Hacker News Reader

Comment by fluxem | Hacker News Reader

I got PhD and I am still not sure how I feel. It took me 6 years, a lot of doubts and putting my life on hold.

You never know if you will finish your PhD. In fact, 50% drop out. Imagine spending 4 years of your life and then quitting with little to show for it.

First of all, PhD is too long (in the US). You spend 1 - 1.5 years taking grad classes that have little relevance to your subfield. If you haven't done research in undergrad, you spend the first year figuring out what is research. IF you did, there's still a lot less hand-holding as a PhD researcher compared to an undergrad researcher.

Now let's talk about research. First, you need to come up with a good idea that no one thought of it before. This requires understanding of all the previous research.

Then you try to implement the idea. Even if you implemented everything correctly, if your method is not better than the state of the art, you just wasted your time. (Of course, you need to implement the state of the art also)

Do you want to become a professor? Forget about. There are so many bright PhDs competing for a few academic openings. Even if you become a professor, it will take 10 years and a lot of work to get tenure. And if you finally become a professor, the pay is terrible compared to the industry.

I'm not even talking about PhD in other subfields like Physics, where people become Google's SWE after finishing their PhD. Such a waste.

View on HN

I think people from the US who consider a PhD should really look outside the US too, as many (not all, of course) of the disadvantages don't apply globally.

For example, in Spain:

- Typical PhD duration: 4 years. It's not even possible to go beyond 5 years (there is a hard cap) and if you want to do it fast, typically 3.5 years or so is enough in most fields if you do it full time.

- Tuition is a symbolic amount (something like 100 €/year or so).

- Most students do their PhD either with a grant or a contract. So it's basically a normal job. Not a high-paying one, mind you, but a job. You typically get something in the range of 1200-1600 €/month and social security. Enough to live reasonably well (not luxuriously, but comfortably) in almost any city that is not Madrid or Barcelona (in those two you'd need to share a flat and it wouldn't be very good, but in most of the rest of the country, a typical rent is around 600 €).

- I don't have figures on drop out but I don't think it's anywhere near 50%, if I had to give a pessimistic estimate I would say 20% at most.

Of course, the bit about becoming a professor is still true (although the competition is somewhat different, less about brilliance and more about sweat... very roughly speaking, churning out more papers than your competitors even if the papers aren't that great. But there are still much fewer professor jobs than PhD candidates).

No remotely reputable STEM PhD program in the US actually charges you tuition, they also pay you a modest wage like you describe in Europe. You might have to TA some semesters if you don't go to a top program or work for a well established prof. But the compensation even for that case includes tuition, health insurance, and stipend. It's not a bad deal especially in the early years of a program when you are actually spending most of your time taking classes and learning other background.

The capped length sounds nice, but I think the years they shave off are mostly just at the start, not the years you spend grinding out research. The first year of a US PhD would typically be separated out as an independent Master's degree in Europe.

I have no idea what the academic culture is like over there, so I can't actually say whether it would've been better with regards to the things I disliked about PhD. But the facts you presented don't make me optimistic. A lot of them I don't consider problems in the US personally, and the absolute biggest problem for me is the ridiculous professorship job pipeline.

I'd also be cautious about that 50% US drop out number. It is likely < 20% at a T25 program in any STEM field in the US, certainly in my field that is the case. Perhaps it is higher in CS because people intentionally master out after 2 years to take a lucrative tech job. But I wouldn't consider that failing with the kind of connotation "drop out" has.

I would suggest extra caution before attending a mediocre (or worse) program of course.

I don't know, when I read US PhD students posting in academic Twitter I'm under the impression that they have to pay huge tuition and go into debt, can barely subsist, are exploited for teaching and can't even do much research, and are often on the brink of depression.

I assumed reality would be somewhat less grim as my sample is probably biased towards the ones who complain, but maybe it's so biased that it doesn't even resemble reality at all.

