Hacker News new | past | comments | ask | show | jobs | submit login
Lessons from my PhD (utk.edu)
403 points by andrewnc 4 months ago | hide | past | favorite | 130 comments



My big eye-openers (some from postdoc) were more about the sociology of science than the day-to-day productivity:

- Even the most blatantly wrong and illogical published work can only be displaced by another publication that explains/does the same phenomenon better; i.e., people are going to keep believing in phlogiston until someone shows them oxygen. If you simply point out inconsistencies in phlogiston theory, in person or in writing, they may well make a variety of unwanted psychological deductions about you.

- Similarly, nobody actually enjoys being around critics or enduring criticism, and therefore you will observe many senior scientists partially avoiding the major downsides of being a critic by artfully concealing criticisms inside what sounds to the uninitiated like mutual affirmation sessions. You have to listen very closely and learn the lingo to pick this up.

- Never question a scientific superior (other than maybe a direct mentor or very close colleague) with any other approach besides "I have a helpful suggestion about how you can maybe reach your intended destination better/faster/more precisely". Regardless of where that destination might be, such as off a cliff or into a wall.

- The opinion/fact ratio you are allowed to have as a scientist is directly and very strongly correlated with seniority, H-index, and so on.

- The incentive structure of scientific publication is such that there are big rewards for being right on an important question, bigger the earlier you are to the party, and little to no penalties for being wrong, so long as the error cannot be provably and directly linked to fraud. There are a variety of interesting consequences to this incentive structure.


In addition to ringing true, this seems largely in line with Thomas Kuhn's thesis in his Structure of Scientific Revolutions [0], a book which despite its shortcomings, should be required reading for anyone in a STEM field.

[0] https://en.wikipedia.org/wiki/The_Structure_of_Scientific_Re...


Kuhn's thesis doesn't have a lot space for the sociology of science to have this kind of influence. Of the classical theses of scientific progress, this comes closer to Lakatos' thesis of research programs.[0][1]

[0] https://en.wikipedia.org/wiki/Research_program

[1] Lakatos, Imre. (1978) The Methodology of Scientific Research Programmes: Philosophical Papers (J. Worrall & G. Currie, Eds.). Cambridge University Press.


> The incentive structure of scientific publication is such that there are big rewards for being right on an important question, bigger the earlier you are to the party, and little to no penalties for being wrong, so long as the error cannot be provably and directly linked to fraud.

This is fantastic insight and I'd like to thank you for sharing it with our group.

Would you agree that the model of rewards for correctness and penalization only in the case of fraud is the core feature of science? And what separates it from business or politics where being an honest failure is worse than being dishonest but successful?

Again, this is a great post, and I think you have a fantastic future in the sociology of science!


It's a good question, and I don't like posting excessively long comments and didn't have time to make it concise, so here's an attempt at an answer:

https://pastebin.com/fsrTtiKY

I think science is too big a thing to have a small set of "core features", and the question of how to usefully define "honesty" in a scientific context is another big topic, but reading about "bullshit" (the term of art that has its own literature, not the colloquialism) is a good place to start thinking about it.

I would suggest that fraud is one of the rarest types of dishonesty, because people who are both smart and dishonest have less risky ways to proceed, and that such people are very glad fraud exists, because it misdirects attention away from their arguably more damaging and prevalent methods. Feynman has a passage about how honesty in science is more a state of mind, which I agree with. But really, the techniques to be dishonest with low risk are the same in science, journalism, politics, and business.

My field isn't sociology of science though; these are just views from the genomics trenches.


Are you trying to provide an example for many of the points of OP above or am I overreading this?

In your message I observe, very careful criticism, uncalled praise, admission and defense of a system that excludes most criticism...


I'm joking a bit with the style and my limited experience leads me to agree with the OP.

The middle paragraph includes my sincere response that a system where discovery is rewarded, failure forgiven, and dishonesty punished is ideally suited to the mission of science.

So I was left wondering if the OP would expand on what their thoughts were about the interesting consequences.


That seems to ring true, but - it also seems a bit defeatist, don't you think?

"Never question a scientific superior?" Not parsing that concept, please elaborate.


Very early on, I noticed that graduate students tend to be idealistic, postdocs extremely cynical, and faculty ruthlessly pragmatic perhaps to the point of occasional shortsightedness. Clearly, something about this progression is expected and normal. I'm a postdoc now, so I'm right on schedule.

I think the way it ultimately works is that you have to be disillusioned from the grade-school fairy tales told to the public about how science works before you can learn to live and work in the environment that actually exists rather than the one you wish existed.

> "Never question a scientific superior?" Not parsing that concept, please elaborate.

tech < grad student < postdoc < junior faculty < full prof < Big Guy/Gal < Nobel Laureate < NIH Director

People above you in that chain will accept limited feedback on methods to attain their chosen goals and will greatly resent questions about whether their selected goals are worthwhile/realistic/rational, or whether their gestalt vision of the field's conventional wisdom is correct.


"People above you in that chain will accept limited feedback on methods to attain their chosen goals and will greatly resent questions about whether their selected goals are worthwhile/realistic/rational, or whether their gestalt vision of the field's conventional wisdom is correct."

Corporate management has the exact same situation.


Yeah but it's a lot easier to not care.


Yeah, in corporate world, at least you're paid to not care, and can change jobs easily. In academia, you're paid shit, and changing labs is not nearly as easy.


I agree with most of the GP's points and I don't think of them as defeatist, but rather a call for realism when dealing with people (versus data, which have no ego to bruise). It's very hard to devise a system that rewards individual achievement without ever falling prey to classic human flaws. The good news is that science over time tends to be self-correcting, and all that requires is a commitment to shared principles and methods, combined with enough anarchy that no one individual can screw up an entire field (Trofim Lysenko being the most extreme example, but any bureaucracy can accomplish this).


The presentation is cynical, as xab31 themself attests, but I don't think it's defeatist.

1. Bad work only being displaced by good work: everything works like this. To replace some useless commercial product (take your pick) someone has to come up with something better. Same goes for information.

2. Nobody liking criticism can be rephrased as it being important to attack ideas, not people, when you have to work with those people.

3. "Never question a scientific superior" is the first piece of advise I think is too cynical. As a warning against undermining a colleague in public when you need their support, I agree, and that's kind of a restatement of #1 and #2. But science really does have a culture of publicly debating contentious ideas. You can definitely be more critical in an event specifically held as a debate / open forum than in a presentation Q&A though, and at a social event it's polite to be at least vaguely supportive.

Kind of a tangent to the later points: Day to day scientific research is mostly chasing dead ends and other activity that is (in hindsight) mostly useless, but there is genuine societal value in having a large body of skilled workers available. That is, science spends a lot of time spinning its wheels trying to figure out the right question to ask, and once this becomes clear there is rapid progress. This means the papers published in between the breakthrough periods aren't really worth paying attention to unless you work in that area. Having a lot of scientists and engineers in the workforce so we collectively have a decent chance at obtaining and exploiting next breakthrough is the point, the papers are just a byproduct.


>"Never question a scientific superior?" Not parsing that concept, please elaborate.

If you think you've been put on a bum topic or your supervisor has put you on the scientific equivalent of a PIP with no way up or out your room for maneuvering is limited, to put it politely.


I understand the impulse to not want to be defeatist, but sometimes it’s both easier and more productive to stop running into the same walls over and over and instead find the path around them.


It is what it is, and it is mostly pretty successful. More or less.


Well, as defeatist as going to work fo a FAANG and not expecting that your managers will give a toss about fairness, their users privacy, or the spirit of regulations. Life is like this. Right now in the African Savannah a lion is mauling a gazelle, it happens daily.


