A PhD at a research focused experimental institution creates a particular kind of human that is absurdly resistant to stress and despair:
Ask an x-ray physicist using DESY about the horrors of “Beam Time” or a chemist about crystallising proteins.
Most people I know don’t use their actual skills anymore, but all of them shrug off whatever you throw at them at work without blinking.
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.
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.
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.
> 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.
> 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.
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.
Use zotero and betterbibtex. By all means type a comment so you know which ideas came from where but I'm a big advocate of taking notes by hand when you really want to understand something, as opposed to reminding your future self about something you already understand.
Ha! You just made me remember how much I used JabRef (open source bibtex reference app) back in 2004 when I did my PhD.
It was the best/worst 4 years of my life. I studied overseas (uk), met my future wife and got a PhD that really wasn't useful for much to me. Fortunately it was under a scholarship.
Not within a PhD, but as a side project I work on a research project on wikiversity about grammatical gender in French. It does reference a bunch of books and academic works, like probably a hundred I guess. The most tedious work though is to check which nouns are used only in a single gender of do have some epicenic or specific inflection used in the wild and giving a reference that attest that when it's not already so consensual that most general public dictionary would already document the fact. For that the research refers to thousand of webpages. I'm glad that most of the time I just need to drop the DOI, ISBN, or page URL and MediaWiki will handle the filing of the most relevant fields. That's not perfect, it generates the output with many different models currently (some don't have an excerpt field), and some required fields might be left blank, url to pdf won't work, and so on. But all in all it make the process of taking note of the reference quick and not going too much in my way. Creating a structured database out of it can certainly be done later.
Zotero and AI have this covered now. If there's one thing AI is good at it's summarising crappy formatted papers. Never understood the 2 and 3 column thing. Horrendous way to format something.
2-column format has narrower columns, which means that your eyes move more vertically than horizontally while reading it. That is considered conducive to “skimming” long texts if you’re a “speed reader”.
Do you mean that you’re using AI as a search engine for your local bibliography? I haven’t seen any AI plugins for Zotero.
I'm severely dyslexic and the columns are a massive hindrance for me, and I also cannot skim read due to this and meares irlen. So my dislike is not universal applicable, just personal experience.
On the zotero front there a bunch of AI plugins. But I've not used them. But yes the premise is that your can speak and ask your library questions. Some are set up differently though. Personally I can fire a paper into an llm and get a good idea of the content immediately and then ask questions about it. It's more interactive and allows me to get a better idea of it prior to reading it.
LLMs make too many mistakes when summarizing papers in their current state, I would never trust it to summarize a whole paper at the moment.
I only use it on a sentence or paragraph basis, otherwise it misses the point 90% of the time.
I would strongly advise against this use for the moment.
The important part of reading a paper is not only to extract general rules, but to build your own internal model. Without it you cannot effectively do research. The main interesting points are often in the subtleties of the details deep in the paper.
Internal tought that come easily to mind when I read :
- 'oh they used that equation, but that could be also be interpreted totally differently, what happens if we change point of view, does it makes sense from this other perspective'
- 'I see they claim to achieve better results than sota, but actually, they compared with other methods that are not solving exactly the same problem, what shortcut or changes did they had to do to obtain a fair comparison, is it a fair comparison, can I trust those numbers? '
- 'oh, the authors didn't realize that they solved this other problem, or did they realize but there was a block somewhere preventing it?'
- 'I like this trick to achieve that result, but at the same time, it will prevent to solve a whole class of other problems, so their method will not work on those cases'
...
Also, notice that a paper IS a summary of multiple months/years of work, and researchers summarize it already to the maximum to stay within the page limit, by summarizing a summary you will always miss many things.
I have found a lot of pearls casually buried in the paper, that there is no way a summary, either human or AI, would extract. Things like changing a method slightly, or recovering an old method to apply to a current problem, menctioned like it is not important but actually you have a project blocked in a similar spot.
Fair points. And likely why I'm not suited to academia too. I've just never really groked the practice. I've obviously only experience from bachelor's and masters but it always seems that you have an idea and the research is finding papers to back it up, and then some that might not. The work you do doesn't really matter as it seems secondary to the nonsense around "the literature".
I've a bachelors of science (first) in computer science, and currently doing a dissertation for a master's in cyber security, on route for a first but that might change depending on the mark for this dissertation.
My experience with the bachelors was that despite my project being derailed by the bullshit around formatting the document, doing "research" by searching the library for peer reviewed papers that backed up my claims, etc, etc; I got a excellent mark. In short I set out to make something and due to the academic processes failed in making anything, but because I was able to critically reflect on it, I got a good mark. Waste of time, unless you were just are a good mark.
For my masters I know the project doesn't matter, I'm concentrating on the academic nonsense because that's where the marks are.
