yamauchi@aic.dpl.scg.hac.com (Brian Yamauchi) (07/14/90)
I'd like to start a thread on what is undoubtedly one of the most often asked questions in grad school (at least in CS) -- "Are you planning to go into academia or industry?" To narrow this question down to a manageable size, by "academia" I mean a tenure-track position at a good research university and by "industry" I mean a position as a researcher at a top industrial lab. My perceptions are that the primary advantages of academia over industry are that you have complete freedom to decide what you want to research and that you are your own boss. The primary disadvantages seem to be that you spend an inordinate amount of time chasing grant money and that you have to teach (which may not be a disadvantage to everyone). Industrial positions seem to vary widely depending on the company and the lab. At the best labs, you may have almost as much freedom as those in academia (although you still have a boss). On the other hand, at some labs this freedom may go hand-in-hand with the need for chasing grants. Both universities and industrial labs tend to be well-equipped and to pay competitve salaries, so these areas don't seem to present major differences. Another issue is politics. While I've never been in a very bad political situation, I've heard horror stories from grad students and professors about the level of internal politics at some universities. Of course, I'm sure internal politics also exists to a certain degree in industrial labs -- but I'd be curious as to whether people think it tends to be worse in academia. Conversely, it seems as though industrial labs may be more affected by external politics (defense R&D cuts, corporate restructuring, etc.). I'm currently a PhD student in CS at the University of Rochester (working at Hughes over the summer), and my perceptions are based on 2 years of grad school, 4 years of undergrad (Math/CS at CMU), and 2 summers working in industry, so clearly, I've had more experience doing research in the "academia" side. I'd be interested to hear whether others (grad students/professors/researchers/etc.) have similar perceptions. ______________________________________________________________________________ Brian Yamauchi Hughes Research Laboratories yamauchi@aic.hrl.hac.com Artificial Intelligence Center ______________________________________________________________________________
shimeall@cs.nps.navy.mil (Tim Shimeall) (07/14/90)
(Note: I don't speak for the Naval Postgraduate School, nor the
Computer Science Dept. thereof, nor am I on the recruiting committee
of that department.)
First, be aware that the academic market for CS Ph.D.s is tightening
rapidly. In 1982, there were about 7 academic jobs open for every
Ph.D. that graduated (based on the Snowbird report). In 1990, many
departments have hundreds of Ph.D. applicants for every open position.
Many of the top schools are not hiring any longer, and the second and
third tier schools are filling rapidly. In short, unless you are an
exceptionally talented student in a rare-but-needed speciality, you
may have a VERY hard time getting an academic job.
I chose academia over industry for several reasons:
+ I wanted to set my own research agenda, and thought my research
ideas interesting enough to be funded (true, thus far.)
+ I wanted to teach. Don't underestimate the sheer fun of getting
students to understand hard concepts, and introducing them to new
ways of looking at the world.
+ I liked the freedom to travel and share information with other
researchers that academic folks have, but industrial folks often don't.
+ I did not want to deal with a profit-oriented company structure
(I'm not denigrating such structures -- they're needed; but I just
preferred to avoid them.)
The particular school I ended up at (the Naval Postgraduate School)
fit my desires almost perfectly. (I'll edit out a testimonial to the
school -- suffice it to say that this is a VERY nice place for a young
CS Ph.D.)
