These pieces of advice are useful. However, they don't touch the bottleneck: mental health. And no, it is not "like any other demanding job". A PhD hits on two fronts - one is "all or nothing". If you spend years and still haven't submitted your dissertation, it is a career-ending failure. The other is its tie to one's identity. You put sweat, blood, and tears into your research, only to be rejected at a journal or conference because the result is "technically correct but not significant enough". Sure, there are similar parts in other careers - from talking with people, it works a bit similarly in medicine (when it comes to "all or nothing") and art (when it comes to this identity).
If people fail, it is mostly because they burn out. If they succeed, it is not unlikely that they will need to heal their burnout wounds anyway.
I am sure Karpathy's experience is different. But most people starting their PhDs are not Karpathy.
The peer review paper requirement puts you in a situation where if your topic of research happen to not be interesting for the reviewers (that you have no control over), you can be a talented student that worked very hard and still fail due to being out of time after multiple successive rejections.
Your supervisor may not understand this until it’s too late, and you may not have the ability to judge your adviser's ability to do so until you are committed.
The main problem is that you were raised in a school system where if you show up, study and do your assignments you are pretty much guaranteed to succeed sooner or later. A PhD is not like that.
Too many people stay in academia out of inertia and being comfortable with the "school" mode of existence and are afraid of the broad wide world and the decisions involved. They finish their masters and liked the classes and the thesis topic and so they stay.
But as you said, a PhD is quite different than all schooling before that. And that's good. A PhD is supposed to signify that you contributed new scientific value as judged by the expert international community, not just your teacher. Of course there are many wrinkles on this story like sloppy knee-jerk reviews etc.
But anything in life where you "just show up" and fulfill some explicit assignments tends not to be very valuable. If just showing up and doing what someone else decided for you is enough for a thing, that thing will lose value very soon. Similarly if you make sure almost everyone can do it, it won't have value, but will become a participation trophy.
But nothing in real life work like that. School is fake. You don't get a job just by showing up or having a diploma. Nobody will fall in love and start a relationship/family with you for showing up and fulfilling some list of criteria. Nobody will fund your startup or strike a business deal with your company because you showed up and did some assigned tasks.
In almost all aspects of life being proactive and exercising agency will get you much further than the teacher's pet mindset that school instills. And unfortunately rather than selecting for it, the PhD selects against such agency again because it's the safe option and people who are ready for an adventure usually dislike the academic environment. Not all of couse, I obviously don't mean every single person fits this. But in my experience this explains part of the mismatch in expectations and reality for the "I was a good student so a PhD felt natural" people. Not those come into the PhD with a well thought out plan, and knowing exactly why they want to pursue it, the upsides and downsides etc.
I recently dated a DMA in her 30s, and while incredibly talented and respected within her field, she had no experience sharing a life with someone else. As selfish and unfair as she said she felt, she wasn't wasn't ever willing to compromise.
I'm not new to dating educated professionals, and it seems to be an unfortunate recurrence balancing being torn between wanting a partner and family and worrying that any misstep could jeopardize their career or cost them opportunity.
> I'm not new to dating educated professionals, and it seems to be an unfortunate recurrence balancing being torn between wanting a partner and family and worrying that any misstep could jeopardize their career or cost them opportunity.
There exist a large number of very smart people who are not that career-obsessed (the traits that you need for a fast career are rather different from "highly smart").
I consider it to be likely that you value character traits that made the partner predestinied for a successful career, but don't like the fact that because of this, they are often obsessed about their career.
But anything in life where you "just show up" and fulfill some explicit assignments tends not to be very valuable.
Including paramedic or trauma surgeon, where the explicit assignment is to stabilize the human being in front of you?
Or how about plumber, to stop water where it shouldn't be, making it come out faucets and go down drains where it should? Had furnace issues when we switched it on in October, the HVAC technicians made the magic box generate warm air: his explicit assignment was 'make the house warm' and he showed up and did that.