The work you were given in your undergraduate and master’s was not research, it was homework. The task was critical reflection, which is repeatable and achievable for students; whereas research is expensive, one off, and generally out of reach for undergrads, and requires intensive oversight by an experienced researcher.
The waste of time would be for a professor to train you up to be a researcher before you’ve proven you are ready, hence the homework assignments.
If that's the case then and researching is way above masters level then how is it you get on a PhD? Genuine question. If everything I've done to date is a pale imitation of the real thing how can I make a fair assessment as to whether I want to pursue a PhD?
I feel like that's true when the font is insanely small, which I guess was good when people would print entire proceedings.
Reading two column super small font on a computer is super annoying though tbh.
The comments you write in to Zotero are not what paper is about - abstract covers this well enough - it’s about what you found interesting or useful about the paper.
I have had some fun exhuming my old LaTeX skills and assembling a BibTeX bibliography from which I automatically extract the right entries presented in whichever style is needed for a given paper and for my own (HTML) site. I even publish the collection in Zenodo in case useful to others. I use the 'annote' field for the reminder you suggest.
The lack of good tools to have good research notes with good search is kind of mind-boggling. I have reverted to having a website for myself, a private one that I run on my machine, using mkdocs which comes close to what I would want.
Presumably the idea is that you put the relevant parts of the list in your thesis. You need to convince your examiner that you understand the background to the original research you did, and a solid reference list (with supporting text in the introductory/background section of your thesis) is part of doing that.
Personally I did the references at the end and didn't feel like I suffered from that decision, but the key references in my particular area were a relatively small and well-known set.
Hmm, yeah. I mean you often see huge reference lists which always just makes me feel like the person can't possible be actually well acquainted with the stuff that's being referenced. So who are you really fooling? Seems all very performative, though I guess I understand the motivation
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.
As if "A" or "C" defined a person capacity. I know some straight A's that went directly for a repetitive and boring but well paid and stable job. Other stayed in academia and turned top scientists.
Academia is a very particular dynamic very difficult to find elsewhere, and some people dig it. You can watch some people finding the same dynamic at Google for example, where they are allowed and encouraged to fiddle around and keep publishing (e.g. the Attention paper, why would Google allow such publication?). Such dynamics are explored in Terence Kealy book "The economic laws of scientific research".
This varies widely between fields and institutions. Getting a PhD position nowadays in ML or computer vision is much harder. You need to already have publications when you apply and need to have experience specifically in the subfield, give a good talk, an interview, a good motivation letter / research statement, recommendation letters from good internships and multiple PIs you worked with, good grades, etc.
It can be different in other fields an in lower tier colleges.
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.
I'd go further and say it's dangerously untrue. What I advise people is that your results are constantly decaying. Only a rate of progress that exceeds the rate of decay will get you out the door. Decay happens for a number of reasons:
* Your records are never good enough to completely replace your memory of what you did. The longer it takes, the more studies, readings, etc., that you will have to repeat.
* In physical and biological sciences, equipment breaks down, gets taken away, facilities get moved, etc. This stuff happens at a constant rate, and is a pure time cost.
* Technological progress gradually raises the bar for the minimum quality of some results, e.g., in computation. Even "theory" is highly computational these days.
There are also risks of career-ending accidents that can be treated as a constant risk per unit time:
* Your advisor dies, retires, gets promoted to administration, loses funding, changes jobs, gets embroiled in ethical / legal issues, etc.
* Some unexpected new result from another team or industry erases the relevance or novelty of your work.
* You get sick, have family crisis, etc.
* Burnout
Results are the wrong unit of measure. A better KPI is results per unit time. The people who look like they fucked around for 4 years then submitted a brilliant thesis were either working hard all along, or were just brilliant, which I certainly wasn't.
My then-fiancee and I were both grad students. We made a pact to meet at 7:00 every morning in the cafe across from the research building for coffee, to force both of us to stay on a work schedule.
This is the kind of advice I give incoming graduate students. The sooner you start to treat grad school like a full-time job, the better. I was in a similar boat: my wife and I were both in grad school at the same time. We worked 9-5 every day, even if we weren't going in to the office. We both finished on time, and generally didn't have a difficult time with our degrees.
He is obviously talking about computer science. Yes, I know in biology or medicine you can often only access the experimental devices during set hours and the lab may not be accessible 24/7 etc. But in computer science the schedule is mostly free, except for meetings and teaching duties but those are specific time slots not a regular clock-in clock-out job like a cashier or bus driver.
In my part of the world (central Europe), the vast majority of PhD-students is actually employed by the university they aim to obtain the PhD from. So in addition to working on your thesis you most likely have to support other research projects as well as do a lot of teaching. The model of a free PhD student certainly exists, but it is rare.
It's pretty common in ivy-leave US universities (which is what the article is biased towards). There, you only have to TA a bit and you certainly don't work for the uni or the dept.
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.
“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!