There is professional school (med school, law school, business school) and there are PhD programs. The former do cost quite a bit in the US, but Ph.D programs in the US (at least in the sciences) don't cost anything if you get in and in fact pay you a salary (I have a doctorate in microbiology and that's how it was for me and everybody I know). That being said, this generosity has a dark side -- there are considerably more Ph.Ds produced than can have jobs in scientific research and that's because grad students are a source of cheap labor for professors.

I think it's true that many PhD students have less time for pursuing their own research interests than they would hope, it is very tied to your advisor's wishes. Brink of depression is probably not a huge exaggeration either, at least for a non-negligible subset.

Barely subsisting is definitely not true for a STEM PhD at a decent program though. You're getting paid way less than you could probably, but as long as you're willing to have a couple housemates (usually fellow students) and are not yet supporting anyone else it's a reasonably comfortable wage. Keep in mind that solid health insurance and student perks like university gym membership are also provided at 0 cost.

This might be controversial, but people who have excess expenses (for whatever reason) in their mid 20s probably shouldn't undertake a PhD in the first place. That's the only way a STEM researcher who didn't settle for a subpar school could possibly end up in additional debt as a PhD student IME.

Post doc wages, given the career stage and associated job culture/benefits, are notably shittier. Debt still feels like an exaggeration to me, but a 35 year old with a lot of training and unclear future career prospects should at least be able to support a middle class family lifestyle with their postdoc wages. A 25 year old that is working towards an actual degree and still benefiting from student status is such a different situation.

In hindsight, the maximally enjoyable thing might be to work for FIRE and then pseudo retire to being a PhD student. A PhD can be inherently valuable, the problem is just how tied it is to an insane career path.

Anyway, it's possible the posting students are humanities PhDs? At a good place they still don't pay tuition, but will often get basically 0 stipend. So they legit need loans for living expenses. Either that or Twitter amplifies complaints from schools that one really should not do a PhD at.

Don't made me start to talking about science in Spain.

> It's not even possible to go beyond 5 years

And we still wonder why nobody is trying to paint a new Giocconda now, and where are the big results

First they repeat "ad-nauseam" that only want the very passionate and ambitious people, but then force you to choose a mediocre theme to work on it, or to solve very complex problems that nobody has figured before in a couple of years, plus one to land and other to write.

- "We have very serious environmental problems that treat our survival as species. What we could do?... Oh; What if we give the people only four years to solve it and save everybody. Just to create suspense?"

This happens by two very simple rules

1. Anybody starting in science must be dumber than this boss. Scientists must be all mediocre and equally talented, have the same background and fit to the current ideas.

If I want to study a complex problem for six or ten, or twenty years with my own money, well... is my problem. I repeat. Is -my- problem.

No other fields have this restriction. You will never see things like:

- "You are an architect? If you can't build an airport or major monument in <4 years you quit and everything is burnt down. This called Sidney Opera is taking too much to build, I can't understand this shape, lets start again and trow the investment in the dumpster"

- "You are a politician?. If you don't became president in three years and a half (or are more than 30 Yo), you are fired"

Who are the politicians to say how much of my time I can invest in something, when they don't apply the same rule to themselves?. Oh, I forgot the second rule:

2. Any grant money graciously given for a politician must benefit the politician career. Period. This means that science is now crushed to fit electoral intervals of four years

Unexpected maternity? Two years of covid?, say bye to your career and the investment of decades of study

And then they delay the grants payment for two or three years, just for laughs.

You seem to make the assumption that large contributions should be made within the duration of a PhD program. In my view, a PhD program's goal is to train new researchers. If awesome breakthroughs are generated before the thesis, of course it's a big plus, but not the point... PhD graduates still have several decades of career ahead of them to paint their Gioccondas. The PhD is just to get them started, when they graduate they should be able to fly alone and generate independent research.

There is the problem that research funding (also after the PhD) is associated to projects that only last for a few years, but that seems to be pretty much universal.

> I think people from the US who consider a PhD should really look outside the US too, as many (not all, of course) of the disadvantages don't apply globally. > > For example, in Spain: > > ... > > - Tuition is a symbolic amount (something like 100 €/year or so).