Depends on your goals. Even as a plumber you have to think about what's worth what, how you specialize, and what the price is. Even as a doctor you may be under appreciated and the explicit goal is not so explicit all the time and sometimes curing the patient requires inventiveness and not only following a script or a textbook. The patients survival is often not purely a cookie cutter application of rules as it works in school.
And in some sense we can make the PhD requirements explicit too: prove that you are able to provide valuable and novel contributions to the scientific community as recognized by them. But it's not box ticking. But so is becoming an especially appreciated and respected doctor in your city. It's not just don't kill too many patients and you're golden.
If you want a mediocre run of the mill PhD it's not super hard. It is hard, but not super hard if you have academic aptitude. You can do something incremental and follow the trends, and eventually some venue will publish it. Not many will read it though and you won't become a leading voice though. There are many incremental paper that help many average PhD students get their degrees.
There are plenty of other non academic fields where you have to work hard and long hours, such as finance or law firms. You can choose less demanding alternatives too but they are less prestigious. Of course finance pays better.
I think the problem is that people want to treat a scientific career as both a regular job and also as something of a pop star personal brand type of thing, and they only want the upside of both.
There are many problems with academia, power balance etc of course. It's not unlike problems for aspiring movie stars or in other somewhat isolated bubbles within society like in sports or the theater scene, or how kitchens work in high end restaurants etc etc. If you want to become an acclaimed top expert in a subfield, it's not easy. And humans are going to do their human things around such opportunities for status advancements.
That's way too far from the article in question and I don't think it helps to discuss them together. Now if you propose that sciences and humanities should be brought more into conversation I can get on board but it's not easy because both sides approach it from a high horse attitude.
In practice these PhDs use the same label but have diverged a long time ago and their cultures, expectations and problems are quite different. As a STEM person, I also don't exactly understand what gets you a humanities PhD. I assume something like coming up with a new interpretation of a text, event or intellectual movement, connecting dots together in new ways that appears insightful to other scholars. But it's probably different in history, literature or other branches of the humanities as well. As I understand novelty is always part of it, as is significance in some way. You need to defend some kind of novel thought of yours, in a thesis.
> The main problem is that you were raised in a school system where if you show up, study and do your assignments you are pretty much guaranteed to succeed sooner or later. A PhD is not like that.
It was not my case and bold of you to assume so. I had peer-reviewed publications before I even applied for PhD.
While I do know some people who expected PhD to be "more classes with more difficult assignments", the mast majority of PhDs I know had nothing to do with mentality you described.
Sorry, my choice of word was not great, I was not targeting you in specific, but the way the education system works prior to a PhD in general.
I’m aware the experience will not be the same everywhere as PhD programs can be quite different, but many students that asked me if they should do a PhD are surprised when I tell them it’s not just up to the university or their advisors that they will get their degree and even more surprised when I tell them how early they have to start submitting their papers.
Even though I was well warned by my seniors, I admit that I also had this misconception to some degree.
This is very true especially nowadays. You typically need several publications to even start a PhD in competitive fields like AI, so people are familiar with the system already.
Your point is valid in many ways. The picture can be a little brighter. The PhD path does not have to be an all or nothing.
1. You can hedge your bets by submitting your work to various conferences of various qualities (without going 3rd-tier, you can bet across 1st-tier and 2nd-tier)
2. You can spend time choosing the professor and the topic before going all in
3. You can seek advice and social interactions within your research group, departement and school
None of this is a silver bullet, but it compounds in the right direction.
PhD programs are very different. The environment Karpathy describes is fairly similar to what I saw as a math PhD in a good school, but not an ivy. My theoretical physics PhD friends had the same setup as I had, but experimentals lived in a different world, long hours in the lab every day, including weekends.
My advisor was well established, tenured prof with a number of students. I had to teach, but the effort was light. We taught large, basic courses that are boring for tenured profs. We usually requested the same 1-2 classes to teach and after the first round had all the materials (homework, quizzes, etc.) and could teach on autopilot. University gave us undergrad graders to grade assignments but I never used them since I wanted to see what my students wrote. Which is a testament that the load was light; if I was drowning I would use all free help I could.