That's what used to be thought about any school at all, then about high school diplomas then about a university diploma. Each time it was decided that by expanding the number of people they would get uplifted to a better standard of life, a higher class etc. But social status is relative and mostly zero-sum, so the value of a diploma simply goes down when everyone has one. Chasing credentials without actual value contributions cannot cash out in anything real.
"How to be a top YouTuber" by MrBeast. "How to be a top athlete" by C. Ronaldo. "How to become a popstar" by Ed Sheeran. You know that such advice will have limited usefulness to most of the aspiring people.
Karpathy is an exceptional person, maybe not as much as Ronaldo in football but taking advice from him similarly won't be guaranteed to work. You can't have guarantees in such things.
In truth, the more literal but not fully literal thing that happens regarding surviving a PhD is that you try to publish a paper in a top venue but after several rejections you publish in a lower tier one, then you do two more followup in similarly second tier but not terrible venues and you get a "magna cum laude" or perhaps a "cum laude" once you reach 5 years and the prof wants to avoid the embarrassment of not having graduated you.
Of course many people don't come into the PhD with such plans, they expect a summa cum laude and papers in top venues and talk invitations and so on, since they've always been a top student so far.
Or lucky! I had a great time during mine because my advisory was amazing. However, my cohort mates, many of whom I'd say are much smarter/intelligent than I, got stuck with terrible mentors.
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.)
Interesting. A couple of questions:
- How young are the kids?
- How do they behave, especially with essentials like eating and sleeping habits?
- Could you carve out a morning and/or evening routine for yourself?
- How much outside help could you rely on (grandparents nearby, lovely neighbours...)?
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).
If you want to work in the biotech/pharma business, you can be a PhD or you can be a technician. The latter needs to have "a good pair of hands" in the lab to be successful. The former, at least in the 00's, needed lab skills as well, but they tended to acquire a group when they moved into more medchem than synthesis. There's a number of support groups (physical or computational chemistry) that has similar staffing.
I did a bachelor degree part time later in life around work and family life. I'm doing a masters full time around work and family life. My experience with academia so far have put me off further study. I really don't get the research thing, and the whole experience seems like bullshit to me. Out of all my experiences doing these things the best has been on the taught modules, that I enjoyed and I didn't feel were out of date, the worse has been the dissertations where you're doing "research". Think of a project off the top of your head and "research" it. Nonsense.
Do I detect a hint of condescension? Hard to tell in text sometimes. I'm not sure how else to read your message, so in the event it is a passive aggressive attempt at being patronising I would caveat my original point with the fact I've spent the better part of a decade in higher education and I've done pretty well for myself in terms of marks, so if after all that I'm still not clear on what is meant by academic research or the point in the practice as presented to me during that time it's hardly my fault.
> so if after all that I'm still not clear on what is meant by academic research or the point in the practice as presented to me during that time it's hardly my fault.
Perhaps, but the mindset of delegating your intellectual advancement to others really is not compatible with being in academia, let alone getting a PhD.
Apologies if that's the way I'm coming across, my intent was the opposite. My mindset isn't one of delegating my intellectual advancement to others, my experience of academia is that academia actively stifled it. I almost dropped out of my bachelors be ause I felt I wasn't learning anything, but my wife talked me out of it.
I'm certainly interested in learning and exploring and finding out stuff, but my experience so far is that the academic processes stand in the way of this.
I'd love to see where you think I was being passive aggressive and/or condescending... Like I said above it's hard to tell in text but what I read in my original comment is an overtly aggressive overview of my experience in academia. You my have a different experience and that's fine.
> You’ll sit exhausted on a beautiful, sunny Saturday scrolling through Facebook pictures of your friends having fun on exotic trips, paid for by their 5-10x larger salaries. You will have to throw away 3 months of your work while somehow keeping your mental health intact.
Sounds strikingly similar to early-stage startup lifestyle.
Gah, don't take advice about doing a PhD from the dude who had the best possible academic experience! The vast majority of people who've gone through the PhD grinder have had radically worse outcomes than Karpathy. It's like taking advice about starting a cult from Joseph Smith.
(This is not to say you shouldn't do it. Just get info and advice from a less biased source).
One thing that's not mentioned here: if you don't come from a top university, you have close-to-zero chances to have that kind of experience in your phd. If you're not incredibly picking some exceptionally relevant project soon enough, your career path after the phd will not be exactly the smooth sailing the author describes.
Loved this article. I'd add a few things I wish someone had told me when I was starting my PhD: 1) Maximize variance, but know when to stop. Karpathy's point is great. Explore early, say yes to different things. But at some point you need to pick a direction and commit. Too much variance and you end up with nothing solid. 2) Consider smaller labs. Big famous groups are tempting, but in a small group of 3-5 people your adviser actually knows your work and gives you real feedback. In large labs you can easily become invisible. 3) Collaborate outside your lab early. Don't wait, reach out to people at other universities working on related problems. Different groups think differently and that's where good ideas come from. 4) Visit other universities. Even a few weeks at another group forces you to explain your work to people with different assumptions. It's one of the most useful things you can do during a PhD. 5)Learn to write good, structured, reproducible and maintainable code. One of the things I regret I didn't, and many working hours were wasted.