Tim
Disclaimer: The above is strictly a personal opinion and is not
binding on any person or organization.uselton@orville.nas.nasa.gov (Samuel P. Uselton) (07/14/90)
Expires: References: <9518@hacgate.UUCP> Sender: Sam Uselton Reply-To: uselton@orville.nas.nasa.gov (Samuel P. Uselton) Followup-To: comp.edu Distribution: Organization: CSC at NASA Ames Research Center, Moffett Field, CA Keywords: I've got some relevant (IMHO) experience, but as always, your mileage may vary. I spent 5 years as a full time grad student at Univ of Texas at Dallas, started teaching full time (Instructor) at Univ of Houston before the dissertation was done (**mistake**). I spent 3 years at Houston, and did finish the dissertation, then spent 7 years at Univ of Tulsa. I know the question asked to compare "top" university and industry labs, and I know these aren't, but I'm going to have my say. (You can hit "n" any time :-) ) UH and Tulsa are both in the "great middle" of universities they each have some good researchers and good departments. I like teaching and I enjoyed that aspect of university life. Grading is a drudge, especially in large classes with insufficient TA's. Curriculum design, both in the large sense of what courses to offer, to require, in what order, etc, and the small sense of what material to try to cover in a particular class (the first time you teach it) takes a lot of time to do well. In most cases, computing resources are not sufficient to allow students do implement / try out everything you would like, even in the places where there are sufficient resources for research. Politics varies WIDELY from places to place. Some departments are made of of groups that amount to armed camps that can never agree about anything. The more camps the less gets done. Other places can be friendly and harmonius. The only way you really discover this is by your own or someone else's experience. The political environment outside the department can be as important as within. (My promotion was shot down by a dean, not my department) The political environment (both acad. and industry) can change fairly abruptly, but stability can be estimated based on reputation, visits and scuttlebutt. The higher the research quality of the U., the more pressure to get grants to support (1) equipment (2) grad students to actually get work done, (3) salary (especially summer, but also course "release time"), (4) administrative "overhead", (5) travel, conference attendance for self AND grad students.... This pressure also increases as one gets more senior (close to tenure). I didn't think I would like industry as well, and never really considered it. I now work for Computer Sciences Corp., on a contract supporting NASA in the area of scientific visualization for Computational Fluid Dynamics (CFD) at the Ames research Center. This is a great job. The equipment is good, th eproblems to be worked on are interesting, there is sufficient freedom to pursue them in your own way, and the colleagues are good scientists. There are (of course) some hassles. Reporting to both bureaucracies is a nuisance. Government rules to avoid the APPEARANCE of impropriety are some times absurd. I miss teaching and getting to know new crops of students and the chance to get on my soapbox now and then (can you tell :-) ). About money,.... My experience is that on a monthly basis universities pay almost, but not quite lab pay scales. But they only pay 9 months salary. You have to hustle for the other, and some sources (eg NSF) will only pay 2 more. Real "industrial" labs probably pay better than government labs, although by using contractors rather than civil servants, gov't can at least get in the game. Real jobs pay 12 months salary including paid vacation. On the other hand the vacation doesn't match Thankgiving, Christmas, Spring Break, between semester breaks, .... What are YOUR priorities? location? pay? freedom? collaborators? interesting problems? ..... Sam Uselton uselton@nas.nasa.gov employed by CSC working for NASA speaking for myself
matloff@mole.Berkeley.EDU (Norman Matloff) (07/14/90)
In article <9518@hacgate.UUCP> yamauchi@aic.hrl.hac.com writes: >Of course, I'm sure internal politics also exists to a >certain degree in industrial labs -- but I'd be curious as to whether >people think it tends to be worse in academia. Conversely, it seems Yes, industry does have some politics, but it is FAR, FAR worse in academia. I say this having been in academia for 15 years, in three departments, and having a number of close friends at various schools (and in various fields) across the country. The degree varies from one place to another, but it *is* universal. I still feel personally that academia is the place to be. But if you are the type who is sensitive to injustice, selfishness, pettiness, meanness, callousness -- shall I go on? :-) -- then you will have to make sure that you are ready to put up with this muck as the price you have to pay for the genuine pleasures that academia has to offer. Norm
liu@beowulf.ucsd.edu (Hai-Ning Liu) (07/15/90)
shimeall@cs.nps.navy.mil (Tim Shimeall) writes: >First, be aware that the academic market for CS Ph.D.s is tightening >rapidly. In 1982, there were about 7 academic jobs open for every >Ph.D. that graduated (based on the Snowbird report). In 1990, many >departments have hundreds of Ph.D. applicants for every open position. >Many of the top schools are not hiring any longer, and the second and >third tier schools are filling rapidly. In short, unless you are an >exceptionally talented student in a rare-but-needed speciality, you >may have a VERY hard time getting an academic job. Thank you all on this discussion. I think each school is still looking for "new blood" to expand their programs. Well I am still a theory student, or a "utility" student, ( we design and publish all sorts of nobody-wanted-algorithms), I would like to know over the next few years which fields will still be "hot" and which will not; e.g. VLSI, networking, AI etc etc. My assumption is that if an area is not hot in industry, then it will be cold in most US schools.
