6 ms·

The PhD problem

36 points by brianwillis 15y ago

melling 15y ago
I think for the hard sciences and engineering we should be creating more research jobs in the private industry for PhD's. We have no shortage of problems that need to be solved and discoveries to be made. Private industry needs to be incentivised to do more long-term R&D.
- brianobush 15y ago
  Agreed. In my company, a fortune 500 company, we are concerned more with revenue than building technology to solve big problems. This leads to quick hacks that solve short term issues only to leave technical debt and little invested in truly solving a problem.
  - zheng 15y ago
    I'd be willing to guess that you are more concerned with revenue because you have shareholders to answer to. R&D is something that doesn't show up on the bottom line with regularity. Sure, you might develop something huge and make tons, but probably not. What would be interesting is to see some heavy quantitative studies done on the ROI of research in a corporate setting. If we could prove that R&D provides a decent ROI, I'd be willing to bet that the business types would be more willing to fund it.
    - achompas 15y ago
      Teasing out returns on R&D would be difficult. Companies that maintain R&D teams (Microsoft, Yahoo, AT&T, Google, Intel, etc.) tend to be industry leaders, which complicates the relationship between research and income. Further, research projects have varying timelines and levels of dedication -- you can work on a project full-time for weeks, only to table it for immediate needs -- so the NPV calculation would be difficult. I bet R&D departments benefit from economies of scale -- 5 PhDs can likely produce much more than one PhD at 5 companies, and a solid team attracts top talent -- which means the aforementioned big shots are the likeliest to produce meaningful research. Still, I would give a lot to work at a something like a "research startup."
      - zheng 15y ago
        Oh I agree, I don't know how you would go about it, and I'm sure it's not easy or it would have been done. Talking with someone who has actual experience in both research and ROI studies leads me to believe however that these companies have a pretty good idea of their research ROI, but that it is merely a hard sell to shareholders due to the time delay.
      - alextp 15y ago
        Microsoft Research seem very evaluatable (they routinely sell their work to product tems, and things like the algorithms driving the ad network from bing came out of MSR) in terms of the bottom line, and Google's research is so integrated with the usual engineering it's hard to know which is which (and hence I doubt it fares any worse than the rest of google in terms of contributing directly to the bottom line).
- URSpider94 15y ago
  What you are suggesting was the norm 20-30 years ago. Major corporate research centers like Bell Labs and IBM Watson were the crown jewels of their respective corporate owners. Today, these programs are mere shells of their former selves. Corporate America just doesn't have the stomach to fund fundamental research with a payback varying from years in the future (optimistically) to never (pessimistically). In fact, we have exactly the opposite trend -- university researchers have taken on all kinds of applied research projects that would have been considered far too commercial in the past. In turn, the universities are looking to license out their IP portfolios and take stakes in companies founded by their students and faculty to raise more cash.
  - pbsurf 15y ago
    The list of inventions that came out of Bell Labs is breathtaking: solar cells, information theory, the first transistor, the first MOSFET, the laser, CCDs, UNIX, and many more. Silicon Valley was essentially founded by William Shockley, one of the inventors of the transistor at Bell Labs. At Bell Labs the absolute best scientists could come together and just work, without having to deal with writing grant applications, teaching courses, or mentoring grad students (the three activities which now consume the vast majority of an academic researcher's time). I am convinced that if Bell Labs was as strong today as it was 50 years ago, we'd have kilo-core microprocessors (and know how to program them), all our genomes would be sequenced already, and we'd all be driving electric cars recharged by ultra-efficient solar power stations. Or rather they'd be driving themselves. Maybe Google can help get us back on track.
    - farout 15y ago