One angle that's missing from this discussion: the cost asymmetry. The effort to do this well is disproportionate to the effort to do it poorly, which means most implementations in the wild are mediocre. That creates a weird market dynamic where quality becomes the differentiator by default — not because it's hard to build, but because most people stop at 'good enough' too early.
> The minimum salary for the 2025 NFL season was $840,000
Raise the minimum salary of a Ph.D. student to that level and we have a deal. (The pocket salary, not $830,000 self pay to the university and $10,000 for the pocket of the student.)
Also, the work in NFL is more standardized, all teams play the same games per week, have a similar amount of training time, ...
In a Ph.D. the topic depends a lot of the advisor. It would be like mixing all the sports in the same bag, and for a weird reason the Waterpolo team from Alaska can pick you that are an expert in Tenis.
Doesn’t really touch what to do when a new political administration comes in and pulls all your funding or makes your research illegal. This happened to me twice as a grad student now as a researcher funding PhD students.
I can imagine that this will be similar to the "Emacs/Vim in the AI age" article - it will just be considered to matter less in the AI age. Why spend 3-5 years of your life with a sometimes frustrating experience to obtain this PhD degree if you have powerful models at your disposal that will just be able to solve everything for you? (Similar to why learn Elisp/VimScript/...) Especially considering the current trajectory, expecting where things will be in 5 or 15 years. It will just feel less and less appealing to get an in-depth education, especially a formal one.
Which is quite ironic, considering who wrote the article.
LLMs fall victim to "garbage in, garbage out." Claude can solve open problems if you know what you're doing, but it can also incorrectly convince you it's right if you don't know what you're doing.
A PhD teaches you how to think, how to learn, and how to question the world. That's a vital set of skills no matter what tool exists.
I don’t really know how to optimize for a world where AIs would be smarter than everyone and able to do everything.
If that comes to pass, I guess there won’t be any economic cost to having done my PhD because the entire economy will be AI driven and we’ll hopefully just be their happy pets.
If that doesn’t come to pass, and AIs just remain good at summarizing and remixing ideas, I guess people with experience generating research will still be useful.
Because you may have fun working in a scientific environment and doing research.
I liked my job at the university - independent of the final PhD. I enjoyed what I was doing. Most of the time I also enjoyed writing my dissertation, since I was given the opportunity to write about my stuff. And mostly I could write it in a way how I felt things are supposed to be explained.
Why spend your life doing anything at all? I'm biased on the topic since im writing up atm, but it was, if nothing else, a very itnerseting way to spend 4 years of my life.
I find it very fulfilling to do a PhD and did so myself. More people should. What I mean is that I'm expecting the general view on it to evolve as described.
Ah. I did indeed misunderstand. Also, as I said, I've got a personal stake, right at the tale end of the PhD, looking for jobs, so I guess im feeling pretty defensive. I certainly hope the general public doesn't feel this way, but I've seen plenty of people say similar things about college degrees now, so it kind of makes sense.
A PhD at a research focused experimental institution creates a particular kind of human that is absurdly resistant to stress and despair: Ask an x-ray physicist using DESY about the horrors of “Beam Time” or a chemist about crystallising proteins.
Most people I know don’t use their actual skills anymore, but all of them shrug off whatever you throw at them at work without blinking.
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.
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.
> 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)
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.
> 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...
> 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.
> 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.
> 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.
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.
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.
I have never been more depressed than grad school working for a complete sociopath. Even people working for “normal” PIs suffered. You are so right
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.
Use zotero and betterbibtex. By all means type a comment so you know which ideas came from where but I'm a big advocate of taking notes by hand when you really want to understand something, as opposed to reminding your future self about something you already understand.
There's also better notes too for this.
Ha! You just made me remember how much I used JabRef (open source bibtex reference app) back in 2004 when I did my PhD.
It was the best/worst 4 years of my life. I studied overseas (uk), met my future wife and got a PhD that really wasn't useful for much to me. Fortunately it was under a scholarship.
Not within a PhD, but as a side project I work on a research project on wikiversity about grammatical gender in French. It does reference a bunch of books and academic works, like probably a hundred I guess. The most tedious work though is to check which nouns are used only in a single gender of do have some epicenic or specific inflection used in the wild and giving a reference that attest that when it's not already so consensual that most general public dictionary would already document the fact. For that the research refers to thousand of webpages. I'm glad that most of the time I just need to drop the DOI, ISBN, or page URL and MediaWiki will handle the filing of the most relevant fields. That's not perfect, it generates the output with many different models currently (some don't have an excerpt field), and some required fields might be left blank, url to pdf won't work, and so on. But all in all it make the process of taking note of the reference quick and not going too much in my way. Creating a structured database out of it can certainly be done later.