Is that €400 total to complete a PhD programme? Or do you mean specifically _tuition_ fees are small, but there are still other fees applicable? (Just clarifying, cause in Ireland when I was at university, they spoke a lot about "free tuition fees", but universities still charged thousands per year to students for "administration fees", to cover everything that wasn't explicitly tuition).

Also worth bearing in mind that international students (i.e. non-EU) often have higher fees, which can be as high as double or triple what home students (EU) pay. (Though if the total is only €400 that's still obviously really good!)

Now that you mention it, there is also a final one-time fee for the thesis defense and PhD diploma.

But putting all the fees together, it's less than €700 for the whole PhD programme. There are no administration fees or other large fees.

I can't speak for Spain, but until recently (a bit over ten years) in France university attendance was completely free. Nowadays there's a registration fee, which is 380€ for a doctorate.

Denmark:

- Usually 3 year (a post graduate education expected, otherwise 5 years) - Considered a job, you are paid USD ~4k a month (depending on field)

All true, but in Europe you need a 5-year degree to start a PhD.

Don’t you normally need an undergrad to start a PhD in the US as well?

In the UK an undergrad is 3 years in most cases, then 3 years for a PhD, so 6 years total.

Some institutions will technically admit you for a PhD with no prior degree at all, but nobody does this in practice.

>Don’t you normally need an undergrad to start a PhD in the US as well?

Yes, but not a graduate degree.

For comparison, at least in Austria, you need to have finished both a Bachelor of Science (3 years) and a Master of Science (2 years) in order to be able to start a PhD.

That's usually not the case in the USA (and some common law countries), where you only need the equivalent of a Bachelor of Science in order to start the PhD. You then do a small version of the Master of Science in parallel with the PhD.

Oh, that explains why I see more Americans going for PhD (successfully or not) than, say, Germans. It never made sense to me because I was used to it being the thing you can do after an MSc/MA and those already seemed like a questionable choice for many jobs.

So the UK is probably the best model - 3 years undergrad, 3-4 years PhD max.

It's less common now in the UK to admit people to PhD programs directly from undergrad. In many cases you'd be expected to do a one-year masters first, although that's not a universal requirement.

Nowhere near universal. It’s common to go from undergrad to PhD in the U.K.

Tbh most Master’s, esp MSc, don’t add anything.

Varies a lot from field to field.

It might depend on the field, but in Math/CS at a number of good unis, the standard path is a 4-year integrated MSc (you get a BSc after 3), then apply to PhD.

It is possible to do the MSc over 5-6 years and spend some time working, whether in parallel with studies as a teaching assistant, or taking a term/year out to get some industry experience. But in terms of credits gained, an MSc is 4 years net of studies.

Yes, this is true. You cannot start directly with a US 4-year Bachelor's, you would need a 1-year Master's. I suppose this can be a deterrent for many, but I still see many people with MSc degrees that complain (understandably) about all the hurdles of US PhDs and could benefit from considering Europe.

That's not necessarily true in the UK. Many masters courses, like mine, take a total of four years. In some cases you can even get lucky and go straight from undergraduate to a PhD. I know at least one person who did that.

This is actually not a hard requirement, you can start after your bachelor's. Not sure about having no degree at all like a sibling comment mentioned, maybe that's possible too.

The thing that bugs me most about the PhD is how much of the final years are controlled by your advisor and the committee. You have a bit of leeway in your first few years to do some "blue sky" research, scoped by the grants your advisor has of course, but the last 1-3 years can be a nightmare since you're expected to produce work that is funded by and within the scope of your grant BUT also be innovative and something that your advisor and committee approve and let you graduate with.

> but the last 1-3 years can be a nightmare since you're expected to produce work that is funded by and within the scope of your grant BUT also be innovative and something that your advisor and committee approve and let you graduate with.

Great summary! This is exactly how it feels at 3 years in. Every sentence in a paper is a careful balance between adhering to science, the funding agency, the supervisors, and myself.