      I used to worship institutions such as Bell Labs. But I guess when you learn how the sausage is made, it takes the luster off of things. The book, My Life as a Quant: Reflections on Physics and Finance, talks about why he left Bell Labs to be a quant on wall street. I remember after graduating college. I was unsure what to do next. My profs wanted me to do PhD. My fiancee, a PhD candidate, was against it. The company I worked at was willing to pay for my masters in electrical engineering but again my fiancee made the argument since I was repulsed by large organizations that I would resent it. I would most likely end up working for those companies. He was right. This is what happened to my friends. 5 years ago, I was reconsidering a PhD, I thought comp sci or stats. So I dipped my toe and got a masters in stats - it was fun. Do a PhD? No. What for. Economically it would not make sense for me. I would only do it if I found a fun topic such as modeling detection of prostate cancer (my masters project) but so far the idea of working for the immoral pharma industry including the unis and rest of them makes me ill. I wish I learnt Marie Curie's secret on how to focus on the work and forget the rest. But I haven't. My husband works as a computer science researcher (AI) - I KNOW I never want to do this. Pittance for ground breaking work done at excruciating slow pace with so many failures. The shit he puts up so he can work on the stuff that interests him - boggles the mind. I almost want to scream - get a bloody commercial job (effectively 2x salary) and use your own money to run your own tests. But he has seen what the prospects and the environment through the eyes of friends smarter than him and knows the true cost of politics. Smart man. I would never advise anyone to do a PhD unless you are in love with the topic.
teyc 15y ago
The same problem was true for engineering/CS graduates in developing countries. The problem in that case was that the skills available did not meet the economic profile of the countries they were working in. For instance, labour may be so cheap it is not cost efficient to automate processes. In many cases, we need to properly analyze the chain of value creation, from basic sciences to applied engineering, and iron out the inefficiencies. Once this occurs, PhDs will again enjoy their proper place in the economy.
- jgamman 15y ago
  >we need to properly analyze the chain of value creation Central planning is ver rarely the answer. Besides, who is the 'we' in this proposition?
  - teyc 15y ago
    > Central planning is very rarely the answer I disagree. That's dogma. Dogma is rarely the answer. The a lot of university's research already relies on public funds which is disbursed by a central committee. There is little private money in this space because the timelines and risks are too high. As for solutions, I heard of a few entrepreneurs who regularly visit university's technology board looking for interesting solutions that they can take to market. This is a kind of gap filling move that can be done broadly, provided the frameworks are realistic. I remembered a tale about IdeaLabs where they were spinning off a startup and decided to give the entrepreneurs 80% share of the company instead of the regular 20%. If a university has unrealistic expectations, then this can also curb the introduction of a technology/technique. Then there are some schools who do very good basic research but little in the way of translational research (e.g. medical trials, FDA approvals etc). The reason these may be occurring is because the way research funding works is that these activities are outside of their scope and ability.
dvse 15y ago
One of the problems with CS PhD graduates at least is the lack of broad familiarity with their and related fields. Outside of a handful of top US schools the resources (i.e. high level graduate courses taught by people who actually understand what they are talking about) are simply not available. Indeed even there one needs to make an effort to take classes from other departments. Somewhere like UK with no course work requirements at all, it is not as easy to find a PhD student with an understanding of the proverbial bigger picture (of course there are exceptions etc). It is the broad understanding of the relations between methods employed in different fields that makes a PhD useful in an industry setting and not the narrow topic of their research (for example someone familiar with ideas used in numerical computing can do research level work in many fields).
pbsurf 15y ago
I would suggest that graduate programs begin requiring applicants work for a year or two before applying for or starting a PhD (not unlike MBA programs). This would help bridge the gap between the academic and working worlds and provide students with some much needed savings to carry them through graduate school as well as a potential placement after graduating. And, having seen what they can earn in the working world, only the truly dedicated would accept a graduate student stipend for five years, solving the PhD overproduction problem.
papaf 15y ago