This is very good advice for a few reasons:
1. It reduces the odds of missing a key reference in your papers and accelerates the write-up of the (often mandatory) Related Work section
2. It helps you maintain a mental map of the field as your research progresses
Zotero and AI have this covered now. If there's one thing AI is good at it's summarising crappy formatted papers. Never understood the 2 and 3 column thing. Horrendous way to format something.
2-column format has narrower columns, which means that your eyes move more vertically than horizontally while reading it. That is considered conducive to “skimming” long texts if you’re a “speed reader”.
Do you mean that you’re using AI as a search engine for your local bibliography? I haven’t seen any AI plugins for Zotero.
I'm severely dyslexic and the columns are a massive hindrance for me, and I also cannot skim read due to this and meares irlen. So my dislike is not universal applicable, just personal experience.
On the zotero front there a bunch of AI plugins. But I've not used them. But yes the premise is that your can speak and ask your library questions. Some are set up differently though. Personally I can fire a paper into an llm and get a good idea of the content immediately and then ask questions about it. It's more interactive and allows me to get a better idea of it prior to reading it.
LLMs make too many mistakes when summarizing papers in their current state, I would never trust it to summarize a whole paper at the moment.
I only use it on a sentence or paragraph basis, otherwise it misses the point 90% of the time.
I would strongly advise against this use for the moment. The important part of reading a paper is not only to extract general rules, but to build your own internal model. Without it you cannot effectively do research. The main interesting points are often in the subtleties of the details deep in the paper.
Internal tought that come easily to mind when I read :
- 'oh they used that equation, but that could be also be interpreted totally differently, what happens if we change point of view, does it makes sense from this other perspective'
- 'I see they claim to achieve better results than sota, but actually, they compared with other methods that are not solving exactly the same problem, what shortcut or changes did they had to do to obtain a fair comparison, is it a fair comparison, can I trust those numbers? '
- 'oh, the authors didn't realize that they solved this other problem, or did they realize but there was a block somewhere preventing it?'
- 'I like this trick to achieve that result, but at the same time, it will prevent to solve a whole class of other problems, so their method will not work on those cases'
...
Also, notice that a paper IS a summary of multiple months/years of work, and researchers summarize it already to the maximum to stay within the page limit, by summarizing a summary you will always miss many things.
I have found a lot of pearls casually buried in the paper, that there is no way a summary, either human or AI, would extract. Things like changing a method slightly, or recovering an old method to apply to a current problem, menctioned like it is not important but actually you have a project blocked in a similar spot.
Fair points. And likely why I'm not suited to academia too. I've just never really groked the practice. I've obviously only experience from bachelor's and masters but it always seems that you have an idea and the research is finding papers to back it up, and then some that might not. The work you do doesn't really matter as it seems secondary to the nonsense around "the literature".
In what, that does not align at all with my experience in the sciences (where the point is the novelty, not summarise the literature)?
I've a bachelors of science (first) in computer science, and currently doing a dissertation for a master's in cyber security, on route for a first but that might change depending on the mark for this dissertation.
My experience with the bachelors was that despite my project being derailed by the bullshit around formatting the document, doing "research" by searching the library for peer reviewed papers that backed up my claims, etc, etc; I got a excellent mark. In short I set out to make something and due to the academic processes failed in making anything, but because I was able to critically reflect on it, I got a good mark. Waste of time, unless you were just are a good mark.
For my masters I know the project doesn't matter, I'm concentrating on the academic nonsense because that's where the marks are.
The work you were given in your undergraduate and master’s was not research, it was homework. The task was critical reflection, which is repeatable and achievable for students; whereas research is expensive, one off, and generally out of reach for undergrads, and requires intensive oversight by an experienced researcher.
The waste of time would be for a professor to train you up to be a researcher before you’ve proven you are ready, hence the homework assignments.
If that's the case then and researching is way above masters level then how is it you get on a PhD? Genuine question. If everything I've done to date is a pale imitation of the real thing how can I make a fair assessment as to whether I want to pursue a PhD?
I feel like that's true when the font is insanely small, which I guess was good when people would print entire proceedings. Reading two column super small font on a computer is super annoying though tbh.
The comments you write in to Zotero are not what paper is about - abstract covers this well enough - it’s about what you found interesting or useful about the paper.
I have had some fun exhuming my old LaTeX skills and assembling a BibTeX bibliography from which I automatically extract the right entries presented in whichever style is needed for a given paper and for my own (HTML) site. I even publish the collection in Zenodo in case useful to others. I use the 'annote' field for the reminder you suggest.