Beautifully put.

Your observations aptly apply to industry as well except for your final one regarding the incentive structure.

Thank you for commenting.


This is true here, as well. I once asked about something related to SSL/TLS (fairly politely) and was kind of mockingly escorted by some groupies to the corner since I apparently responded to an Apache developer.

I was just trying to learn. Learning bad is what I learned.


The sociology of science is so interesting. (Not the field, but the subject.) Here are some of my favorites quotes/thoughts:

This one is a direct contrast to your advice (which speaks volumes about what's wrong with academia): "A good scientist, in other words, does not merely ignore conventional wisdom, but makes a special effort to break it. Scientists go looking for trouble."[0]

This was written about physics at Caltech, but applies more broadly. It explains why the ability to 'manage up' is so critical for early-career success. "[...] departments are run, for better and worse, by the professors who often lack managerial experience. Worse, they are generally unaware of this shortcoming, assuming incorrectly that management is trivially easy compared to their topics of study and merits minimal effort. We have now seen the consequences of this lack of attention." [1]

Academic politics is a great reason not to stay in academia: "Look for environments where competitors see themselves as playing a game, rather than fighting for survival — this prevents rankings within the hierarchy from becoming an existential problem." [2]

This book has a great chapter of career advice, here's a gem: "Don't build a pyramid. Everyone seems to build one pyramid per career. A pyramid is an ambitious system that one person really cares about and that winds up working well, but then just sits in the desert because nobody else cares the same way. This happens usually just after leaving graduate school." [3]

"In general, status-conscious places are miserable for everyone, and the more, the worse." [3, next page]

Gatekeeping is predictable from the incentive structure: "For all the high-level talk about how we need to plug the leaks in our STEM education pipeline, not only are we not plugging the holes, we're proud of how fast the pipeline is leaking." [4]

"So why am I not an academic? There are many factors, and starting Tarsnap is certainly one; but most of them can be summarized as 'academia is a lousy place to do novel research'." [5]

"...whereas Newton could say, 'If I have seen a little farther than others, it is because I have stood on the shoulders of giants,' I am forced to say, 'Today we stand on each other's feet.'" [6]

[0] http://www.paulgraham.com/say.html

[1] https://caseyhandmer.wordpress.com/2019/08/09/caltech-astrop...

[2] https://www.briantimar.com/notes/mimetic/mimetic/

[3] Phillip Hobbs, "Building Electrooptical systems: making it all work" 2nd Ed, p392

[4] https://danluu.com/teach-debugging/

[5] http://www.daemonology.net/blog/2020-09-20-On-the-use-of-a-l...

[6] Richard Hamming 1968 Turing Award lecture, Journal of the ACM 16 (1), January 1969, p. 3–12


> "...Scientists go looking for trouble."

It is several repeated and very costly attempts that I made to do just that which leads me to give the advice I did.

The pyramid quote is an interesting one. Obviously there is a tension between being passionate about an idea/goal/cause but not being overly siloed. It seems the best-case scenario is: pick your passion, find some people who're thinking in the same general direction, and compromise the vision among yourselves.

Let's just say that the thought of solving some of the problems I'm interested in from outside academia has occurred to me. But I'm sure it's not all sunshine and rainbows on the outside, either, and moving from academia whose primary motivator is risk aversion to something like a startup is an extreme culture shock, the more so because my objective would be building something real, rather than bilking gullible VCs into an acquihire.

Really good thoughts there.


Thanks! What kind of problems outside academia are you interested in?


Well, I do aging research (mostly from a computational+biochemical perspective). I've met most/all of the important players in the field, and it baffles me how this important area of research continues to be a backwater, as far as the public's concerned.

It's hard for me personally to think of something more important than aging, so if I were to expand outwards, it would be to pursue the same goal, but maybe with fewer constraints. In general, I'd work towards streamlining and automating certain aspects of it. Technologically, the field is in the Dark Ages. There are realistically ~200-300 (max: 5000 including subordinates and techs) people in the entire world working on this seriously, which is fairly mind-boggling, considering that it is the primary risk factor for cardiovascular disease, cancer, and indeed COVID-19, along with many other diseases and the more transhumanist and futurist implications.


This reflects my experience as well.


Young people need to realize that the things we love about science: the uncompromising search for the truth, its international and no-boundaries character, the ability to bow down to evidence, the ambition of the ideas, are just a very distilled fraction (basically the highlights) of a what it is a very mundane, fragile, political human activity, full of petty and lame characters, absurd situations and pathetic developments.


> full of petty and lame characters, absurd situations and pathetic developments.

Yeh, I wish I could understand this a few years ago. Pursuit of truth is not the primary goal of many tenure profs it seems.


Well there's science and then there are social institutions that undertake science. But science can exist outside the institutions.


Missing from this actionable survey is the skill of answering questions that haven't yet been answered in the secondary literature (books, reviews) or even primary literature (journals). This is, of course, what research and a PhD is all about in the end.

In my experience, this fact was the #1 reason why some peers dropped out of PhD programs. They joined expecting a continuation of undergraduate education, which consists largely of recapitulating what appears in the secondary literature. Instead, what they got in graduate school was the expectation that they would be producing the primary literature. That's a very different game.

It was a game these peers discovered they hated playing. Nothing in college can prepare you for the isolation of spending your time becoming the world's expert on a narrow technical topic. Your usual reinforcement mechanisms of approval from family and friends gives way to slight comprehension at best. Then there is all of the alone time doing research requires. But I suspect the hardest part of all is the seemingly endless lineup of dead ends and false hope. Not only is success not assured, you often have no idea whether the result will have any utility even if you succeed.

Then, just when you've gotten the hang of this finding answers game you discover that the real expectation is to be the one who formulates good questions. The kinds of questions that, although they will certainly involve dead ends, will ultimately pay off in some meaningful way. Very little in a bachelor's prepares you for doing this. It's a hard-won skill that comes from a round or two (or three or four) of months (or years) spent answering questions that nobody cares about. A lot hinges on your relationship with your advisor on this one.

The PhD isn't just a bachelors degree but harder. It's a completely different animal. The skills in this article are very useful toward that end. But there's a lot more to the story when it comes to skills for finding answers to those unanswered questions, and formulating worthwhile questions without answers.

The benefit of all of this work and discomfort is that you come away with the ability to answer worthwhile questions that haven't yet been answered. And that's a highly transferrable and applicable skill.


> In my experience, this fact was the #1 reason why some peers dropped out of PhD programs. They joined expecting a continuation of undergraduate education, which consists largely of recapitulating what appears in the secondary literature. Instead, what they got in graduate school was the expectation that they would be producing the primary literature. That's a very different game.

Hit the nail on the head. I would like to add one point though - it's not just the unanswered questions, one sometimes doesn't even know which questions are unanswered.

Typically, up until a Ph.D - you are given a question and then asked for an answer which more often than not exists. Suddenly, in a Ph.D - not only do you not know the answer, you don't know the question too. The craft to come up with an important question, create a well-defined scope and then answer the question from different perspectives is the heart of a Ph.D program. The true skill is the ability to "learn to learn". The transferrable skill is to probe around for questions which are important, define them and then go ahead to answer them.


From my experience failing my PhD it's not so simple. The head of my lab have a bunch of topics he want to (make other people under him) investigate, sometimes really precise. One of the PhD student literally got his thesis question handed down after an experimented researcher worked on it for 6 months. Other like me had to found one themselves. It's obvious that the first student got a head start of about a year. The irony is he is now in difficulty writing his thesis, despite having published the required number of papers, which is not too surprising since he didn't get the problematic by himself.