Until recently I was living this problem. I finished my PhD in Computational Biology and then joined the ranks of the long term unemployed. I got 1 job interview in 1 year. Things have improved and I now have a position in academia (the first academic position that I applied for). I personally don't see that doing a PhD should be a issue for recruiters in industry. I worked as a programmer for 10+ years before doing a PhD and can confidently say that doing research was one of the most rewarding times of my life. I learnt an incredible amount - especially how to independently research and tackle hard problems. Does industry not have any hard problems? Perhaps this is just a period of adjustment. The people recruiting don't have PhDs so they don't know what research involves or what someone with a PhD can bring to a company. I have seen both sides of the fence and if I was ever in a position to recruit I would certainly consider someone with research experience on an equal footing with a new graduate. I often joked that I would have had the same employment chances if I'd have spent 4 years sitting on a beach instead of doing a PhD. In a complex modern world such as ours this is truly twisted and bizarre.
- apl 15y ago
  > I got 1 job interview in 1 year. Things have > improved and I now have a position in academia > (the first academic position that I applied for). > I worked as a programmer for 10+ years before > doing a PhD What kind of jobs were you applying for? It seems somewhat unusual to do 10 years in the industry, then get a somewhat but not really related PhD, then get back into the industry. It'd be great if you elaborated a bit.
varjag 15y ago
One rarely mentioned thing is that most PhD degrees are worthless. This IMO is a key factor to understanding many problems plaguing postgraduate world. Of maybe dozen PhDs (spanning two different countries and fields of mathematics, physics and CS) I've been acquainted with, only one had uncompromising, endurance runner attitude that serious research takes. He was the only one who landed at research career after the defense. I fail to fully realize the reasons behind the motivation of the less successful ones. Whatever it was, it was certainly not a thirst for knowledge. Their contribution to advancing the state of art was zero. Those in the fields were I have a glimpse of understanding, should not have defended. They were net waste of grants/taxpayer money, and their own life. As it is now, postgraduate research is like uranium enrichment: a costly process with majority of output being waste. Of course the thing wasted in this case is people's effort and time which could (maybe) better spent elsewhere.
- michaelochurch 15y ago
  This is sad but true. A lot of people go into PhDs at 22 thinking they're intellectually invincible. They've gotten A's their entire life, never had any kind of serious setback, and fail to process the fact, when they hear PhD programs' abysmal job placement numbers, that those numbers will apply to them. This is even worse in the humanities, where the numbers are closer to 1-in-5 than 1-in-2. People who have been in the top 10% for their whole life fail to apprehend that all their classmates have been in the top 10% forever as well, and that they're counting on luck more than anything else. Five years in, the more successful PhDs have moved on or are at least in the process of finishing up, but about half of the students are long past burned out but will still be around for 2-5 years. Most careers encourage the less successful people to leave or fire them (which seems harsh, but it's less harsh than wasting peoples' time) but academia doesn't, or to whatever extent that it does, people don't get the hint. The result is that you have a lot of people who reach 29-31 when they finish their PhDs, and because they took too long they can't get great post-docs, and it's downhill from there. I think graduate school is a great way to learn a lot for people who have the ambition and ability, but the problem with it, from a career perspective, is that it's regressive rather than progressive for a lot of people. What I mean is that I'd rather be a first-year graduate student than a bottom-half investment banker, for sure, but the people in the "real world", even if they take some lumps early on, have already moved on, learned from a mix of success and failure (in most cases) and had a chance to establish themselves by 29-30. The less successful half of the graduate students haven't had this opportunity; they start out as graduate students at elite departments and, 7 years later, they're "wraiths" with nothing to show for their time. It's a really broken and unfair system. They should be given the hint much earlier (like after 1-2 years) so they can move on and do something where they have a better shot at success.
numeromancer 15y ago
tl;dr: I'm on board the SS PhD now, so let's pull up the boats before it gets too crowded up here.
sebkomianos 15y ago
The problem with not only PhDs but even the first levels of higher education is only one and it is universal: We see it from an economic perspective. A bachelors degree is good if it helps you do a masters degree. A masters degree is good if it improves your chances at finding a job. And then a PhD is good if it increases your expected salary. Well, I am sorry, but we shouldn't be complaining about "the PhD problem" then. If you do a PhD, do it to learn. Don't do it to find a job that will pay more.