asears@hcil.East.Sun.COM (Andrew Sears - Sun BOS Software) (07/16/90)
In article <9518@hacgate.UUCP> yamauchi@aic.hrl.hac.com writes: > > I'd like to start a thread on what is undoubtedly one of the >most often asked questions in grad school (at least in CS) -- "Are you >planning to go into academia or industry?" This is definitely one of those questions I've been asking myself. I agree, in general, with quite a bit of what you said. However, I have to question the point below. > > Both universities and industrial labs tend to be well-equipped >and to pay competitve salaries, so these areas don't seem to present >major differences. > I would have to disagree with the point about saleries. From my experience (both personal and friends) I'd have to say that you will probably make more money in industry. I don't remember the exact numbers, but you can look them up a relatively recent issue of the CACM, starting saleries for PhD's in academia are somewhere under $45,000 (I think it was closer to $35-40). Starting saleries, based on what industry pays PhD students for the summer (my experience and that of friends), can easily exceed $50-60,000. Granted this is based on a limited number of examples, but I would have to say that I think industry will probably pay much better. Andy ----------------------------------------------------------------------------- Summer: Sun Microsystems Real life: Human-Computer Interaction Lab Boston Development Center University of Maryland asears@East.sun.com sears@cs.umd.edu
rbl@nitrex.UUCP ( Dr. Robin Lake ) (07/16/90)
In article <9518@hacgate.UUCP> yamauchi@aic.hrl.hac.com writes: | > | > I'd like to start a thread on what is undoubtedly one of the | >most often asked questions in grad school (at least in CS) -- "Are you | >planning to go into academia or industry?" To narrow this question | >down to a manageable size, by "academia" I mean a tenure-track | >position at a good research university and by "industry" I mean a | >position as a researcher at a top industrial lab. | > | > My perceptions are that the primary advantages of academia | >over industry are that you have complete freedom to decide what you | >want to research and that you are your own boss. The primary | >disadvantages seem to be that you spend an inordinate amount of time | >chasing grant money and that you have to teach (which may not be a | >disadvantage to everyone). | > | > Industrial positions seem to vary widely depending on the | >company and the lab. At the best labs, you may have almost as much | >freedom as those in academia (although you still have a boss). On the | >other hand, at some labs this freedom may go hand-in-hand with the | >need for chasing grants. | > | > Both universities and industrial labs tend to be well-equipped | >and to pay competitve salaries, so these areas don't seem to present | >major differences. | > | > Another issue is politics. While I've never been in a very | >bad political situation, I've heard horror stories from grad students | >and professors about the level of internal politics at some | >universities. Of course, I'm sure internal politics also exists to a | >certain degree in industrial labs -- but I'd be curious as to whether | >people think it tends to be worse in academia. Conversely, it seems | >as though industrial labs may be more affected by external politics | >(defense R&D cuts, corporate restructuring, etc.). | > | > I'm currently a PhD student in CS at the University of | >Rochester (working at Hughes over the summer), and my perceptions are | >based on 2 years of grad school, 4 years of undergrad (Math/CS at | >CMU), and 2 summers working in industry, so clearly, I've had more | >experience doing research in the "academia" side. I'd be interested | >to hear whether others (grad students/professors/researchers/etc.) | >have similar perceptions. | > | >______________________________________________________________________________ | > | >Brian Yamauchi Hughes Research Laboratories | >yamauchi@aic.hrl.hac.com Artificial Intelligence Center | >______________________________________________________________________________ I've seen both sides of the fence. 12 years in academia and 8 years in industry. My Humble Opinions follow: Politics: They are bad on both sides of the fence when resources are scarce. You are probably more blessed by better senior management in industry than in academia, where one can watch university presidents and passive boards of trustees quickly run an institution into the ground. Stockholders would not stand for such non-feasance! Politics of funding, space, staff and time for research are about equivalent. Freedom: You must market your research ideas on both sides of the fence. The time scales may be different; the "clients" may be different. In both cases, the most certain funding is where you provide a needed service to clients and bootleg the research onto the service function. Used to be in industry we had "15%" of our time for "personal research", which became "15% for personal research and professional development", which became "0%" for personal research (that is unspecified, unapproved research) and "fight for" professional development support. This is not a complaint! I understand the need for "justification" and "financial control". It is reported that a new "culture" is coming down from above that will delegate more responsibility and authority. Fact of life: If you don't produce something that your institution perceives as valuable, you won't last as long as you may like to last. There are precious few institutions where teaching actually