An important caveat I think: your mentor is critical. I’m a PhD student, and I have an incredibly lucky mix of subject matter and mentorship (my mentor has actually gotten some awards in the past for being a good one). (Also, he’s funny.) Every day my research is a blast. My mentor is truly an expert in his field (he’s actually mentioned in Norvig’s AI textbook introduction) and has the answers to my questions. I get just enough steering to make sure I’m on track to producing something novel, but I’m totally free to explore. This is all bolstered by me being truly extremely intrinsically interested in my sub field. It’s really a very pleasant experience. I’m not terribly worried about the prospects afterwards as furthering humanity’s knowledge is rewarding enough to me. Heading to industry as a code monkey afterwards wouldn't upset me as I’m really just having fun as a PhD student, and it’s all because my mentor is incredible.

It’s important to try working with multiple mentors if you can. I actually started with a different person, and he probably would have been considered by most to be ideal. I was basically totally funded to do whatever I wanted with very little mentor interaction. Interestingly, this wasn’t great for me. In part I wasn’t interested in the subject, but more than that I greatly benefit from weekly mentor interaction to check in with my direction. Now I don’t have funding (working to support myself while doing research), but I’m happy as a pig in mud.

Isn't working towards the top of any field going to have just as many chances of failure at every step? My wife was the very top candidate for her surgical sub-specialty in the year she completed her fellowship and she still barely got an attending position due to how few openings were available. I am nowhere near as talented as she is, but I still had a tremendous battle to get where I am in tech and I watch brilliant engineers tap out before hitting even PE routinely. Working hard for 4 years and having nothing to show for it is de rigueur. Most effort spent in the world is spent in vain, but you have no chance of winning if you don't play the game.

The difference with software engineering is even if you "fail" you usually have a nice pile of money to show for it. If you drop out of a PhD all you have is some cool stories and obscure knowledge.

> all you have is some cool stories and obscure knowledge

As if obscure knowledge is worthless. Trading the opportunity to make money for obscure knowledge is the whole value proposition of a Ph.D.! If one doesn’t value obscure knowledge, yeah, a Ph.D. Might be a bad value proposition for that individual.

Definitely this. You're still very comfortable if you only end up as a Senior SWE.

Nope. You can go ahead and make a startup which can fail and then you're left with burnout and an empty bank account. Wasting time is an integral part of life.

Most fields don't require all of the people working in them to be at the top of their field.

And the ones that do pay orders of magnitude more.

Orders? Plural? This is not true. Youre likely to make 90 percentile

Sorry, I suppose being a professional musician is one field with similar requirements for skill versus pay. Nobody's going to pay any money to a middle-quintile piano player, just like nobody in academia is going to pay any money to a freshly-minted middle-quintile PHD.

Middle quintile?

In percentiles, 40% to 60%.

Someone in the middle 20% of competence at a skill or profession.

Sorry, what's PE? /Junior engineer in non-english country

I assume it’s principal engineer.

Professional engineer. That is, a official state licensed engineer.

> I'm not even talking about PhD in other subfields like Physics, where people become Google's SWE after finishing their PhD. Such a waste.

There are senior high - paying FAANG roles outside webdev where quantitative skills like the ones acquired in subjects like Math/Physics/Theor. CS are imperative, in which you can get straight into after graduation without even having to go through the absurd LeetCode hiring process (it's significantly easier if your advisor recommends you to an old friend on the industry).

> Do you want to become a professor? Forget about. There are so many bright PhDs competing for a few academic openings. Even if you become a professor, it will take 10 years and a lot of work to get tenure. And if you finally become a professor, the pay is terrible compared to the industry.

This is very accurate, especially in the US where tenure seems to be largely reserved for people that come from financially strong families who effectively support them from undergrad to tenure track.

For comparison - UK/mathematics, I went into research on day one, my supervisor set me an 'easy' problem straight off the bat, and I learned on the job. A couple of extra research problems came up naturally during my PhD. Had a paper submitted by end of year 1, was ready to write-up my thesis by end of year 3. I had some teaching responsibilities, but nothing heavy like running a course. Not saying it always goes this smoothly, but I rarely felt like I wasn't moving towards the end goal. +1 to 'forget about becoming a professor though/pay is terrible' though!

