Hacker News Reader

Comment by bertr4nd | Hacker News Reader

While it may be true that many PhDs are not a financially good decision, I think the answer is not necessarily so clear cut for a computer science PhD.

I didn’t have any sort of foot in the door at FAANG level companies before my PhD (ask me about my 3x rejections from Microsoft!), but I was able to parlay my CS PhD into such a job, and the earning potential there is profoundly higher than would be found elsewhere. Enough, even, to offset 6 years of lost earning potential.

Obviously this is just my experience, and those talented undergrads who can ace a Google interview are likely better off just taking it, but it’s not like the PhD was a vow of poverty.

View on HN

CS PhD is the outlier and not the norm. Most fields even in STEM its impossible to get a high paying job in the field of your thesis.

In chemistry at least, the PhD is the entry card for non-technician work. Post-docs are seemingly required for the academic path, and are of some use for pharma, but the union card is a must.

This seems more like a reason to avoid Chemistry than to get a PhD.

You don't always get to choose what you're interested in, as much as it might seem on HN.

Plus with a PhD you don't always choose where you go, you go where the door is open.

I think there’s something to be said for cultivating a passion rather than falling into one

Does the salary compensate the amount of effort to get there though?

I had a friend who was a chemist but she left the field to pursue data science instead after her pharmaceutical company started even more layoffs in R+D - it seems pharma companies are more interested in intellectual property law than drug development these days.

Not true. I worked in an engineering company that has a lot of jobs that require a PhD. Pay well, too.

Lots of engineering companies have positions where they value PhDs.

Will it offset the cost of not earning all those years? Probably not.

Will you get those jobs without a PhD? Probably not.

On other hand, are the jobs that are not those jobs much worse? For pay and career development at least? Counting the 3-4 years you have extra experience in industry.

IMHO they're not worse, they're different. The PhD-enabled jobs there aren't worth it purely in pay, they're worth it iff you really want to do a different type of work.

From what I can tell in engineering fields in industry, the biggest differentiator is the PhD tends to give you more autonomy.

So get a job in a field outside your thesis. FAANGy companies are happy to hire PhD from "other quantitative fields".

Oh boy, I was having a quick brunch at a software industry conference before Corona and I happened to sit next to this dude: he was getting into software testing after having worked his ass off a PhD and a post-doc in physics. Not to berate software testing but… wow, I lost my appetite thinking how bad his hand had been.

I work in civil engineering (in Europe), and all companies I've worked at have been full of PhDs doing work related to the thing they did their PhD in.

In my experience, civil engineering may not have cross-topic applicability like some other engineering disciplines. For example, someone with a PhD in concrete may not be able to jump to hydrology. But am industrial engineer who did a PhD in healthcare optimization could more easily parlay that into other optimization work. A aerospace PhD could work on windmill design, etc.

There are bootcamp grads with jobs at Google. There are self taught people. A PhD is a very expensive way of getting into a FAANG if there are people who manage it without a university degree of any kind.

There is a difference between front-end HTML/CSS typesetting and bleeding edge cryptography work.

Yes, and there are self-taught people working at Google who do bleeding edge cryptography work. Many, many fewer than those with PhDs but I know tptacek and daeken are excellent and have no formal qualifications.

If your goal is to get a job at Google a PhD is a poor way to get it.

> Yes, and there are self-taught people working at Google who do bleeding edge cryptography work

And we’ve come full circle to the topic of survivorship bias

We might have different definitions of "bleeding edge". I meant cryptography research, with implication that work is publishable.

Google Research employs 2000+ people. I believe not only they all have PhDs, but also from top-tier schools.

a PhD is really an absolute minimum requirement for those jobs, and with extremely fierce competition for those jobs, I would be seriously surprised there is anyone there without a PhD who is doing actual cryptography research. Not aware of any exceptions to this general rule.

"a job" could also be a janitor. yes, don't need a phd for that. for some jobs it's a bare minimum qualification.

(not to say people should necessarily strive for those, but such is life)

Googler here. Believe it or not, a lot of cutting edge work (even cutting edge crypto) is happening outside of the research organization. We even publish papers! And yes, some of those folks doing crypto work don’t have PhDs.

Hiring is weird. You definitely don’t need an advanced degree to get your resume looked at, but a PhD from a top program will at least get you interviews and so it gets you over one difficult hurdle.

It would be great if you could point to a specific example of a published article where a named author does not have a PhD. I 100% believe you but just to shut down those who believe that a PhD is a necessity to do research.

> Google Research employs 2000+ people. I believe not only they all have PhDs, but also from top-tier schools.

This would imply that Google Research is a great deal less meritocratic than Economics research, where someone like John List, who went to an unranked school, can end up teaching at UChicago. More generally ranking is not as big a dealn in Europe in at least one country, Germany. There are definitely people with doctorates from schools you've never heard of working at Google Research.

I don't think Google Research is as credentialist as you paint it. I'm willing to bet $200 at even odds that I can find one person working at Google Research as a researcher who does not have a PhD. My email is in my profile.

it is possible, as i stated just below in my comment i'm simply not aware of any exceptions.

to be sure, i specifically selected cryptography in the very first comment for a specific reason. math subspecialties require rather unhealthy amounts of training that is best started as early as practically possible. it is very hard to do, a mentor often helps and speeds up the process, and by the time you start doing original research as a fully self-taught mathematician, well, you might as well get get some recognition of your efforts by way of getting a degree, plus potentially get funded to do it. why not?

due to several factors, in all probability, successful applicants for mathematical research (cryptology) positions will have a phd, if nothing else simply because prior original research is likely to be required. competition pushes up the entry barrier as high as possible, for better or worse, and so the exceptions in research without a phd are going to be exactly that - exceptions.

the school ranking is a bit of a red herring here. while reputation of the school itself does matter quite a bit of course despite everyone doing their best to pretend it doesn't (but insist on sending their kids to the best school possible), it is simply that very often talent is sought out from specific groups/departments/professors who just happen to teach at this institution, wherever they are located. it's just that most of the time large well known institutions have budgets and brand recognition that attracts top talent, often from around the world. brain drain is real.

just to be sure, its not that i recommend everyone goes out and gets a phd in math (personally, i think that even a masters degree is a complete waste of time for the vast majority of jobs), but that in some situations it is a de facto pre-requisite. it likely will not pay off financially and it is a big commitment that can crush you mentally, emotionally and potentially even physically. some departments have sadly a reputation for high suicide rates.

btw. john list does have a phd, and spent 8 years teaching elsewhere prior to UC. not sure if that supports your meritocracy argument, as one would think someone with his reputation today could advance quicker.

> john list does have a phd, and spent 8 years teaching elsewhere prior to UC. not sure if that supports your meritocracy argument, as one would think someone with his reputation today could advance quicker.

I know John List has a PhD. I said he went to an unranked school. You had previously said

> Google Research employs 2000+ people. I believe not only they all have PhDs, but also from top-tier schools.

and I was pointing out that Google Research seemed unlikely to be more credentialist than university economic research, implying that there would be quite a few people at GR with PhDs from non-top schools.

I don't think there's much disagreement between us. I doubt there are more than 20 researchers at GR without PhDs, if that many. May your research and life be fruitful.

Reply on news.ycombinator.com