counts, unless it raises a LOT of money. Continuing sponsored research and publication in peer- reviewed journals counts. Often, your "management" won't be able to tell the quality of your research work nor the quality of your papers --- therefore scientific publications are full of work of questionable quality and slight reworkings of the same ten dog experiments..... perhaps more papers than experiments! Don't set your standards too high! In industry, given today's competitive world, you will still need to justify your existance. Dollars saved, products invented/patented/licensed/marketed, services rendered. And watch out for divestiture of the business/market-segment you are involved with. Time: There is administriva on both sides of the fence. Reporting can easily consume 30% of a week's work in industry. The same effort may be necessary to propose and report on project efforts when you are in an R&D lab that requires "business unit" support each year. Whether it is grant writing in adcademia or project proposals to meet business needs in industry, the time sink is still there. There may be differences in the dynamics. If a business needs something quickly, it will have funding quickly. If you deal with grant proposals, you may find much longer waits. Capital Equipment.: MUCH easier to come by in industry. If you need expensive hardware to do your work, it is much easier to acquire in industry. Lab space may be just as tight as in academia, but capital dollars still seem to flow. Pay: Make no mistake. Industry salaries are typically FAR better than in academia. Vacation policies may be better in academia, depending upon how tough a negotiator you are when you have been accepted for a job in industry. Health care benefits in industry are eroding quickly. Tests: How many signatures does it take to order a roll of toilet paper? (I know you won't be ordering your own toilet paper, it is just that you would still need the signatures....) How many peers have stayed/left in the past decade? When things get "tight", the more talented who are not "married" to a geographic area tend to vote with their feet. How many major management reorganizations have occured in the past 5 years? (Stability may be as valuable as funding in doing productive research.) Do your personal computing needs conflict with the prevailing computing culture? (Ever try to promote UNIX in a VMS culture?) NOTE: I AM VERY HAPPY IN INDUSTRY! THE COMMENTS ABOVE DO NOT REFLECT UPON MY EMPLOYER --- it is just what I typically advise my former graduate students. Each year I visit at least one university campus and try to keep my perspective --- the grass is equally as green on both sides, just shaded a bit differently.
siegman@sierra.STANFORD.EDU (siegman) (07/17/90)
Re economic factors: 1) For most any technical (science or engineering) area general impression is that base salaries in industry are better than in academia, especially in earlier years. And (again, this is only "general impression") academics work harder: more evenings and weekends, more stress. This is partly self-imposed competititve drive, especially in later years, partly tenure competition in earlier years. 2) Consulting opportunities -- if you're willing and able to take advantage of them -- can make up a lot of the base salary difference, maybe go well beyond it. Opportunities for patent or start-up income much larger in the academic sphere...though of course you have to be the right combination of lucky and entrepreneurial. 3) Job security in later years: MUCH better in academia. In industry, if you don't go into management -- which of course has it's own problems and threats -- then somewhere past 45 or 50 you are more and more likely to become an increasingly expensive, and increasingly out of date, liability rather than asset. There is now NO legally supprted mandatory requirement age in universities. This is, of course, a growing problem for the universities themnselves, and this situation may change in the future; but for the present it's a big plus for academic versus industry.
lumsdon@dtoa1.dt.navy.mil (Lumsdon) (07/17/90)
In article <9518@hacgate.UUCP> yamauchi@aic.dpl.scg.hac.com (Brian Yamauchi) writes: > I'd like to start a thread on what is undoubtedly one of the >most often asked questions in grad school (at least in CS) -- "Are you >planning to go into academia or industry?" To narrow this question >down to a manageable size, by "academia" I mean a tenure-track >position at a good research university and by "industry" I mean a >position as a researcher at a top industrial lab. Are you sure that you want to narrow your definition of academia to "a good research university"? If teaching really turns you on, why not teach at a school attended by 'less prepared' students. Helping a disadvantaged student learn a skill that can be used to support his/herself and family might give you a big charge. My mother is not in the CS field. One of the jobs that she enjoyed most was teaching nursing. Many of the students were working as nurse's aids, and earning an RN diploma raised their pay scale by at least $4.00 per hour (in the early 70s). She found it very rewarding to tutor struggling students, and to have a real impact on their lives. Yes, she also found it rewarding to teach the smarter ones :-) My mother didn't burn out; she wasn't so good at politics. She eventually took a 12-month job elsewhere. -------------------------- Esther Lumsdon -------------------------------- lumsdon@dtoa1.dt.navy.mil lumsdon@dtrc.dt.navy.mil lumsdon%dtrc.navy.mil@uunet.uu.net "Wherever you go, there you are" -Buckaroo Bonzai