I recall the same things. I remember being bitterly envious of students that seemingly were given topics to examine while I had to spin my wheels for a good year or two coming up with my own ideas to explore.

Looking back my perspective is very different. First of all, my ideas generated a minor spike in publications for the lab all centered around my work. After I graduated I continued to advise new students to continue what I started.

I now think in these teems, which might sound cynical but simply reflect the vicious nature of academia.

    1. The students given topics, the ones I was jealous of, all kind of sucked. They were given topics to advance a short term goal and gtfo. It may seem paradoxical or cruel but a savvy advisor will maximize both short term and long term gains. I was a long term bet, the others were short term plays.

    2. I benefitted greatly from being forced to identify my own topics. This is the one skill that I use every single day. Every hour of every day. As time goes on the ability to evaluate ideas deeply and with some speed effectively defines what it is I do for a living. 

    3. Students given topics were cheated out of more valuable long term skills for the lab’s short term gains. This is not always universally true, of course. Some super stars really can crank through a deep serious topic quickly and continue on to generate novel research if their own. One such person may appear in a university department once every ten or twenty years, they are extraordinarily rare.

    4. I was well aware of the exploitive nature of grad school, did it anyway with a clear head for what I wanted to get out of it and my only disappointments came from when I giddily let my guard down and expected more than what I already realized would be forthcoming. A specific example, my advisor would use my conference paper acceptances to fund their own personal travel and vacations; I was not allowed to present my own first author papers. Silently tolerating that sort of bullshit, in part, allowed me to graduate.


"A specific example, my advisor would use my conference paper acceptances to fund their own personal travel and vacations; I was not allowed to present my own first author papers."

That's really very bad. Learning to present to your peers is an important part of the process, as is getting your name and face out in the field.


Wow, not being allowed to present your own first author paper is pretty bad - unless someone else contributed a lot of the work and was made second author + presented?


Not at all. The papers where I was first author were essentially 100% all my work. Second or third authors were advisor and a committee member that helped edit, closely read proofs and pseudocode. Certainly helpful and deserving of authorship but a steep drop off in contributions from first author (me) to them.

Advisor had family in Europe (I am in US), and would use these conferences to create extended family vacation plans.

Of course when I think back on this I am still pissed! But years have passed and I am doing well in my career. I am able to not get too fixated on it.

And yes, I agree with the people saying it was terrible and very wrong. In terms of all the possible terrible and very wrong things that can happen to a graduate student this is maybe, in the grand scheme of things, about mid-range maybe?


Wow, those who could find their own research topic were lucky. I've never seen anyone in my environment get that much freedom. The supervisor sets the problem and the student must solve it.

Edit: I was under the impression that even the postdocs are hired for a specific task.


> Suddenly, in a Ph.D - not only do you not know the answer, you don't know the question too.

Mirroring what tasogare said: There are a lot of research professors who will not give you the flexibility of finding the question. They often are paying you to be an RA, and will want you to work on their topics, not yours.

This may vary per discipline. In the circles I was in, this was the norm, though. Some professors were open to you choosing your own topic, but the "contract" was similar: If they are funding your research, then you should work on your own topic "on your own time".


I know what you and tasogare refer to and I've seen it happen. I had to fight to change my thesis topic 3 times and also change advisors. Of course, it is not simple.

I wouldn't change anything from my initial comment though. Flexibility is not binary - it's a gray scale. If you have ZERO flexibility, you should accept the implication that such a Ph.D will be stripped of some valuable lessons. On the other hand, you can always decide to not do it and move to a different professor. You could also decide if the broad area is ok with you before you take an admission to a lab.


I'd frame it this way: a good senior researcher knows when to give a junior researcher the question or not. First-year grad student? Pair them off with a postdoc for fast iterating on an established project. Fourth-year grad student? Push them by letting them flounder a bit and learn to find their own questions.

I agree that in practice, this simply varies a lot by discipline and advisor.


You missed the most fun option: pair them up with a problem that leads to nowhere!


This is all completely true.

"Then, just when you've gotten the hang of this finding answers game you discover that the real expectation is to be the one who formulates good questions."

And that is where I personally failed.

On the other hand, there's that funny moment...

One of the things I've heard repeatedly from pilots is that first solo flight changes everything. Before that, you're just some human. Afterwards, you are some human who can fly. Everything is somehow different, although I've never seen anyone really successfully describe how. I suspect it's different for everyone. But then I'm not a pilot.

In your dissertation defense, someone whose knowledge and intelligence you respect immensely will ask a difficult question. When you answer that question confidently and to their satisfaction, the world is a different place. For one thing, you're no longer student and teacher; you are peers. But that's not all it is.


Spot on! I had the same misconception, but it worked out OK b/c I was motivated at least in part by curiosity. When you're curious, you ask enough questions to get to the edge of knowledge and then pose a novel question. If you enjoy the coursework purely b/c you like having nice tidy answers to everything, being at the edge is uncomfortable & research isn't for you. OTOH, in my PhD coursework, the HW questions were almost always solved ones where we just had to reproduce the steps to get the answer that was included in the question formulation; this burned much of my curiosity out by the time I was done.


Those dead ends are negative results. While not easily publishable they can form the bulk of a thesis. Many people get demotivated because they treat a thesis like a journal publication. I, for one, was glad I finally didn't need to sex up the language to convince some editor.


> need to sex up the language to convince some editor.

I hated this pressure. I wrote up the core of my thesis as a manuscript for a second-tier journal, but my advisor though I had a shot at a first-tier publication. I disagreed, but I rewrote the paper anyway, and had to significantly rework/descope it. It ultimately wasn't accepted for the first-tier journal, so I rewrote it a second time for the original journal. The whole process was immensely frustrating (cat-herding coauthors, playing volleyball with editors/referees, trying to discern whether my concerns about overselling my results were legitimate issues of integrity vs. instances of imposter syndrome, ...).

I fell in love with the hard sciences because "reality must take precedence over public relations, for Nature cannot be fooled." [Richard Feynman] Finding out how much PR is actually involved was hugely disillusioning.


> The benefit of all of this work and discomfort is that you come away with the ability to answer worthwhile questions that haven't yet been answered. And that's a highly transferrable and applicable skill.

This is why I have come to see that PhDs can in some cases make excellent founders. Source: CS PhD turned founder ;-)


And this makes everything else in the survey easier. Topic sentences and presentation skills are useful but most important is having something original and substantial to say. The rest follows and is easy by comparison.


Two things I would add to this that I learned early on in my PhD:

1. Presentations aren’t really about conveying information.

I sat though so many dull presentations, they were very informative but I can read a paper quicker than they can badly present the same information.

The best presentations were the ones that covered the whys of the work, the applications, the next steps, the specific problem areas - often these aren’t covered in the paper but, armed with that extra insight I am far more likely to read the paper and remember it.

Presentations are (as the author says) about telling stories.

2. Show up. So many PhDs waft around not doing a whole lot, and so land up being on the program forever. This only benefits the uni and is detrimental to the student. I noticed in the first month of my PhD that most people did a lot more work at the end than the beginning - so I flipped it, worked consistently from day one and got done in just under 3 years.

Carry this over to your daily life and it’s almost a super power for getting stuff done. Consistently showing up and plugging away in something reaps rewards.


> So many PhDs waft around not doing a whole lot, and so land up being on the program forever. This only benefits the uni and is detrimental to the student. I noticed in the first month of my PhD that most people did a lot more work at the end than the beginning - so I flipped it, worked consistently from day one and got done in just under 3 years.