The lack of good tools to have good research notes with good search is kind of mind-boggling. I have reverted to having a website for myself, a private one that I run on my machine, using mkdocs which comes close to what I would want.
why? like what are you doing with this mega list later? Ive never felt the need for something like this in my research but maybe im missing something
Presumably the idea is that you put the relevant parts of the list in your thesis. You need to convince your examiner that you understand the background to the original research you did, and a solid reference list (with supporting text in the introductory/background section of your thesis) is part of doing that.
Personally I did the references at the end and didn't feel like I suffered from that decision, but the key references in my particular area were a relatively small and well-known set.
Hmm, yeah. I mean you often see huge reference lists which always just makes me feel like the person can't possible be actually well acquainted with the stuff that's being referenced. So who are you really fooling? Seems all very performative, though I guess I understand the motivation
Citeulike [1] was great but is sadly no more.
Add any paper you pick up to your tracking system before you read it. Make that part of your reading ritual.
Save the PDF right away, too. You may later lose access to the journal. Or, CiteULike (where I^Hyou uploaded all those articles) may go away.
[1] https://en.wikipedia.org/wiki/Citeulike
I did a PhD in the late 90s.
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.
>What was the motivation? Honestly, I was too lazy to get a job and staying in academia for another 3+ years seemed amazing
This is actually how I view academia. "Couldn't get a job"
It really lowered the prestige of a PhD for me. Heck, if I think through my PhD friends... none of them were A students. They were all C-tier.
As if "A" or "C" defined a person capacity. I know some straight A's that went directly for a repetitive and boring but well paid and stable job. Other stayed in academia and turned top scientists.
Academia is a very particular dynamic very difficult to find elsewhere, and some people dig it. You can watch some people finding the same dynamic at Google for example, where they are allowed and encouraged to fiddle around and keep publishing (e.g. the Attention paper, why would Google allow such publication?). Such dynamics are explored in Terence Kealy book "The economic laws of scientific research".
This varies widely between fields and institutions. Getting a PhD position nowadays in ML or computer vision is much harder. You need to already have publications when you apply and need to have experience specifically in the subfield, give a good talk, an interview, a good motivation letter / research statement, recommendation letters from good internships and multiple PIs you worked with, good grades, etc.
It can be different in other fields an in lower tier colleges.
I discounted the remainder of the piece after reading this:
For some programs, this is untrue. Your advisor, your experiments, or your conference deadlines govern your schedule.I'd go further and say it's dangerously untrue. What I advise people is that your results are constantly decaying. Only a rate of progress that exceeds the rate of decay will get you out the door. Decay happens for a number of reasons:
* Your records are never good enough to completely replace your memory of what you did. The longer it takes, the more studies, readings, etc., that you will have to repeat.
* In physical and biological sciences, equipment breaks down, gets taken away, facilities get moved, etc. This stuff happens at a constant rate, and is a pure time cost.
* Technological progress gradually raises the bar for the minimum quality of some results, e.g., in computation. Even "theory" is highly computational these days.
There are also risks of career-ending accidents that can be treated as a constant risk per unit time:
* Your advisor dies, retires, gets promoted to administration, loses funding, changes jobs, gets embroiled in ethical / legal issues, etc.
* Some unexpected new result from another team or industry erases the relevance or novelty of your work.
* You get sick, have family crisis, etc.
* Burnout
Results are the wrong unit of measure. A better KPI is results per unit time. The people who look like they fucked around for 4 years then submitted a brilliant thesis were either working hard all along, or were just brilliant, which I certainly wasn't.
My then-fiancee and I were both grad students. We made a pact to meet at 7:00 every morning in the cafe across from the research building for coffee, to force both of us to stay on a work schedule.
This is the kind of advice I give incoming graduate students. The sooner you start to treat grad school like a full-time job, the better. I was in a similar boat: my wife and I were both in grad school at the same time. We worked 9-5 every day, even if we weren't going in to the office. We both finished on time, and generally didn't have a difficult time with our degrees.
He is obviously talking about computer science. Yes, I know in biology or medicine you can often only access the experimental devices during set hours and the lab may not be accessible 24/7 etc. But in computer science the schedule is mostly free, except for meetings and teaching duties but those are specific time slots not a regular clock-in clock-out job like a cashier or bus driver.
In my part of the world (central Europe), the vast majority of PhD-students is actually employed by the university they aim to obtain the PhD from. So in addition to working on your thesis you most likely have to support other research projects as well as do a lot of teaching. The model of a free PhD student certainly exists, but it is rare.
It's pretty common in ivy-leave US universities (which is what the article is biased towards). There, you only have to TA a bit and you certainly don't work for the uni or the dept.
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.
“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!
True, but on the other hand, maybe only a very small subset of people should be doing a PhD.
That's what used to be thought about any school at all, then about high school diplomas then about a university diploma. Each time it was decided that by expanding the number of people they would get uplifted to a better standard of life, a higher class etc. But social status is relative and mostly zero-sum, so the value of a diploma simply goes down when everyone has one. Chasing credentials without actual value contributions cannot cash out in anything real.