36 points by brianwillis 15y ago

I think for the hard sciences and engineering we should be creating more research jobs in the private industry for PhD's. We have no shortage of problems that need to be solved and discoveries to be made. Private industry needs to be incentivised to do more long-term R&D.

brianobush 15y ago
Agreed. In my company, a fortune 500 company, we are concerned more with revenue than building technology to solve big problems. This leads to quick hacks that solve short term issues only to leave technical debt and little invested in truly solving a problem.
- zheng 15y ago
  I'd be willing to guess that you are more concerned with revenue because you have shareholders to answer to. R&D is something that doesn't show up on the bottom line with regularity. Sure, you might develop something huge and make tons, but probably not. What would be interesting is to see some heavy quantitative studies done on the ROI of research in a corporate setting. If we could prove that R&D provides a decent ROI, I'd be willing to bet that the business types would be more willing to fund it.
  - achompas 15y ago
    Teasing out returns on R&D would be difficult. Companies that maintain R&D teams (Microsoft, Yahoo, AT&T, Google, Intel, etc.) tend to be industry leaders, which complicates the relationship between research and income. Further, research projects have varying timelines and levels of dedication -- you can work on a project full-time for weeks, only to table it for immediate needs -- so the NPV calculation would be difficult. I bet R&D departments benefit from economies of scale -- 5 PhDs can likely produce much more than one PhD at 5 companies, and a solid team attracts top talent -- which means the aforementioned big shots are the likeliest to produce meaningful research. Still, I would give a lot to work at a something like a "research startup."
    - zheng 15y ago
      Oh I agree, I don't know how you would go about it, and I'm sure it's not easy or it would have been done. Talking with someone who has actual experience in both research and ROI studies leads me to believe however that these companies have a pretty good idea of their research ROI, but that it is merely a hard sell to shareholders due to the time delay.
    - alextp 15y ago
      Microsoft Research seem very evaluatable (they routinely sell their work to product tems, and things like the algorithms driving the ad network from bing came out of MSR) in terms of the bottom line, and Google's research is so integrated with the usual engineering it's hard to know which is which (and hence I doubt it fares any worse than the rest of google in terms of contributing directly to the bottom line).
URSpider94 15y ago
What you are suggesting was the norm 20-30 years ago. Major corporate research centers like Bell Labs and IBM Watson were the crown jewels of their respective corporate owners. Today, these programs are mere shells of their former selves. Corporate America just doesn't have the stomach to fund fundamental research with a payback varying from years in the future (optimistically) to never (pessimistically). In fact, we have exactly the opposite trend -- university researchers have taken on all kinds of applied research projects that would have been considered far too commercial in the past. In turn, the universities are looking to license out their IP portfolios and take stakes in companies founded by their students and faculty to raise more cash.
- pbsurf 15y ago
  The list of inventions that came out of Bell Labs is breathtaking: solar cells, information theory, the first transistor, the first MOSFET, the laser, CCDs, UNIX, and many more. Silicon Valley was essentially founded by William Shockley, one of the inventors of the transistor at Bell Labs. At Bell Labs the absolute best scientists could come together and just work, without having to deal with writing grant applications, teaching courses, or mentoring grad students (the three activities which now consume the vast majority of an academic researcher's time). I am convinced that if Bell Labs was as strong today as it was 50 years ago, we'd have kilo-core microprocessors (and know how to program them), all our genomes would be sequenced already, and we'd all be driving electric cars recharged by ultra-efficient solar power stations. Or rather they'd be driving themselves. Maybe Google can help get us back on track.
  - farout 15y ago
    I used to worship institutions such as Bell Labs. But I guess when you learn how the sausage is made, it takes the luster off of things. The book, My Life as a Quant: Reflections on Physics and Finance, talks about why he left Bell Labs to be a quant on wall street. I remember after graduating college. I was unsure what to do next. My profs wanted me to do PhD. My fiancee, a PhD candidate, was against it. The company I worked at was willing to pay for my masters in electrical engineering but again my fiancee made the argument since I was repulsed by large organizations that I would resent it. I would most likely end up working for those companies. He was right. This is what happened to my friends. 5 years ago, I was reconsidering a PhD, I thought comp sci or stats. So I dipped my toe and got a masters in stats - it was fun. Do a PhD? No. What for. Economically it would not make sense for me. I would only do it if I found a fun topic such as modeling detection of prostate cancer (my masters project) but so far the idea of working for the immoral pharma industry including the unis and rest of them makes me ill. I wish I learnt Marie Curie's secret on how to focus on the work and forget the rest. But I haven't. My husband works as a computer science researcher (AI) - I KNOW I never want to do this. Pittance for ground breaking work done at excruciating slow pace with so many failures. The shit he puts up so he can work on the stuff that interests him - boggles the mind. I almost want to scream - get a bloody commercial job (effectively 2x salary) and use your own money to run your own tests. But he has seen what the prospects and the environment through the eyes of friends smarter than him and knows the true cost of politics. Smart man. I would never advise anyone to do a PhD unless you are in love with the topic.