I left the US to do a UK PhD for this reason. No teaching responsibilities and you can just do the PhD. And they really try to kick you out after four years (at least in Cambridge).

I did 1 year pre-apprenticeship, then 4 years apprenticeship (Baker). Then shortly after that I left the industry and got into IT.

shrug I think that life is a process of navigating obstacles while acquiring skills. Its not a RPG with you choosing a character that you have to stick with until the end.

Branch out... you'll be surprised at where your PHD learned skills will lead you.

A trick that made my PhD a great experience: getting married right at the start. My life wasn’t on hold, I enjoyed a very long honeymoon :)

I did that too, then I got a divorce at my third year. :)

Sounds cool, how did you finance the honeymoon? :)

there is more than one way to honeymoon; they aren't all trips to the Maldives.

Most important part is quality, low-stress time with your beloved.

I have friends who recently lost most of their physical possessions (including house) to a volcano, and they don't have high income or good insurance. They are rebuilding from a low place.

The wife is an absolute gem. Keeps talking about the flow and accepting the moment. She radiates optimism and joy.

I tend to agree with you, but I don't think it's that black though.

> Even if you implemented everything correctly, if your method is not better than the state of the art, you just wasted your time.

You didn't waste your time as you know now the state of the art, and also why your method isn't better.

> Do you want to become a professor? Forget about. There are so many bright PhDs competing for a few academic openings.

The world is a big place. There's more to academia than Ivy League in the US. And don't believe all professors are geniuses! I think it's possible to somewhat assess your odds to get an academic position when you start a PhD.

On the downside, academia is very competitive and being a second rank professor drowning in teaching and administrative duties isn't the most rewarding thing to do.

> Even if you become a professor, it will take 10 years and a lot of work to get tenure. And if you finally become a professor, the pay is terrible compared to the industry.

This also depends in which industry you want to work. Not all fields/companies pay as a well as Google. And having a PhD can also open doors in industry.

> Do you want to become a professor? Forget about.

Unless you want to be a professor of computer science. There, everyone is hiring.

Because who wants to be an assistant professor making $60k/yr or god forbid an adjunct making $30k/yr when you can work for a startup making $150k or FAANG making $400k or more?

I'm frugal by nature, and from Eastern Europe. How comfortable a life can one afford with $60k/yr or $30k/yr in the US? With this level of income, would you actually find any essential things or services lacking in life?

In other words, what's the actual, practical need for a person (or a smaller nuclear family) in the US to actually earn $400k a year?

Similar to the US, university-employed PhDs definitely tend to make less than many industry professionals in my country as well. Then again, it is nonetheless a very acceptable standard of living. As civil servants, life is very stable (as compared to e.g. one-person companies), and they are by no means poor. Most importantly, though, a doctorate program will very probably train you into a better, more analytical thinker (I've seen very few examples of the contrary).

For that reason, I've always found it questionable when PhD vs no PhD is debated solely on the basis of expected income. Maybe I should remove my pink glasses, but there is, still, a little more to it, isn't it?

The US is big in every sense of the word, cost of living varies hugely in different areas. There are places where $60k is more than enough to live nicely, and there are places where $60k would mean sharing an apartment, and being very careful about spending your money.

One thing to keep in mind is that you might $100k+ in student loan debt by the time you actually get your phd. So for that reason you go for the higher paying job.

Note that those figures vary wildly by field and institution.

The lowest salary offer I got on my faculty job search, with an apology re: salary bias against folks without clinical degrees, was half again what's quoted there.

> For that reason, I've always found it questionable when PhD vs no PhD is debated solely on the basis of expected income. Maybe I should remove my pink glasses, but there is, still, a little more to it, isn't it?

There's certainly an argument to take a pay cut to do a PhD if you enjoy doing it.

The thing is - go ask real PhD students 2 years into their PhD whether they're enjoying themselves. Given the 50% drop-out rate, you can probably guess what the answers will be...

> How comfortable a life can one afford with $60k/yr or $30k/yr in the US? With this level of income, would you actually find any essential things or services lacking in life?