But there was a cult of academia. "Get an academic job or you are a loser" mentality was prevalent. My advisor was disappointed, but OK when I decided to go into industry after PhD, but a friend's (Physics PhD from Harvard, CEO of a profitable startup now) advisor does not talk to him anymore because he did not stay in academia.
And I only realized long after finishing my PhD how incredibly lonely PhD path is. You live in your bubble for years, with minimal interactions outside a few other folks at the same grad school. Stipend was enough for basic living, but not much else. No good vacations, ski trips with friends, etc. And a few somewhat creepy characters that grow in this lifestyle. This is all surmountable, but the mental toughness required is certainly something to keep in mind. I did not have that mental toughness, but was an introvert, which helped a lot. But looking back I see that I also could have gone off the rails. My 2c.
It depends highly on the field. In history, sure. The point of getting a history PhD is to become a history professor, and you can't do that if you don't get the PhD, and meanwhile history PhDs don't meaningfully open up any other job prospects, so attempting and failing to get a PhD provides negative value.
In CS and many engineering disciplines, there is a long history of people dropping out of PhDs and landing in industry. The industry is therefore much more accustomed to, and therefore accommodating to, people taking this path. Whether it's a maximally efficient use of time is another question, but it's certainly not wasted effort.
But I do agree that it's stressful nonetheless because it still feels like a failure even if it is not actually in reality. I wrote about this when I put down my own PhD journey here [1]. In particular after the control replication (2017) paper, I very nearly quit out of academia entirely despite it being my biggest contribution to the field by far.
Eh. It's different, but framing it as uniquely challenging seems silly. There are very few other jobs where you don't need to deliver any specific, measurable results for months or years. And your "career-ending" outcome is that you go and get a cozy industry job in the same field because you already have a degree. Now, you might have a difficulty adjusting to that because they will want you to get stuff done.
The one piece of advice I give new PhD students is to maintain a list of your references for a bibliography ahead of time. For every paper you read, copy the citation in BibTeX format and write a couple of sentences to remind yourself what the paper was about. Do this for every source, even if it doesn't seem important at the time.
I discounted the remainder of the piece after reading this:
Personal freedom. As a PhD student you’re your own boss. Want to sleep in today? Sure. Want to skip a day and go on a vacation? Sure. All that matters is your final output and no one will force you to clock in from 9am to 5pm. Of course, some advisers might be more or less flexible about it. . .
For some programs, this is untrue. Your advisor, your experiments, or your conference deadlines govern your schedule.
Karpathy is an interesting case of PhD gone industry and he mentions this topic in the article. In my field of computational social science it is sadly very taboo to happily leave the academy. Yet, they don’t do much to make it more appealing. My biggest win was to find a group of people outside of my research group that I liked collaborating with. Research is more fun as a team sport.
What was the motivation? Honestly, I was too lazy to get a job and staying in academia for another 3+ years seemed amazing (probably not recommended, but it worked out OK for me).
What helped get me through it:
1) Doing something I genuinely enjoyed - I approached the Computer Vision professor who gave me some ideas. I super enjoy writing code, and the idea of processing gigabytes of video to produce answers seemed cool. I treated it as a super difficult programming project.
2) Breaking my leg - Just before starting, I broke my leg badly. And that meant working from home with a weekly visit from the professor with a stack of reading papers. That time spent understanding state of the art was super useful.
3) Funding - At some point, DARPA gave enough money for me not to worry about funding, so I never had to work a job or get distracted.
4) Marriage - The final straight of writing a thesis was tough and I was super lucky to have a supportive wife who pushed me to get-shit-done.
I am a bit surprised that this article talks so much about actual PhD stuff than high level guidance. Maybe it has to do with the author's personal background/experience or field.
Something I didn't see in the article:
Depending on your field, it can be extremely hard to get tenure. Unless you are a genius and are seeing signs you are well on your way to getting professorship and tenure (publishing good and important papers, really good at communication, checking all the other boxes), you'd better consider an off-ramp as early as possible.