Amen. PhD is a marathon. Other degrees may be a 100m or 400m race but PhD is about consistency.


100% agree with number 2. A PhD is a job. 5 days a week, 9 to 5, or you’ll never finish in 3 years. Fastest I saw was 2 years, which was a guy that put in all the hours. Slowest was a guy that took 8 years, ‘full time’, though he was never there, but clearly didn’t have a job. Goodness knows how they supported themselves for all that time. In the uk you had a grant for 3 years when I did mine.


Yep, this was my attitude - I was technically done in 2.5 years but my uni wouldn’t let you submit until 2yr 10 months - so I started my first company in those intervening months.

To anyone reading this who is considering a PhD, start writing up your thesis as soon as you can, like 6 months in if you can and have enough to start. You can always go back and change when you’ve written but it makes life so much easier if you’re “always writing up” then you’re not terrified of starting.

Oh and yeah: 9 to 5, full time, give yourself a standard holiday allowance and stick to it.


Thanks for that perspective. I just finished my applications last month and am anxiously waiting on decisions. I always thought 4 years was the absolute minimum.


It depends on the discipline, the programme, and (particularly) the country. I did mine in the UK, where the funding is for 3 years and the expectation is ≤4. I managed that (modulo two terms of sick and paternity leave) and so do most others in my field. The quickest I saw was 2 years 6 months, by someone with an impossible combination of intelligence and relentless 9-hour productive days. Some take longer (financially and professionally problematic), not that many drop out.

In some places in the US taking >8 years is normal. In some parts of Europe it's an actual job, with delineated teaching responsibilities, a pension scheme and everything. In Russia the equivalent isn't even called a PhD. It's not a standardised process.


I guess that makes sense - in the EU a masters degree is usually required for a PhD which is not the case in the US. That accounts for 1-2 years at least.


Yes. The standard "1+3" funding programme covers a 1 year MRes or other masters degree with methods training in advance of the PhD programme starting. I got mine separately, so just had "+3" funding. Like in the US, a PhD without funding is normally a bad idea.

Edit: It generally remains shorter than a US programme, though. We tell ourselves that our focused BA/BSc programmes provide a better foundation than the broader US undergraduate degrees, but I suspect the truth is just that it's cultural differences.


I was very lucky to not need a masters, 3 years BEng and then 3 PhD. I couldn’t have done another year. I was dying to get out but the end. My wife did 3,1,3 and I have no idea how she did that extra year.


>In Russia the equivalent isn't even called a PhD.

Are you thinking of Candidat Nauk or habilitation?


Candidat Nauk. As I understand it the Russian doctorate/habilitation is closer to our 'higher doctorates' (DSc/DLitt/DM etc.) which are rarely awarded and are mid/late career distinctions. There's no requirement for habilitation here, so PhD is almost always a final degree.


While I agree with no. 2, I think part of the reason that makes me not do it is realizing that I will have to do it all my life if I become a faculty. I have seen my friends graduated from Ph.D. and they literally told me that their life is basically the same, except that they now have service tasks to do on top of research. To think that I will always have to plug away and not have enough time for family or relationships makes me a bit demotivated.


Research, teaching, and service.

As a faculty member, each of your three constituencies is almost completely invisible to the others. So each one thinks you work hardly at all. Only your family sees the total hours, and only your tenure and promotion committee sees the total contributions (and typically they up-weight research, so don't skimp there).


“Showing up” is the best advice of all, better than any of the (great) ones presented by the article. Actually, none of the other advices will work unless you show up. I’ve seen people digging around for advice in the hopes that it will save them from disaster or help them do more with less time, but the truth is that the advice only works if you are actually willing to suffer through the working hours.


I also did a PhD and this is all true. I realized many of these things years later.

I think many of these points come down to confidence. When you are in the trenches, you really, really do know a lot, and you know it in incredible detail. In fact, in your career, if you leave academia you will probably never know a unique small "thing" in such detail ever again simply because you will have to make something as opposed to studying it. Not even your professor knows everything about what you do, and so she may give advice that seems to contradict what you think. It is vital that you trust yourself enough to speak up. Yes, the professor is really smart, and knows more than you, but she didn't spend 3 weeks in the lab wrestling with some optical setup like you did and you know some things better then her, better than anyone in fact. It's hard to admit, I know.

Also, you may really have wrong assumptions about the progress you're going to make in the project. You may feel very bad after a year of messing around while the prof thinks you're doing well. Talk about these feelings. The prof knows what's normal, you on the other hand may think you're the next Einstein (and assume Einstein wrote something great every other month) and constantly disappoint yourself.


Similar to learning software engineering on the job. Once you're leaving the baby level you stop being able to take the more senior engineers' word as golden. You will know some things better temporarily due to recent intense exposure. The tricky part is figuring out when that's true and when others can see something you can't. This never leaves you I guess.


> It is vital that you trust yourself enough to speak up. Yes, the professor is really smart, and knows more than you, but she didn't spend 3 weeks in the lab wrestling with some optical setup like you did and you know some things better then her, better than anyone in fact. It's hard to admit, I know.

It's really not. It was obvious to me about 9 months in that my advisor really didn't know all that much. The professors who really seemed to have technical chops were either new faculty still trying to get tenure, or the rare iconoclast who didn't play the game and had a single grad student. The tenured professors with large research labs were frankly better politicians than they were scientists.


>It is vital that you trust yourself enough to speak up. Yes, the professor is really smart, and knows more than you, but she didn't spend 3 weeks in the lab wrestling with some optical setup like you did and you know some things better then her, better than anyone in fact.

Amen to that! It's better to have the discussion than to silently disagree (well, assuming your thesis advisor isn't a raging narcissist, and assuming you are sufficiently tactful about speaking up) because there's a chance you are mistaken & the feedback would be helpful.

>You may feel very bad after a year of messing around while the prof thinks you're doing well. Talk about these feelings.

Another one that I wish I had known (again, needs caveats about unhealthy advisors, though). It's easy to underestimate the scale of a task as a grad student (the devil is in the details), and to therefore bite off more than you can chew & feel guilty for choking.


I agree, I was fortunate to have a very nice prof, really dedicated to the development of his PhD students, who saw the importance of social events and tried to have some fun himself, eager to roll up his sleeves and help in the lab, he enjoyed it. He was a bit further in his career with no need to publish or perish anymore.

That's also an advice I give to aspiring PhD students, look for a warm place, talk to the other PhD students about the working atmosphere. You don't want to end up a "measurement slave", as one of the 4 PhDs that (and I quote a prof during a talk) "was burned on this subject".


There is another, more fundamental lesson, that I learned during my (failed) PhD - make sure that the environment suits you. By this, I mean do some research beforehand about the supervisor and the alumni. If possible talk to one of the other PhD candidates in their department and find out if you are compatible with the working environment.

This could be hard to do such early in your life, as one does not have much experience. Usually it falls in one of two categories - either you are someone that can do the work but needs support and guidance, or you prefer working on your own, in which case a more hands-off supervisor would be OK.

If you are of the former type and find yourself working for a supervisor that doesn't offer much support, it will be very hard to finish anything, and most likely you will become demotivated and drop out. Likewise, if you want to try things on your own but your supervisor wants to dictate where to go next, there will be a lot of conflicts and even the possibility that they block the thesis until it is done their way.

Having other PhD colleagues around and bouncing ideas off of them is worth its weight in gold, make sure that there is at least one that is working on something similar as you are.