Kind of feels a bit like
How To Get Rich?
Step 1: be rich
Step 2: you're now rich
"How to be a top YouTuber" by MrBeast. "How to be a top athlete" by C. Ronaldo. "How to become a popstar" by Ed Sheeran. You know that such advice will have limited usefulness to most of the aspiring people.
Karpathy is an exceptional person, maybe not as much as Ronaldo in football but taking advice from him similarly won't be guaranteed to work. You can't have guarantees in such things.
In truth, the more literal but not fully literal thing that happens regarding surviving a PhD is that you try to publish a paper in a top venue but after several rejections you publish in a lower tier one, then you do two more followup in similarly second tier but not terrible venues and you get a "magna cum laude" or perhaps a "cum laude" once you reach 5 years and the prof wants to avoid the embarrassment of not having graduated you.
Of course many people don't come into the PhD with such plans, they expect a summa cum laude and papers in top venues and talk invitations and so on, since they've always been a top student so far.
Draw remainder of owl.
"Never confuse education with intelligence, you can have a PhD and still be an idiot."
Or as a good friend told me when I was starting my PhD: "those of us that finish our PhDs are not the most intelligent, but the most stubborn "
Or lucky! I had a great time during mine because my advisory was amazing. However, my cohort mates, many of whom I'd say are much smarter/intelligent than I, got stuck with terrible mentors.
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.)
I finished a PhD while working full time with 3 young kids. Feel free to reach out if you've been interested and I can share my experience with you.
How did you keep the motivation up?
(I tried doing a PhD while working full time, and quit the idea after 3 years.)
Interesting. A couple of questions: - How young are the kids? - How do they behave, especially with essentials like eating and sleeping habits? - Could you carve out a morning and/or evening routine for yourself? - How much outside help could you rely on (grandparents nearby, lovely neighbours...)?
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).
If you want to work in the biotech/pharma business, you can be a PhD or you can be a technician. The latter needs to have "a good pair of hands" in the lab to be successful. The former, at least in the 00's, needed lab skills as well, but they tended to acquire a group when they moved into more medchem than synthesis. There's a number of support groups (physical or computational chemistry) that has similar staffing.
I did a bachelor degree part time later in life around work and family life. I'm doing a masters full time around work and family life. My experience with academia so far have put me off further study. I really don't get the research thing, and the whole experience seems like bullshit to me. Out of all my experiences doing these things the best has been on the taught modules, that I enjoyed and I didn't feel were out of date, the worse has been the dissertations where you're doing "research". Think of a project off the top of your head and "research" it. Nonsense.
You are right, sir, you should stay away from research. Don't worry, there are others that will handle it.
Do I detect a hint of condescension? Hard to tell in text sometimes. I'm not sure how else to read your message, so in the event it is a passive aggressive attempt at being patronising I would caveat my original point with the fact I've spent the better part of a decade in higher education and I've done pretty well for myself in terms of marks, so if after all that I'm still not clear on what is meant by academic research or the point in the practice as presented to me during that time it's hardly my fault.
> so if after all that I'm still not clear on what is meant by academic research or the point in the practice as presented to me during that time it's hardly my fault.
Perhaps, but the mindset of delegating your intellectual advancement to others really is not compatible with being in academia, let alone getting a PhD.
Apologies if that's the way I'm coming across, my intent was the opposite. My mindset isn't one of delegating my intellectual advancement to others, my experience of academia is that academia actively stifled it. I almost dropped out of my bachelors be ause I felt I wasn't learning anything, but my wife talked me out of it.
I'm certainly interested in learning and exploring and finding out stuff, but my experience so far is that the academic processes stand in the way of this.
About the same amount you had in your comment, to be honest.
I'd love to see where you think I was being passive aggressive and/or condescending... Like I said above it's hard to tell in text but what I read in my original comment is an overtly aggressive overview of my experience in academia. You my have a different experience and that's fine.
> You’ll sit exhausted on a beautiful, sunny Saturday scrolling through Facebook pictures of your friends having fun on exotic trips, paid for by their 5-10x larger salaries. You will have to throw away 3 months of your work while somehow keeping your mental health intact.
Sounds strikingly similar to early-stage startup lifestyle.
Gah, don't take advice about doing a PhD from the dude who had the best possible academic experience! The vast majority of people who've gone through the PhD grinder have had radically worse outcomes than Karpathy. It's like taking advice about starting a cult from Joseph Smith.
(This is not to say you shouldn't do it. Just get info and advice from a less biased source).
One thing that's not mentioned here: if you don't come from a top university, you have close-to-zero chances to have that kind of experience in your phd. If you're not incredibly picking some exceptionally relevant project soon enough, your career path after the phd will not be exactly the smooth sailing the author describes.