teyc 15y ago

The same problem was true for engineering/CS graduates in developing countries. The problem in that case was that the skills available did not meet the economic profile of the countries they were working in. For instance, labour may be so cheap it is not cost efficient to automate processes. In many cases, we need to properly analyze the chain of value creation, from basic sciences to applied engineering, and iron out the inefficiencies. Once this occurs, PhDs will again enjoy their proper place in the economy.

jgamman 15y ago
>we need to properly analyze the chain of value creation Central planning is ver rarely the answer. Besides, who is the 'we' in this proposition?
- teyc 15y ago
  > Central planning is very rarely the answer I disagree. That's dogma. Dogma is rarely the answer. The a lot of university's research already relies on public funds which is disbursed by a central committee. There is little private money in this space because the timelines and risks are too high. As for solutions, I heard of a few entrepreneurs who regularly visit university's technology board looking for interesting solutions that they can take to market. This is a kind of gap filling move that can be done broadly, provided the frameworks are realistic. I remembered a tale about IdeaLabs where they were spinning off a startup and decided to give the entrepreneurs 80% share of the company instead of the regular 20%. If a university has unrealistic expectations, then this can also curb the introduction of a technology/technique. Then there are some schools who do very good basic research but little in the way of translational research (e.g. medical trials, FDA approvals etc). The reason these may be occurring is because the way research funding works is that these activities are outside of their scope and ability.

dvse 15y ago

One of the problems with CS PhD graduates at least is the lack of broad familiarity with their and related fields. Outside of a handful of top US schools the resources (i.e. high level graduate courses taught by people who actually understand what they are talking about) are simply not available. Indeed even there one needs to make an effort to take classes from other departments. Somewhere like UK with no course work requirements at all, it is not as easy to find a PhD student with an understanding of the proverbial bigger picture (of course there are exceptions etc). It is the broad understanding of the relations between methods employed in different fields that makes a PhD useful in an industry setting and not the narrow topic of their research (for example someone familiar with ideas used in numerical computing can do research level work in many fields).

pbsurf 15y ago

I would suggest that graduate programs begin requiring applicants work for a year or two before applying for or starting a PhD (not unlike MBA programs). This would help bridge the gap between the academic and working worlds and provide students with some much needed savings to carry them through graduate school as well as a potential placement after graduating. And, having seen what they can earn in the working world, only the truly dedicated would accept a graduate student stipend for five years, solving the PhD overproduction problem.