In some fields, 100 people compete for 1 open positions, and that's rough. Having been involved in such a hiring process, I know it's extremely challenging.
I was smart enough to realize I'd never be a good researcher relatively early during my PhD and started preparing for job interviews. Sadly, I see too many people still having unrealistic dreams about being a professor late in their PhD. They even do postdoc and do that for many years until they finally discover they are not hireable. Good news is that they often find an industry job quickly after switching.
I could see that clearly -- their research, originality, communication and understanding of the field is just not there, and one doesn't become a professor without being completely in control of the direction of their academic research.
But it's hard to tell people "you are just not good enough for this". In most situations, you'd be considered unsupportive. However, in this case, it's the best thing you can say to a struggling PhD/postdoc.
(Many professors are completely incapable of advising on students' career. They often find it surprising that a student wants to go into industry. They hold completely incorrect assumptions like a PhD student just goes into a management role as soon as they graduate. Rarely a professor tells you that you should not pursue tenure.)
“How to get into a top PhD program: get ~3 famous professors to write letters saying you’re one of the five best students they’ve ever worked with.”
I feel like this particular advice applies to a very small subset of people. If I’d had professors telling me that I certainly would have considered doing a PhD!
His guide to doing well in undergraduate courses is decent enough that I've sent it around to my students as well. I sometimes have to teach first or second year students and the amount of questions I get about how to study or how to do well is significant. We kind of forget that this is a learned behavior, and everyone learns it at different times in their lives (or not at all).
112 comments
If people fail, it is mostly because they burn out. If they succeed, it is not unlikely that they will need to heal their burnout wounds anyway.
I am sure Karpathy's experience is different. But most people starting their PhDs are not Karpathy.
See also "The Lord of the Rings: an allegory of the PhD?" http://danny.oz.au/danny/humour/phd_lotr.html
Sure you may survive. But even if all goes well, you succeed, there will be a void in you after the quest.
Your supervisor may not understand this until it’s too late, and you may not have the ability to judge your adviser's ability to do so until you are committed.
The main problem is that you were raised in a school system where if you show up, study and do your assignments you are pretty much guaranteed to succeed sooner or later. A PhD is not like that.
But as you said, a PhD is quite different than all schooling before that. And that's good. A PhD is supposed to signify that you contributed new scientific value as judged by the expert international community, not just your teacher. Of course there are many wrinkles on this story like sloppy knee-jerk reviews etc.
But anything in life where you "just show up" and fulfill some explicit assignments tends not to be very valuable. If just showing up and doing what someone else decided for you is enough for a thing, that thing will lose value very soon. Similarly if you make sure almost everyone can do it, it won't have value, but will become a participation trophy.
But nothing in real life work like that. School is fake. You don't get a job just by showing up or having a diploma. Nobody will fall in love and start a relationship/family with you for showing up and fulfilling some list of criteria. Nobody will fund your startup or strike a business deal with your company because you showed up and did some assigned tasks.
In almost all aspects of life being proactive and exercising agency will get you much further than the teacher's pet mindset that school instills. And unfortunately rather than selecting for it, the PhD selects against such agency again because it's the safe option and people who are ready for an adventure usually dislike the academic environment. Not all of couse, I obviously don't mean every single person fits this. But in my experience this explains part of the mismatch in expectations and reality for the "I was a good student so a PhD felt natural" people. Not those come into the PhD with a well thought out plan, and knowing exactly why they want to pursue it, the upsides and downsides etc.
> Nobody will fall in love and start a relationship/family with you for showing up and fulfilling some list of criteria.
The "date-me doc" community might disagree:
* https://www.nytimes.com/2023/08/02/style/date-me-docs.html
* https://www.astralcodexten.com/p/in-defense-of-describable-d...
* https://www.astralcodexten.com/p/highlights-from-the-comment...
I'm not new to dating educated professionals, and it seems to be an unfortunate recurrence balancing being torn between wanting a partner and family and worrying that any misstep could jeopardize their career or cost them opportunity.