It really depends where in US. $60k/yr in NYC or SF or Seattle will be hard -- you'll need to get housemates to share the costs of housing. But $60k/yr is some small town in Vermont or Montana is comfortable in my guesstimate.

The question is not "comfortable", but "how comfortable", which is an important distinction.

Does comfortable mean being able to find specialist doctors? Or even just enough doctors? Being able to eat a variety of fruits/vegetables year round? Constant supply of avocados? Being able to live near family? How secure is one's income, or how many possible buyers are there for what you are selling? How are the schools for one's kids? Will the school community be of acceptable quality?

Of course, these answers will vary person to person.

MIT has a living wage calculator that claims in my U.S. county a single person not in a relationship needs $ 37,169 before taxes to support themself.

If you believe this is accurate 60k is doable, 30k is not.

I doubt you could do 38,169 without a roomate.

From my experience your first number is off by 3X and sometimes more depending on the university.

But yes, you make less as a professor than you do in industry.

That definitely depends on the school. At my undergrad, tenure track CS faculty only got around 70k, going up to 90k over time. Lecturers got 55k. Adjuncts got paid more in exposure and good feelings than money. My grad school is a much bigger research school instead of a teaching school and the pay is surprisingly not much better (better, but not 180k until full professorship if ever from what I've found) from what I've heard.

For reference, at my uni (public) full tenured profs are making about $120k-$150k. Adjuncts are about $90k. Lecturers are around $70k. But that's in the CS department. Still, all of these people could be making at least double that in industry (some do double dip though).

Note: The following isn’t intended to be bragging or anything, I’m just making the case for this profession in a place where the role is routinely diminished, so I’m going to make a good case.

Well first of all the front page of HN with tons of companies laying off massive amount of their workforce is one reason. Professors have insane job security.

Another reason is that professors only have to work 8 months of the year. Quoted figures for professor salaries are for 9 month contracts. Your quoted figures are too low by 2-3x. Only the smallest no-name colleges pay $60k, and there’s no point in being an adjunct in cs with so many instructor openings at top schools.

To make up the difference in salary, I’ve established my own consulting firm and I do tech consulting. So I have the best of both worlds: stable salaried employment and I’m my own boss.

Speaking of being my own boss, I really don’t have one. The chair isn’t really a supervisor, and while I have accountability, how I teach and what I do day to day is determined by me.

Work life balance is really nice. Some days I’ll go in at 11:00 and be done by 4. Today I won’t be going in at all and I’ll work from home for 30 minutes and be done for today.

The rest of the time I spend on “research”, which is just my open source passion project. Most hackers are trying to figure out how to work on their own open source project full time, this is how I’ve done it.

Before being a professor it was impossible to get people to contribute. Now I have dozens of students who make contributions, and every year there is a fresh group who wants to get involved.

Also I’ve got access to alternative funding sources not available to people who aren’t professors. Yes, you may have a 400k salary at FAANG, but I’ve got a $4M government grant that I use as fun money to buy my own hardware for my own flights of fancy. Literally whatever I want to do, not what some FAANG CEO like Musk or Zuckerberg wants me to do.

Anyway, I like it, and I’ll never these kinds of job perks are hard to put a price on.

Also in academia with a comparable lifestyle. I am not as established but pursuing a similar path.

I didn’t realize it was possible to do this without a PhD. I want to do research, I’m working as SWE at a startup and finishing MS now, but it is online without advisor or anything. If that’s possible, that’s awesome, how does that work, like to get the position and so on, I’m guessing you would need to publish some work on your own first.

If you want to publish papers, you'll need to be a PhD or extremely good in the research division of a major company - e.g. Microsoft Research, but it'll be difficult to get there without the PhD and previous research to show (but possible).

I wouldn't hesitate to take that assistant professorship for a second. A life of the mind beats the hell out of building yet another web site or CRUD app for decades on end, which is what most in our profession do.

Sorry, are the down votes because people don’t believe CS departments are hiring? My department has been hiring continuously for 6 years and has had many failed searches. Our peer institutions haven’t fared much better. There are tons of openings, just visit any academic job board to see for yourself.

Reply on news.ycombinator.com