Yeah, this is great advice for undergraduates considering graduate applications. I lucked out with my supervisor in terms of his advising style (rather hands-off, which suited me), but I could have easily gotten stuck with a bad match. I picked the school, rather than the advisor, which I now realize wasn't the best way to go. [Edit: I will say that picking a school where there were multiple faculty to choose from in the topical area was a good decision, because that provides options.]


All of these tips are good, but the “get excited” one has been my secret weapon through life.

A professor in undergrad gave me the tip to get excited or even feign interest when reading dense written material in order to retain more.

After trying it throughout a difficult class I was amazed at how well it worked. I applied it to every other academic thing I didn’t want to do and noticed immediately how much easier and enjoyable school was. I still use the “fake excitement” trick for my work all the time.

Also, it’s kind of like a Trojan excitement because after I fake the intense interest I do genuinely become interested more often than not.


It's hard to remain excited when so many professors don't prepare for lectures, teach outdated material, rely on question banks for exams, and play the part of ball-breakers. I went back to school for a different degree program in my 30s with a fresh perspective and a much bigger drive from 15 years ago when I did my first two degrees. Mind you that it was during COVID but what did I get? A bunch of "read these chapters, take these exams" kind of lectures. The current education model is shot, particularly the PhD degree where you practically grind out nonsense for years on end only for your advisor to collect the funds and stick their name on top of your papers.


> stick their name on top of your papers.

Maybe you have a more independent mindset having gone into a PhD program a little later than most, but the whole point of a PhD program (at least in the sciences) is that it’s an apprenticeship. You study under an established researcher using their grant, so it’s not “your” paper. You are supposed to work together using grant money from your advisor.

If you have obtained grant funding on your own and are working independently on a novel research project you thought of yourself, then you can call it your paper. But that scenario usually doesn’t happen, because it’s hard to come by funding without a good proposal, and it’s hard to write or qualify for a grant without the training one gets in a PhD program.

If you are working using grant money, lab equipment, lab space, data, models, software, or methods acquired and developed in your advisor’s lab, then even if you write an entire paper yourself it’s still both your names that go on the paper. I’ve had a few like that and was glad to share the credit, because it wouldn’t have been possible otherwise.


In my field (experimental plasma physics) the general convention is that the PI's name goes last on the paper, while the first author is the one who did the bulk of the work and wrote the manuscript, with the other coauthors usually playing small supporting roles.

This field also tends to have very large grants (~$10-100mn/yr) that support dozens of researchers, because of the large centralized facilities, so it's easier for students to have some self-direction.


I totally disagree with you. Giving a single word "guidance" and pretending to read a manuscript is not enough to qualify for authorship in most journals submission guidelines. Likewise, all the thing about money, lab space etc. (which in my case is funded by a national scholarship, not my supervisor) has nothing to do with research ideas, which is what papers are about. If people who are making my live easier as a grad student were to be giving authorship, the secretary and the cleaning lady would both rank higher than the head the lab. I'm putting the name of my supervisor because I'm forced to and because I belong to his lab but anyone who worked with him knows his involvement in lot of papers is close to 0.

So if someone has done the research and wrote the paper it is normal he got credit as first author for it. Whatever money is lying around isn't writing paper by itself. The monetary compensation is meager enough not to be robbed on top of that of what we created.


There seems to be an increasing expectation that undergraduate education should be like high school. By the time you start studying at university you should be able to study independently. The lecturer isn’t there to entertain you or to hold your hand.


All true. I learned about topic sentences only recently, I wish I'd heard of them years ago!

I'll add something else I have realised:

Your Gantt chart is not for you.

I hate Gantt charts - they're out of date the second they're created; they take too long to update; there's very little decent free software for them that everyone uses; etc etc.

But your supervisor will probably want to see one. Or your funder, or examiners, and so on.

That's the point: sometimes you just gotta transform information into the format that's expected. From your perspective it may be easier to say "I've completed task X but task Y will drag on for another two weeks" than it is to update a spreadsheet and render a Gantt chart, attach it to an email and stick it in a shared drive. But from the supervisor/funder/examiner perspective, they need a way to very rapidly assimilate complex detail and spot problems.

A lot of academia is about clear communication of complex material. Your supervisor probably has several students, as well multiple projects of their own, teaching duties, management duties, and so on. Your Gantt chart is for them, not for you!

Simple and obvious in hindsight, but it really helps me put aside the grinding resentment I feel whenever it comes to updating a Gantt chart :)


I did a PhD in physics and feel like I missed all the great "meta lessons" some people seem to learn in their PhD. Mostly I just spent my time alone doing calculations either with pen and paper or computer. Most of the stuff I did was either suggested by my supervisor or was obvious continuation of some previous work. Even after I got my PhD I didn't feel like I was really a member of the research community or that I had a PhD level command of my field. I just did a bunch of calculations, wrote papers on what I did and got a PhD. It was almost like doing homework on a really long course, but just more difficult.

I left academia after a failed postdoc because I realized I had no clue how to conduct research on my own; I didn't know how to pick good research topics, or how to manage my time, or how to find people to collaborate with, or how to collaborate productively with someone for that matter.

I'm not sure if the fault was my supervisors or mine. I'm a bit "on the spectrum" and have lots of difficulties with social interaction, but I guess so do many other people drawn to technical fields and still they manage to navigate the system somehow. I certainly never sought for any kind of mentorship because I didn't realize it was needed and, also, because it felt extremely awkward.

Also, the whole academic system seemed a bit fucked up. People do research and write papers because they have to produce something measurable, not because the research they do is actually interesting or important. I published five papers during my PhD and I would say that maybe only one of them was slightly interesting or important, and even that could have been much better. All of the papers were published in proper, highly regarded journals (mostly Physical Review). Towards the end of the PhD I started having some vague ideas of stuff that would be _actually_ interesting and more worth my time, but also more difficult and less certain results. When was I supposed to do those? I was still in the mindset that I wanted to stay in academia so I couldn't take any risks.


To be honest, that's kind of normal and 5 papers is pretty good going for a PhD (assuming at least some of them were at reputable venues). The truism is that you should view the PhD as training you how to do research, but not necessarily that the results you produce will be in anyway ground breaking. Of course there are exceptions. As you develop it would be expected you apply for funding/fellowships to pursue more difficult problems etc. and demonstrate more independence.


As the joke goes, once you realize it's all bullshit is the day they go "Congratulations, you finally understand the field, so here is your PhD". Then you just have to decide if you want to continue on and get paid to do bullshit.


The thing is, I think there are people in academia who are not just bullshitting. Occasionally real scientific advances do happen. It's just that if I don't personally have a breakthrough in sight, I'm supposed to just produce garbage and pretend that I'm doing a good job while trying to do the actual good research on my free time or something.


Yes, that is exactly how it works. Well said.


>People do research and write papers because they have to produce something measurable, not because the research they do is actually interesting or important.

Yes, so much of this. I think it's a direct consequence of your next point:

>I wanted to stay in academia so I couldn't take any risks.

That's how boring research gets prioritized.


> That's how boring research gets prioritized.

Yes, I realized I was part of the problem, but couldn't help it (except by leaving). If it was only a bunch of PhD students and postdocs wasting their time, the boring research wouldn't be such a problem. It becomes a problem, however, when everyone is doing it and the actual good publications get drowned in noise.


The "lead or be led" trope is apt, and certainly became the paradigm of my doctoral years. I might, however, amend this slightly. There's another idea of going rogue too soon or too late (https://matt.might.net/articles/ways-to-fail-a-phd/).