> I'm not new to dating educated professionals, and it seems to be an unfortunate recurrence balancing being torn between wanting a partner and family and worrying that any misstep could jeopardize their career or cost them opportunity.
There exist a large number of very smart people who are not that career-obsessed (the traits that you need for a fast career are rather different from "highly smart").
I consider it to be likely that you value character traits that made the partner predestinied for a successful career, but don't like the fact that because of this, they are often obsessed about their career.
>
But anything in life where you "just show up" and fulfill some explicit assignments tends not to be very valuable.Including paramedic or trauma surgeon, where the explicit assignment is to stabilize the human being in front of you?
Or how about plumber, to stop water where it shouldn't be, making it come out faucets and go down drains where it should? Had furnace issues when we switched it on in October, the HVAC technicians made the magic box generate warm air: his explicit assignment was 'make the house warm' and he showed up and did that.
And in some sense we can make the PhD requirements explicit too: prove that you are able to provide valuable and novel contributions to the scientific community as recognized by them. But it's not box ticking. But so is becoming an especially appreciated and respected doctor in your city. It's not just don't kill too many patients and you're golden.
If you want a mediocre run of the mill PhD it's not super hard. It is hard, but not super hard if you have academic aptitude. You can do something incremental and follow the trends, and eventually some venue will publish it. Not many will read it though and you won't become a leading voice though. There are many incremental paper that help many average PhD students get their degrees.
There are plenty of other non academic fields where you have to work hard and long hours, such as finance or law firms. You can choose less demanding alternatives too but they are less prestigious. Of course finance pays better.
I think the problem is that people want to treat a scientific career as both a regular job and also as something of a pop star personal brand type of thing, and they only want the upside of both.
There are many problems with academia, power balance etc of course. It's not unlike problems for aspiring movie stars or in other somewhat isolated bubbles within society like in sports or the theater scene, or how kitchens work in high end restaurants etc etc. If you want to become an acclaimed top expert in a subfield, it's not easy. And humans are going to do their human things around such opportunities for status advancements.
>A PhD is supposed to signify that you contributed new scientific value
You forgot about the humanities, but it's hard to blame you. So do most grant offices.
In practice these PhDs use the same label but have diverged a long time ago and their cultures, expectations and problems are quite different. As a STEM person, I also don't exactly understand what gets you a humanities PhD. I assume something like coming up with a new interpretation of a text, event or intellectual movement, connecting dots together in new ways that appears insightful to other scholars. But it's probably different in history, literature or other branches of the humanities as well. As I understand novelty is always part of it, as is significance in some way. You need to defend some kind of novel thought of yours, in a thesis.
> The main problem is that you were raised in a school system where if you show up, study and do your assignments you are pretty much guaranteed to succeed sooner or later. A PhD is not like that.
It was not my case and bold of you to assume so. I had peer-reviewed publications before I even applied for PhD.
While I do know some people who expected PhD to be "more classes with more difficult assignments", the mast majority of PhDs I know had nothing to do with mentality you described.
I’m aware the experience will not be the same everywhere as PhD programs can be quite different, but many students that asked me if they should do a PhD are surprised when I tell them it’s not just up to the university or their advisors that they will get their degree and even more surprised when I tell them how early they have to start submitting their papers.
Even though I was well warned by my seniors, I admit that I also had this misconception to some degree.
1. You can hedge your bets by submitting your work to various conferences of various qualities (without going 3rd-tier, you can bet across 1st-tier and 2nd-tier)
2. You can spend time choosing the professor and the topic before going all in
3. You can seek advice and social interactions within your research group, departement and school
None of this is a silver bullet, but it compounds in the right direction.
> 2. You can spend time choosing the professor and the topic before going all in
To a point. If you stray too far from your personal interest, you increase the risk of burnout.
My advisor was well established, tenured prof with a number of students. I had to teach, but the effort was light. We taught large, basic courses that are boring for tenured profs. We usually requested the same 1-2 classes to teach and after the first round had all the materials (homework, quizzes, etc.) and could teach on autopilot. University gave us undergrad graders to grade assignments but I never used them since I wanted to see what my students wrote. Which is a testament that the load was light; if I was drowning I would use all free help I could.
