their career outside science. Look at how scientists worked a century ago: Their life was often equally precarious, but still their motivations helped them sustain constant progress.
[1]: https://www.alexanderdunkel.com/cv_alexanderdunkel.pdf
This individual chose his field of study, university, department, collaborators etc etc based on time leverage. Everything is done remotely and there are practically no meetings. Every paper written produces material for another three. Every sentence is recycled from publication to publication if possible. Automation (python scripts) performs most of the analytical work. Conferences are attended exclusively to find more collaborators to feed into the system. Tier 1 journals are pursued relentlessly. Extra-research commitments are brutally minimised. Most importantly his data collection methodologies are hundreds of times more time-effective than his peers. We're at the beach together most summer afternoons. Having read some of his work, it's much better science than most as well.
The fact is most people aren't tactical about their careers. They don't want to hear what it takes to be successful in the system and would rather treat their PhD as a hoity-toity vocation than admit they have more in common with a blue-collar miner than with Kant or Mendel. If you want to cosplay as Darwin then you need to get on a boat for 5 years on your father's money and cross your fingers you have a brilliant revelation. But if you see your education as a job and apply a hacker mentality academia is easy ground to till.
Will the next Pasteur or Semmelweis work this way? Probably not. But let's not pretend 99.9999% of PhD's are on that route anyway. And no big deal. Science would do better optimising for an army of effective workers than the odd genius. Those tend to make themselves known anyway.
When you don't care to have your name plastered on numerous publications or on some company report, saying "no" to peer/company pressure is much easier.
Bonus point: Propaganda. Movies and tv shows are full of hard working heroes with impossible schedules and deadlines (literal most of the time) which makes for interesting entertainment but terrible lives.
* Paper deadlines. If everyone planned well there would be no need to work on weekends, but nobody plans well, so a lot of the work is done last minute. Also research is international and as a result many conference paper deadlines are set randomly so as not to always be convenient for USA and inconvenient for everyone else. Hence some deadlines are on weekends. Again, if the work was well planned there would be no need even then to work on a weekend, you could submit the paper before the weekend, but no one plans well.
* Students time schedule. Students have classes. They sleep late and work nights. They go to parties and drink. They have exams and assignments that take all their attention during the week and then they get back to the research on weekends. If you are collaborating you often have to be connected in real time, even if it is remote work. This means working nights, weekends, really any time of the day. It can be like being on call.
* Profs and advisors try to get you to cram as much as possible into the paper to make it better. This means working up to the absolute last minute continuing to make minor revisions. I've seen changes made to papers in the last seconds before a deadline, and changes continue to be made for the hours grace period some conferences offer after the deadline. As long as the paper submission portal is open changes continue to be made. It's a stupid way to work, but it is common.
That said, since I was an employee I had more freedom to restrict my hours. Grad students on the other hand are slaves and are expected to work 80+ hour weeks. They are essentially almost always working. So they will try to access other people any time they encounter an issue. It's difficult to not interface with them during "off" hours because you know it might mean they can't get anything done until they talk with you, which can mean wasting days of time for them. They'll mention this at meetings, which makes it look like YOU are blocking the project. And you are, but for reasonable reasons. But that doesn't look reasonable to the student. And profs used to be students and in many cases that's basically the only life they know so they tend to sympathize with the students and not with you.
Some things that would help:
Plan for papers, posters, meetings so it's not all last minute. The conferences happen around the same time every year. It's not like it's a big surprise when the date is set.
Get grad students to form unions so they are not so overworked.
Set rules on what profs can task grad students with. Getting coffee, creating lessons for the profs classes, getting the profs laundry, being volunteered for work, being assigned more than 40hrs work/week, should not be on the list. This is a large part of why grad students are overworked.
Require management courses for profs so they learn that people are not machines. There is a reason we have a 40hr work week, and while going beyond 40hrs is _sometimes_ required, working longer hours continuously just damages productivity and is actually making the work go slower and burning people out.