Students, new employees, and other inexperienced folks need to be led initially, and then rapidly, transition into a self-directed paradigm. Success emerges if and only if the advisor and student recognize the need for this transition at a similar moment. The alternative is either the student who runs down rabbit holes repeatedly despite being guided elsewhere (those students tend to at least get SOMETHING done and while they take forever to graduate, do find some interesting results along the way) or the student who after a couple years is still just reading papers and waiting to be told what to do (these students often fail outright as advisors get fed up with the hand-holding).


Great point. I clashed a bit with my advisor early on - they were trying to shepherd me away from a rabbithole, but I was a bit pigheaded and determined to fix something that I felt was sloppy. I ended up with a publication on it, but it probably wasn't a great use of my time (aside from providing me with a chance to flex my wings).


It seems like, in your case, the publication is evidence that the rabbit hole was at least original and interesting. Students in that category tend to graduate, have a few compelling results to their name, and simply expend a couple extra semesters relative to the average.

The real question is whether they learn to focus efforts on the relevant goal (in which case, these original thinkers with innate curiosity can be fantastic hires) or continue their rabbit-hole-exploring ways (in which case they generate publications as post-docs, but generally struggle in the private sector where folks want THEIR questions answers ASAP). Which are you?

(Fascinating discussion!)


I'd like to think I've gotten better about focusing on a relevant goal. I'm working in industry now, although still doing physics R&D.


One point that I seem rarely mentioned especially in the life sciences is learning when to tell your mentor that you have enough to leave. In my experience a lot of labs will try to keep their senior PhD students around as long as they can because they don't have a suitable replacement and they are cheap labour. I know people who stick around for 6 to 10 years to chase after a high impact paper (that doesn't often manifest) or they can't let their work go and pass it off to someone else, or in the very malicious cases the PI won't let them go until they publish another paper etc. The department I was in was determined to get the average PhD down to less than six years but still hasn't reached that point.

The trainees supervisory committee is usually there to push them out but in many cases they also have a close relationship with the PI and aren't going to force a productive student to graduate. Those extra years are rarely useful for their overall career prospects.

I think students need to be aware of when they should draw the line and move on. Spending three more years in their PhD probably won't pay off nearly as much as three years of accumulated experience in industry job or in a post doctoral fellowship in a new lab.


Yeah, I have seen instances where it's the student's lack of drive/progress, the PI being unreasonable, the student's project was too ambitious, plain bad luck (lab equipment caught fire & destroyed), or time taken off for family considerations (child-birth, caring for ailing parents).

I graduated in 6 years (including 2 years of coursework), which was the mode (not the median, though) for my department. The distribution skewed to the long side, and 5 years was the shortest I can recall.


The topic sentence idea is particularly valuable, and it's something I try to pass on to students. I also use latex macros to turn this on and off (and to put in margin notes, also). All of this advice was so similar to that I give my students that I went to the author's homepage here on HN, to see if it was somebody I had taught. (Nope, wrong field.)


I did something similar that was quite useful. I wrote a complete outline of my thesis, down to the paragraph level. Then I sat down with my advisor and went over it, before I did any writing. This had a couple useful effects. First, I knew he was in general agreement with my plan. Second, it acknowledged that I was in fact in the writing phase and wasn't doing any more experiments.

A useful side effect is that whenever I wasn't feeling really inspired, I could pick a paragraph at random and just fill it in. I would not call any of my paragraphs "filler" but there was stuff that needed to be written down, that didn't require profound brain work to produce.

Anyway, that's how we're supposed to write code, right? It was, 30 years ago. ;-)


Wow, that's very detailed. I can see it being beneficial, especially in the case of an unreasonable advisor who might demand late-breaking additions to the research.

For me, I started from the slides I had presented in my immediate group meetings (~6 ppl, including my advisor, typically once per week, 2-3 slides each) plus the larger group meetings (~40 ppl, including the lab director, typically twice per year, 20-30 slides each). That gave me bullet points and figures. I wrote one chapter at a time, starting with the central chapters & ending with the introduction & conclusion. I had a 6-month time table for writing, and I was only delayed 2 weeks in the end. Remaking figures and messing with LaTeX took more time than I wish it had.


I do that too, I tell students to first make a skeleton document, with titles that read like a story.

I do the same when programming btw, my function names read like a story with their complexities hidden lower in the class/library. Yes I have function names that some may find ridiculously long but it helps me a lot.


Would it be possible to share the macro for highlighting topic sentences? I'd love to try that out on a recent draft I'm working on.


Sure, I'm pasting it below. Since I might sometimes want bold-face, I am also colouring it red, which my journals will not permit. (I am not sure how this will format in hackernews, but the main thing is that you uncomment either of the two `newcommand` lines.) If you want to do other things, you might want to use the `\if` method, so that altering just one line will let you alter a bunch of properties at once.

  \documentclass{report}
  \usepackage{color}
  \begin{document}
  
  \newcommand{\topic}[1]{\color{red}\textbf{#1}\color{black}}
  %\newcommand{\topic}[1]{#1}
  
  \topic{Lorem ipsum dolor sit amet, consectetur
  adipiscing elit.} Praesent vel consectetur est,
  sed accumsan dolor.
  
  \topic{In malesuada in nulla eget aliquam.} In 
  facilisis erat neque, non sollicitudin felis finibus a.
  Sed pellentesque suscipit lorem, quis lacinia mi
  suscipit at.

  \end{document}


Lessons from my Ph.D.:

(1) Role of Math. In most fields of research, the most respected research mathematizes the field, that is, makes progress with math techniques and results. So for Ph.D. research, try to have math play that role.

(2) Ugrad Preparation. To be successful with that role of math, have a good ugrad math background. Then maybe get some more math from independent study, work in a career, a Master's program, or whatever. Likely the math topics that both come first and are the most important are calculus and linear algebra.

(3) Find a Good Problem. In your career, independent study, whatever, find a good problem to solve. Pick a practical problem and intend to get an engineering Ph.D. where a solution to that problem is regarded as good research. Make some progress on solving the problem.

(4) Pick a University and a Department. Want a department that respects applied research, maybe in a school of engineering. Hopefully the university will state their standards for a Ph.D. dissertation, e.g., "An original contribution to knowledge worthy of publication." Look at their description of their Ph.D. qualifying exams. Do enough study at the ugrad or Master's level and/or independent study to be well prepared for the exams. If the department offers courses for preparation for the exams, in addition plan to take those courses.

(5) Enroll. Become a grad student in the chosen department.

(6) Progress. In your first year, take some courses, especially in subjects you already know well. Continue your research. Pass the qualifying exams. If you see some opportunities for doing some fast publishable research, as co-author, better as sole author, do that. Show the department that you have done publishable research. Then, sure, technically will have done a Ph.D. dissertation (I did that).

(7) Finish. In your second year, finish your research project, stand for an oral exam, and graduate. Of course, if there is any question about your research being publishable, then just PUBLISH it.

Done.


I think a 2 year PhD is very atypical in my field (CS). In my experience I learned a different lesson than you when it comes to 3 and 4. I found that more important than anything, picking an advisor is the way to choose your PhD. They have such an outsized role over your experience, much more than the University, Department, or even the problem IMO.

To give some context as to my experience:

- PhDs are funded. You get a stipend and tuition is paid for. This funding is either through a research assistantship (RA), or a teaching assistantship (TA). Either way you are expected to devote 20 hours a week to this task, and the rest would be devoted to your coursework. Typically you take 9 credits per semester for about 4 years, and then after you enter candidacy (you're not really considered a PhD candidate until you pass qualifiers, before then you're a mere PhD "student") you reduce that to a 1 credit "dissertation maintenance" per semester.