papaf 15y ago

Until recently I was living this problem. I finished my PhD in Computational Biology and then joined the ranks of the long term unemployed. I got 1 job interview in 1 year. Things have improved and I now have a position in academia (the first academic position that I applied for). I personally don't see that doing a PhD should be a issue for recruiters in industry. I worked as a programmer for 10+ years before doing a PhD and can confidently say that doing research was one of the most rewarding times of my life. I learnt an incredible amount - especially how to independently research and tackle hard problems. Does industry not have any hard problems? Perhaps this is just a period of adjustment. The people recruiting don't have PhDs so they don't know what research involves or what someone with a PhD can bring to a company. I have seen both sides of the fence and if I was ever in a position to recruit I would certainly consider someone with research experience on an equal footing with a new graduate. I often joked that I would have had the same employment chances if I'd have spent 4 years sitting on a beach instead of doing a PhD. In a complex modern world such as ours this is truly twisted and bizarre.

apl 15y ago
> I got 1 job interview in 1 year. Things have > improved and I now have a position in academia > (the first academic position that I applied for). > I worked as a programmer for 10+ years before > doing a PhD What kind of jobs were you applying for? It seems somewhat unusual to do 10 years in the industry, then get a somewhat but not really related PhD, then get back into the industry. It'd be great if you elaborated a bit.

varjag 15y ago

One rarely mentioned thing is that most PhD degrees are worthless. This IMO is a key factor to understanding many problems plaguing postgraduate world. Of maybe dozen PhDs (spanning two different countries and fields of mathematics, physics and CS) I've been acquainted with, only one had uncompromising, endurance runner attitude that serious research takes. He was the only one who landed at research career after the defense. I fail to fully realize the reasons behind the motivation of the less successful ones. Whatever it was, it was certainly not a thirst for knowledge. Their contribution to advancing the state of art was zero. Those in the fields were I have a glimpse of understanding, should not have defended. They were net waste of grants/taxpayer money, and their own life. As it is now, postgraduate research is like uranium enrichment: a costly process with majority of output being waste. Of course the thing wasted in this case is people's effort and time which could (maybe) better spent elsewhere.

michaelochurch 15y ago
This is sad but true. A lot of people go into PhDs at 22 thinking they're intellectually invincible. They've gotten A's their entire life, never had any kind of serious setback, and fail to process the fact, when they hear PhD programs' abysmal job placement numbers, that those numbers will apply to them. This is even worse in the humanities, where the numbers are closer to 1-in-5 than 1-in-2. People who have been in the top 10% for their whole life fail to apprehend that all their classmates have been in the top 10% forever as well, and that they're counting on luck more than anything else. Five years in, the more successful PhDs have moved on or are at least in the process of finishing up, but about half of the students are long past burned out but will still be around for 2-5 years. Most careers encourage the less successful people to leave or fire them (which seems harsh, but it's less harsh than wasting peoples' time) but academia doesn't, or to whatever extent that it does, people don't get the hint. The result is that you have a lot of people who reach 29-31 when they finish their PhDs, and because they took too long they can't get great post-docs, and it's downhill from there. I think graduate school is a great way to learn a lot for people who have the ambition and ability, but the problem with it, from a career perspective, is that it's regressive rather than progressive for a lot of people. What I mean is that I'd rather be a first-year graduate student than a bottom-half investment banker, for sure, but the people in the "real world", even if they take some lumps early on, have already moved on, learned from a mix of success and failure (in most cases) and had a chance to establish themselves by 29-30. The less successful half of the graduate students haven't had this opportunity; they start out as graduate students at elite departments and, 7 years later, they're "wraiths" with nothing to show for their time. It's a really broken and unfair system. They should be given the hint much earlier (like after 1-2 years) so they can move on and do something where they have a better shot at success.

numeromancer 15y ago

tl;dr: I'm on board the SS PhD now, so let's pull up the boats before it gets too crowded up here.

sebkomianos 15y ago

The problem with not only PhDs but even the first levels of higher education is only one and it is universal: We see it from an economic perspective. A bachelors degree is good if it helps you do a masters degree. A masters degree is good if it improves your chances at finding a job. And then a PhD is good if it increases your expected salary. Well, I am sorry, but we shouldn't be complaining about "the PhD problem" then. If you do a PhD, do it to learn. Don't do it to find a job that will pay more.