But there was a cult of academia. "Get an academic job or you are a loser" mentality was prevalent. My advisor was disappointed, but OK when I decided to go into industry after PhD, but a friend's (Physics PhD from Harvard, CEO of a profitable startup now) advisor does not talk to him anymore because he did not stay in academia.
And I only realized long after finishing my PhD how incredibly lonely PhD path is. You live in your bubble for years, with minimal interactions outside a few other folks at the same grad school. Stipend was enough for basic living, but not much else. No good vacations, ski trips with friends, etc. And a few somewhat creepy characters that grow in this lifestyle. This is all surmountable, but the mental toughness required is certainly something to keep in mind. I did not have that mental toughness, but was an introvert, which helped a lot. But looking back I see that I also could have gone off the rails. My 2c.
> it is a career-ending failure
It depends highly on the field. In history, sure. The point of getting a history PhD is to become a history professor, and you can't do that if you don't get the PhD, and meanwhile history PhDs don't meaningfully open up any other job prospects, so attempting and failing to get a PhD provides negative value.
In CS and many engineering disciplines, there is a long history of people dropping out of PhDs and landing in industry. The industry is therefore much more accustomed to, and therefore accommodating to, people taking this path. Whether it's a maximally efficient use of time is another question, but it's certainly not wasted effort.
But I do agree that it's stressful nonetheless because it still feels like a failure even if it is not actually in reality. I wrote about this when I put down my own PhD journey here [1]. In particular after the control replication (2017) paper, I very nearly quit out of academia entirely despite it being my biggest contribution to the field by far.
[1]: https://elliottslaughter.com/2024/02/legion-paper-history (written without any use of LLMs, for anyone who is wondering)
What was the motivation? Honestly, I was too lazy to get a job and staying in academia for another 3+ years seemed amazing (probably not recommended, but it worked out OK for me).
What helped get me through it:
1) Doing something I genuinely enjoyed - I approached the Computer Vision professor who gave me some ideas. I super enjoy writing code, and the idea of processing gigabytes of video to produce answers seemed cool. I treated it as a super difficult programming project.
2) Breaking my leg - Just before starting, I broke my leg badly. And that meant working from home with a weekly visit from the professor with a stack of reading papers. That time spent understanding state of the art was super useful.
3) Funding - At some point, DARPA gave enough money for me not to worry about funding, so I never had to work a job or get distracted.
4) Marriage - The final straight of writing a thesis was tough and I was super lucky to have a supportive wife who pushed me to get-shit-done.
Something I didn't see in the article:
Depending on your field, it can be extremely hard to get tenure. Unless you are a genius and are seeing signs you are well on your way to getting professorship and tenure (publishing good and important papers, really good at communication, checking all the other boxes), you'd better consider an off-ramp as early as possible.
In some fields, 100 people compete for 1 open positions, and that's rough. Having been involved in such a hiring process, I know it's extremely challenging.
I was smart enough to realize I'd never be a good researcher relatively early during my PhD and started preparing for job interviews. Sadly, I see too many people still having unrealistic dreams about being a professor late in their PhD. They even do postdoc and do that for many years until they finally discover they are not hireable. Good news is that they often find an industry job quickly after switching.
I could see that clearly -- their research, originality, communication and understanding of the field is just not there, and one doesn't become a professor without being completely in control of the direction of their academic research.
But it's hard to tell people "you are just not good enough for this". In most situations, you'd be considered unsupportive. However, in this case, it's the best thing you can say to a struggling PhD/postdoc.
(Many professors are completely incapable of advising on students' career. They often find it surprising that a student wants to go into industry. They hold completely incorrect assumptions like a PhD student just goes into a management role as soon as they graduate. Rarely a professor tells you that you should not pursue tenure.)
I feel like this particular advice applies to a very small subset of people. If I’d had professors telling me that I certainly would have considered doing a PhD!