- Grant money is the lifeblood of a PhD granting research-focused department. Here's how the economy of a typical CS department works:

-- Newly hired faculty are given a "startup grant" that they use to bootstrap a lab. Their motivation is they want to get tenure in 6-8 years. To do that, they will need to justify to the Dean that they are capable of generating sufficient grant revenue.

-- Grant agencies award grants largely based on published research papers. Therefore the primary directive of a new academic is to publish research, and use that to obtain grants. Hence the phrase "Publish or perish"; if a researcher fails to get enough grants when the tenure clock is up, they will usually be put out to pasture; failing to get tenure is the death knell for a young academic's career.

-- So they hire a couple PhD students as RAs and they work on producing research papers for conferences. The new academic uses the published research in grant applications (the first target is usually the CAREER award). Soon enough grant this grant money is flowing to the researcher and they use it to pay for all sorts of things. Chiefly though, it is used to pay for the stipends and tuition of graduate RAs.

- As an RA, you will be expected to spend 20 hours per week on grant funded research. This means you don't have room to explore your own research topics! All of the grant money is allocated for the funded grant research, not your own whims. The best you can do is carve out some interesting angle on the research that you can call your own.

- By the time you get to maintaining your candidacy, you're already knee deep in publications on the funded research project. The path of least resistance at this point is to bundle them up into a dissertation and defend it.

- If you have your own research agenda, now is the time to execute it as a faculty member at another University. One of their primary concerns during hiring will be: "How is your research agenda different from your advisors?" You will perhaps not be surprised to find that many candidates fresh out of a PhD program will not have their own original thoughts yet. This is why many departments prefer that a new PhD actually take some time doing a postdoc where they can gain some independence from their mentor.

Anyway, what I would say is that instead of picking a problem or a university or a department, pick a person you want to work with for the next 5-8+ years. Like I said, 2 years is very atypical. In my department, we have built in buffers that would make the minimum I think 3 years with a Master's, and even then I think the typical time would be 4 years.


Interesting description of the system of academic computer science research and no doubt crucial for some computer science students.

For funding, I got tuition but no stipend.

For an advisor, I didn't want one or really have one. On paper I had two advisors, but I brought my own problem, did my own research, both for the dissertation and some publishable research I did before the qualifying exams, and didn't want, need, or get any advice from either of my advisors.

The best I got from my Ph.D. work was just terrific, fantastically good, powerful, valuable material. But there was a downside: I was attacked by some profs who resented me, wanted me to fail, and tried hard to have me fail. The actual academic work, including the research, was easy; most of the effort was just defending myself from attacks.

I do not now nor have I ever had any desire to be a college professor. I got a Ph.D. to be better qualified for a good career in applied math and computing I had going before my Ph.D.

Now I'm in business for myself. Math is not all there is to my business, but it is an advantage, likely a crucial one. The math is some math I derived together with some advanced pure math, a bit amazing, long in some advanced textbooks but not well appreciated for its potential for applications. The business is based on computing, and I've written all the code, all in Microsoft's .NET (which I like). The computer science used is just (a) the heap data structure used as a priority queue and (b) AVL trees for a cache. At one point I make use of LINPACK -- downloaded the Fortran version; got the Bell Labs program F2C to translate the Fortran code to C; compiled the C code as a DLL; and call it with Microsoft's platform invoke.

I've published in applied math (optimization), mathematical statistics (multivariate, distribution-free), and artificial intelligence. I didn't publish my dissertation research because I wanted, maybe, to SELL it and certainly didn't want to give it away.


I’m curious, who paid your tuition? Were you a TA for those two years?


I'm not sure just where the tuition came from. I was a TA for some of the time, but the duties were trivial. At one point a department prof confessed that the department had a lot more tuition scholarships than qualified grad student applications. Net, the university didn't much want to charge grad students for tuition.


Unmotivated details are also my pet peeve. I want to know why something is important or valuable before I decide to invest my time and energy into it.

Other people generally don't care about your personal struggles with a problem, so leave them out. Or at the very least don't lead with them. Lead with something that piques the interest of your target audience.


Solid advice, not just for PhD students. This is also invaluable for helping final year/capstone students navigate their supervisor interactions.

There’s nothing better than a student you wind up and they go off and solve a bunch of problems in interesting ways. They’re having fun, you’re workload is reduced and there’s even potential for a publication. Meetings are indeed about giving feedback and learning on both sides.

In contrast other students show up empty handed, unmotivated and expect a list of instructions some of which they might attempt. You feel like repeating yourself constantly and that they are not listening.


>In contrast other students show up empty handed, unmotivated and expect a list of instructions some of which they might attempt. You feel like repeating yourself constantly and that they are not listening.

What do you do in that situation? When I encountered it, I assumed it was a communication problem at first, so I asked the student to take notes on what I had requested. This didn't help. I then realized they didn't understand what I had asked for in the first place. I suppose I could have requested they repeat my instructions back in their own words. Ultimately, I figured it was a lack of motivation, b/c they would half-jokingly complain about whatever I requested, and I usually found them watching videos on their computer when I walked by.


It’s hard. In the case of final year students they just need to do enough to get by. This usually means finding something in the project description that they can actually do or are motivated to do.

Sometimes you need to have a talk about what is going on. This usually happens after Xmas when they screw up their interim report/presentation. They get to see their peers succeed so it’s a strong motivator.

With PhD students it’s even tougher. You have to work with them for years and build them up. Some lack confidence, some are over confident but can’t actually do anything. You need to avoid doing too much for them — that’s the hardest part for me.

Occasionally you get postdocs that are difficult. They really should know better at this stage and should not have been hired.


My lesson from my failed Ph.D (graduated with a Master's in Chemistry after far too long) is that after the first year take a long critical look at what you're planning to accomplish and whether you really made the progress you needed too.

I didn't, and simply tried to power ahead on the assumption I'd pull it out of the fire: this was absolutely the wrong conclusion. You already have a university degree, and you'll get paid more in industry: the right answer is to abandon ship it you're not looking at a clear path ahead by then.


A first year is a pretty harsh deadline to set yourself to know what you’re doing. Most people I know’s PhDs only came together in the last 18 months of their degree.


For most people in my program and others like it (molecular biology, US) the decision point tended to be at the end of the second year, after the qualifying exam(s). This is because the first year tends to be full of lab rotations and some classes, and you don't really start doing research until nine months in.


My program was pure research, so I'd say that decision point lines up pretty well with the "1 year of research in, are you looking good?"

Though it seems weird to me to be in classes and labs without it being associated with it's own academic achievement (did that qualify as a certificate of any sort?)


If you dropped out after passing the qualifying exam, you got an MS.


The author became a professor unlike the 90% of the other PhD graduates, so you need to take all the lessons with a heavy dose of selective bias. Most of the things the article talks about is the effective processes the author has learned during PhD. This is definitely useful in any work where agency is involved. However, these effective routines can be learned from a decent job in a good organization and does not need a PhD. There is nothing in the article that suggests a unique learning that can be achieved only through pouring years into an endless pursuit like PhD.

In my opinion, there is nothing unique that can be learned only through a PhD for a successful career (except maybe for a tiny slice of outlier of CS researchers). Most people will be well better served to take a job that provides some agency, or better try to start a company and fail. They can learn a lot more this way without jeopardizing their financial future.


"Most of what I learned during my PhD had nothing to do with my dissertation topic, grad school, or even computer science.

"These lessons are so ingrained into me now that I'm shocked when I find out that not everyone knows them! I think they can be applied to virtually any office job."