Loved this article. I'd add a few things I wish someone had told me when I was starting my PhD: 1) Maximize variance, but know when to stop. Karpathy's point is great. Explore early, say yes to different things. But at some point you need to pick a direction and commit. Too much variance and you end up with nothing solid. 2) Consider smaller labs. Big famous groups are tempting, but in a small group of 3-5 people your adviser actually knows your work and gives you real feedback. In large labs you can easily become invisible. 3) Collaborate outside your lab early. Don't wait, reach out to people at other universities working on related problems. Different groups think differently and that's where good ideas come from. 4) Visit other universities. Even a few weeks at another group forces you to explain your work to people with different assumptions. It's one of the most useful things you can do during a PhD. 5)Learn to write good, structured, reproducible and maintainable code. One of the things I regret I didn't, and many working hours were wasted.
Good luck to anyone starting out.
I am really curious to know how Karpathy would update this survival guide in the 2026. Hope to hear something from him!
One angle that's missing from this discussion: the cost asymmetry. The effort to do this well is disproportionate to the effort to do it poorly, which means most implementations in the wild are mediocre. That creates a weird market dynamic where quality becomes the differentiator by default — not because it's hard to build, but because most people stop at 'good enough' too early.
Why don't we assign grad students to PhD courses the way NFL draft works?
Let directional universities pick first and Ivies (and other prestigious universities) pick last.
Linked by Gemini: From https://en.as.com/nfl/these-are-the-lowest-paid-players-in-t...
> The minimum salary for the 2025 NFL season was $840,000
Raise the minimum salary of a Ph.D. student to that level and we have a deal. (The pocket salary, not $830,000 self pay to the university and $10,000 for the pocket of the student.)
Also, the work in NFL is more standardized, all teams play the same games per week, have a similar amount of training time, ...
In a Ph.D. the topic depends a lot of the advisor. It would be like mixing all the sports in the same bag, and for a weird reason the Waterpolo team from Alaska can pick you that are an expert in Tenis.
What would the benefit of that procedure be?
A more equitable distribution of quality among the universities, in the long run.
Good to see this again resurface !
Doesn’t really touch what to do when a new political administration comes in and pulls all your funding or makes your research illegal. This happened to me twice as a grad student now as a researcher funding PhD students.
Its so esoteric.
You do what your priest/advisor tells you. You honor the priest. You do the latex ritual.
Did it actually do anything? Ah in 8 years someone is going to replicate your study, it wont work, but too late! You got a PhD!
I can imagine that this will be similar to the "Emacs/Vim in the AI age" article - it will just be considered to matter less in the AI age. Why spend 3-5 years of your life with a sometimes frustrating experience to obtain this PhD degree if you have powerful models at your disposal that will just be able to solve everything for you? (Similar to why learn Elisp/VimScript/...) Especially considering the current trajectory, expecting where things will be in 5 or 15 years. It will just feel less and less appealing to get an in-depth education, especially a formal one.
Which is quite ironic, considering who wrote the article.
LLMs fall victim to "garbage in, garbage out." Claude can solve open problems if you know what you're doing, but it can also incorrectly convince you it's right if you don't know what you're doing.
A PhD teaches you how to think, how to learn, and how to question the world. That's a vital set of skills no matter what tool exists.
It seems your question largely boils down to: “why do anything when AI could do it instead?”
I think there are many answers to this, not the least of which is that AI can’t really do it instead.
I don’t really know how to optimize for a world where AIs would be smarter than everyone and able to do everything.
If that comes to pass, I guess there won’t be any economic cost to having done my PhD because the entire economy will be AI driven and we’ll hopefully just be their happy pets.
If that doesn’t come to pass, and AIs just remain good at summarizing and remixing ideas, I guess people with experience generating research will still be useful.
Because you may have fun working in a scientific environment and doing research.
I liked my job at the university - independent of the final PhD. I enjoyed what I was doing. Most of the time I also enjoyed writing my dissertation, since I was given the opportunity to write about my stuff. And mostly I could write it in a way how I felt things are supposed to be explained.
Models can solve the problem, but they can't tell you if the problem was worth solving in the first place.
Doing hard things has consistently made me more generally (not only in the narrow hard thing) competent and comfortable with myself.
Why go to the gym if you don't need physical strength? One needs to do something to not degenerate into a miserable state.
Why spend your life doing anything at all? I'm biased on the topic since im writing up atm, but it was, if nothing else, a very itnerseting way to spend 4 years of my life.
People seem to get my comment wrong.
I find it very fulfilling to do a PhD and did so myself. More people should. What I mean is that I'm expecting the general view on it to evolve as described.
Ah. I did indeed misunderstand. Also, as I said, I've got a personal stake, right at the tale end of the PhD, looking for jobs, so I guess im feeling pretty defensive. I certainly hope the general public doesn't feel this way, but I've seen plenty of people say similar things about college degrees now, so it kind of makes sense.