Taking a job that provides some agency is harder than it sounds. As is starting a company and failing without jeopardizing one's financial future. (And not everyone is really enthusiastic about learning those lessons that can only be learned that way.)


You can LC for a few months closed door to get a decent SWE job in a good company that can provide you enough agency. PhD on the other hand requires spending 5+ years in a lab cutoff from reality and ending up with no skills to get a real world job. Yes you lose some money (mostly other people money) if a company fails, but the learnings are well worth the loss of money. Its definitely not the case with a PhD.


> Lead or be led > If you show up to a meeting/internship/job expecting to be told what to do, then chances are someone will tell you something to do... Alternatively, if you show up to a meeting/internship/job with a convincing game plan, then chances are people will get out of your way so you can go do it.

Wholeheartedly agree.


My wife is finishing up her PhD in American History and I can tell she's getting bored, at least demotivated reading a billion books a week. I've seen this motivation decline in a lot of founders as well 4/5 years into their startups, but I've not seen much reignite the fire. Can anyone recommend any specific tips for staying motivated through a PhD?


Your wife is bored because she’s gotten what she can out of the program, and the dissertation is a slog. I had the same experience.

This is why there is a term for people who do the whole program but drop out without finishing the dissertation: ABD (all but dissertation). It’s the one non-degree people feel justified to list on their resume, because it takes at least 4 years to get there, and it’s still quite an achievement.

I was ABD for 3 years when I got bored, and I almost quit. I figured since I had all the skills, it didn’t matter that I didn’t have the degree. It’s just a meaningless credential. I asked a friend of mine who had gotten his degree whether it was worth it, and he said “Don’t do it, the plus side isn’t that great”

Then I went to his wedding. He was the only PhD in his family, and his mother made the DJ introduce him as “Doctor”. What he considered a meaningless credential made his whole family so happy and proud. That moment made me change my mind, and I finished my PhD.

And you know what, he was wrong about the credential being meaningless. Employers look at you differently when you are ABD versus PhD. He didn’t experience that because he was never ABD. Truly it was like night and day. Governments care when applying for visas. Grant agencies care. There is also a lucrative market for expert opinions, and ABD are not considered at all in this market. Credentials matter in our society, even if they don’t matter so much in the tech sector.

Anyway, I hope something I said here will convince your wife to stick it out! If you want a specific tip, I would say take a leave of absence. For me I could take up to 2 years off no questions asked, and rejoin. If her school has a similar policy, she can use that time to recharge, and come back fresh and ready to bang out her dissertation.


Just have to chime in and state that this extremely sage advice that is rare to find on HN, where credentialism is (rightfully so) looked down upon. Whether we like it or not, credentials are a very powerful social signal in our society, with very real benefits. If you've spent 3 years in a PhD program, you should absolutely finish unless you have a very compelling alternative opportunity that cannot wait.

A secret that no one often admits is that most PhDs get more out of the credential than advertised, because they aren't a von Neumann or a Fermi, whose credentials never mattered because everybody knew they were one in a billion geniuses.


Isn't there an issue that if you take two years off, in the meantime someone somewhere else is doing the same topic and bangs it out and steals your thunder? Does this kind of thing ever happen, or is it more of a theoretical possibility?


It's more theoretical. While it's possible to write a dissertation in a short amount of time (I wrote mine in 2 months), it really should represent multiple years worth of dedicated research. For example, I had a figure in my dissertation which took 4 months to create, in terms of data collection and processing. Just one single figure out of several dozen!

That's one aspect is that most dissertations are not very impactful. The general idea is that a good dissertation should extend the field of knowledge in just a small way. It doesn't have to represent a titanic shift in thought or be revolutionary in any way. So really most dissertations are not supposed to have much thunder at all to steal.

At that level of specificity, it's possible to know all the big players by name and what they're working on. It's easy to find your own corner in such a small group, and it's very rare for a dark horse researcher to enter the field and suddenly steal your topic. It's hard to become an expert in a field without being noticed by already experts.


Thanks for replying. I'm not implying that someone else would "redo" your entire body of work in that two years, but rather, that someone else had also already been working on it at the point when you took your two years off, and they wrapped it up in that two years, taking you somewhat by surprise. But yes, I imagine that for a sufficiently compact addition to Knowledge, it's straightforward to keep tabs on everyone working in that small area.


wow this is really great advice! And funnily parallels what I see in founders, founders who start their startups for the wrong reasons (Sam raised a $3MM seed and I'm smarter than Sam so I bet I can raise a $5MM seed. Shit, I raised a $5MM seed).

Thanks for the thoughtful reply, I'll share your words with her.


> My wife is finishing up her PhD in American History and I can tell she's getting bored

[PhD candidate] is finishing up [gender pronoun] PhD in [field] and I can tell [gender pronoun]'s getting bored.

I just turned your statement into a template that works for every PhD candidate of the last 100 years.


If I had written this article myself and titled it “Lessons from my First Year as a Full Time Software Dev” a lot of the bullets would have been the same. Especially the Daily Progress and Get Excited bullets. Those two tend to go hand-in-hand given how often you need to do a thing with the only tangible reward being the accomplishment of the thing.


I do not have a PhD, but this is how I believe professional work should (and probably used to) operate. I’m extremely bored and annoyed by management practices that do nothing but track and assign work. When I started my career, it seemed a like a lot like a PhD described here. It was in space systems, and everybody was responsible for finding a problem to solve independently within the mission requirements. In that scenario, a PhD felt like a waste of time compared to professional work. Today it feels like I sorely need it (or an MBA) to climb the micromanagement hierarchy.

Is there such a place today as a professional?


The value in formal education is mostly reading, writing, and thinking skills. How to express and communicate increasingly complex ideas to others. PhD is not much different.

There was just a Reddit post saying that 54% of US adults have a reading level equal to or below a sixth-grade level according to the US Department of Education. Many communication problems can be attributed to differences in prose, document, and numeracy literacy.

"If we want to have an educated citizenship in a modern technological society, we need to teach them three things: reading, writing, and statistical thinking." – H. G. Wells


Nice list! A bit micro on a few points, but nice things to highlight. On a zoomed out View, a PhD will teach you a lot about the life cycle of creative projects in general - idea inception, prototyping, feedback, iteration, presentation. It will also teach you perseverance - oh boy, will it teach you perseverance.

At the same time, it will not teach you some things you'd pick up in industry - team work in particular.


Great article. I especially loved the parts about presenting, greying out the boring stuff makes a ton of sense.

What is the appeal of posting your daily progress on a public form? Do you not feel this to be a kind of invasion of your privacy? Is there some benefit that isn't immediately clear?


Really helpful! I love the 'Managers as input/output machines' part.


This is really good and I've saved it to revisit it later. Very concrete and insightful. Unlike a lot of meta-posts about PhDs that tend to be pretty watered down or abstract.


This is really good advice, with actionable tips for once. Bravo!


You could have learned all of this stuff at a good mid sized company.


Invaluable.

I've put that on my list of things to distil, review, and put into action.


YMMV. What I learned from experience that I already knew from education is that absolute power corrupts absolutely. There is a real tension when picking a graduate school. Do you go with the topic of interest and a tyrant? Or do you pick an amenable advisor that's in a different research field? I guess one can get lucky and find both but that wasn't an option the year I applied. The thing about graduate schools and research domains is that there a typically only a handful of choices if one is lucky. The other thing I learned is graduate schools never admit absolute power corrupts absolutely and when one points that out one is immediately ostracized. Go along to get along should be a sign over ever graduate schools doorway. Or dog eat dog.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: