Ask HN: How do you get a “PhD required” job without a PhD?

https://news.ycombinator.com/item?id=5890752

by mezsmi • 13 years ago

40 44 13 years ago

news.ycombinator.com

Computers. It used to be a strong academically controlled field. It was an area where you needed needed a PhD in Math to even touch the machines. That was then.

I want to get deeply involved in a field that was started by academics. It is dominated by academics. Guess what their minimum requirement is for a job?

I don't have a PhD. My academic credentials suck. However, the field fascinates me. It's getting to the point where I'm reading papers on the subject in my spare time. What realistic options do I have?

I want to build something that goes beyond NASA's Curiosity.

duked• 13 years ago

In my opinion there are two kind of jobs out there: 1- PhD required for real 2- PhD required but not really ;)

Let me explain what I mean for 1, PhD is required because they expect you to publish your work. And I have to say unless you did a PhD you might find it very hard to publish in top tiers conferences, understand the review process and of course be familiar with the state of the art (mandatory citations or your paper would suffer terrible reviews). Or they need the new hire to be a PI on some NSF/DARPA fundings and it's expected that the PI has a PhD.

Case 2, startups or big corp like to show off PhD in their work force just to impress customers/investors or they genuinely look for someone that can narrow down a problem, read about it and solve it. PhD have track record of that to get their degree but a talented engineer could achieve the same.

My personal advice, is stay away from jobs 1 as you probably will have a hard time fitting in. However you can definitely ace the interview for 2 and show what your capable of.

doorty• 13 years ago

You could try to get published or get similar field-specific attention by making some advancement in the field.

alanctgardner2• 13 years ago

If you really have a PhD level understanding of the topic, you could probably try to prove that in an interview. You have to seriously ask yourself, is that true? Reading papers in your spare time isn't dedicating years of your life to that topic. I can read CS Master's and PhD theses, and even give feedback, but that doesn't mean I could do research at that level. I'm working with a PhD in maths right now, and he is completely on another level; it's not just the technical knowledge to be able to read and understand papers, he's also immersed in the field to the extent that he has strong opinions and ideas about current research.

I guess my question is, if you love this field so much, why not do your PhD? Dedicate a whole bunch of time and effort to actually becoming an expert in this field, make a significant contribution by publishing papers. Then you'll know you're qualified, and you'll already have gotten to do the job.

On a side note, what field is this exactly? There's a big difference between wanting to do stats for machine learning, and wanting to write control systems for nuclear power plants.

Fomite5• 13 years ago

The first bit is important: What people imagine a PhD is, and what a PhD actually is, are often disjoint. "Really, really likes the subject, and reads a lot" is necessary but not sufficient.

marme• 13 years ago

the problem is at the majority of institutions you will not get the phd experience you are expecting. Most of phd work is doing what your professor tells you until he decides you have essentially apprenticed him long enough that you can graduate. If you really enjoy your subject you will become a master of the topic but just as many people go through the motions just enough to graduate but dont really care that much about their subject and end up being bad researchers

ylem2• 13 years ago

I have to comment on this--my advisor told me very early on not to become a technician--to understand what I was doing and why. At least in physics, the usual pattern is to give a student a project to work on--as they advance, they gain more independence. Sadly, some people never reach this stage...

rtpg• 13 years ago

think that depends on the field though. At least in CS, the experience seems to have been a lot more open.

skaevola2• 13 years ago

Applying.

coherentpony• 13 years ago

I'm sure OP already thought of that. Thanks, though, for your truly insightful suggestion. I'm sure you'll go far if you ever look for employment in a position that doesn't require you to state the obvious.

eli• 13 years ago

Snark aside, I think that is the 100% correct answer. It's unlikely the job actually requires a phd, they probably just need someone with skills equivalent to those of a phd holder.

skaevola2• 13 years ago

It's not obvious that applying for a job you're "underqualified" for is ever a good idea. But, it's a good way to start.

sjg0072• 13 years ago

Write a paper. Get it published. Easiest places are IEEE or ArVix.

Fomite5• 13 years ago

I'm not sure I'd really consider arXiv to be "published" in the peer-reviewed sense - it's a preprint service, and while that can attract attention, you need to get it in a genuine journal/conference (depending on your field as to which is the bigger deal).

Also, outside CS/Physics/Math, be aware that standards differ, and arXiv is often met with "...what?"

sjg0072• 13 years ago

Dude wants to build a space robot. That should be physics/cs/math.

Fomite5• 13 years ago

Publications may help - and as folks like duked mentioned, there are some jobs that are actually "We'd really rather prefer a PhD but we might be able to swing it for the right person".

But honestly, if they mean PhD required? Odds are the answer is get a PhD. It's not just that its an exposure to the field thing, though that's part of it. A PhD is equal parts credentializing, expertise building and a cultural thing. It's possible that they need a PhD for things like competing for grants, and even if they don't, it's likely going to be an odd cultural fit.

orokusaki2• 13 years ago

FWIW, I won't bother with anybody in my life that requires a degree for anything, even teaching. Having a degree for something is, IMHO, no better than having a JD Power & Associates award, which is not very valuable. Now, are there countless people with PhDs who are terrific and very intelligent? Yes. Are they terrific and very intelligent because they received degrees? No, and furthermore, they'll probably achieve less than their drop-out counterparts (e.g., Steve Jobs, Bill Gates) or their "bad student" counterparts (e.g., Albert Einstein). Even Napoleon Bonaparte dropped out of Berkley at age 22... just making sure you were paying attention :)

orokusaki2• 13 years ago

In other words, do what you love and do it well, and don't ever let "papers, please" get in your way.

ynd• 13 years ago

FYI, Einstein was not a bad student. (http://www.todayifoundout.com/index.php/2011/12/albert-einst...). It's just a myth.

Having a degree does not magically make you more qualified, it's the work you put it over the years to get your degree. The paper is useless, but the experience learned remains in your head.

joshguthrie• 13 years ago

Applying for the job is the first step to getting a "PhD required" job without a PhD. Then it's up to you to prove not having a PhD doesn't impede your skills in any way.

auctiontheory• 13 years ago

"Fascinates" is a terrific start, but it is not enough. What have you built or accomplished in the relevant area?

You seem to be focusing on academic credentials. (I realize that's what the job posting asks.) But at the end of the day, credentials are only proxies for ability to do the job. And the best way to prove ability is to demonstrate past success in similar or related projects.

Industry, even an industry apparently dominated by PhDs, is not academia.

eddyparkinson• 13 years ago

A PhD answers a question, it is a process for answering a question. Do you know the process already, do you know how to answer the question, do you understand the research process. Are you able to tick all of the todo list items that are used to do research.

How to Organize Your Thesis - describes the output of a PhD. If you already have created output that includes most of the items listed, then sure maybe you can skip the PhD bit, won't be the first time.

Bunch of PhD related links that I found invaluable: http://www.dmoz.org/Reference/Education/How_To_Study/Postgra...

mikegagnon• 13 years ago

Are you an expert in the field? If yes, then apply for the position(s). If you are rejected for lack of Ph.D., then move on. It's not good to work in an organization that values credentials over expertise.

If you are not an expert in the field, then become an expert. One way to become an expert is through a doctoral program. Other ways include self-study, apprenticeship, real-world work experience etc.

graycat• 13 years ago

Curiosity was a huge project with many quite separate parts with people with qualifications of enormous variety. Thus, Curiosity was not just some one thing. Also, going "beyond" Curiosity might be an objective to be considered by a committee reporting to the head of NASA but is not really a suitable goal for just one person.

One way to look at Curiosity is that it is just a platform for a collection of devices for making scientific measurements. Each such device was developed by a team, and no doubt the teams were largely independent. A single device team had experts in maybe geology, chemistry, optics, mechanical engineering, electronics. Likely the leading academic subject matter experts had Ph.D. degrees, but I have to doubt that usually more than 50% of a team did.

Once Curiosity was ready to be launched, there was the rocket, the ground stations that communicated with the rocket and its payload, lots of people back at JPL working on trajectory, guidance, data analysis, software updates, etc., lots of people with likely less than 10% of them with a Ph.D.

If you want to make a big splash in, say, 'autonomous vehicles', then making progress in some part of that field might be a suitable goal for one person. So, look at what has been done at Google, Stanford, CMU, etc. and funded by DARPA, etc. Also consider aviation from early autopilots to autonomous drones and what more that people want. Consider sea-based autonomous vehicles. And, of course, the hypersonic scram jets will have to be autonomous until they are big enough and trusted enough to carry a person, which stands to be a long time.

For sending humans to Mars, my approach, which you are welcome to borrow if you want, is first to do a lot with autonomous vehicles. So, before even the first human leaves earth, have dozens of autonomous vehicles on Mars, awash in redundancy, with a good camp set up and running, and able, reliably, to launch payloads back to earth. Then, almost as an afterthought, let a team of humans go, with appropriate cosmic ray shielding, etc. So, don't send any humans until apparently nearly all the risk is gone. So, do nearly all the work with autonomous vehicles first. Maybe there's some work there you'd like to do!

But, be careful: Even if autonomous vehicles are your real interest, you may find that mostly the qualifications needed are in mechanical engineering, aeronautical engineering, control system engineering, software engineering, electronic engineering, etc., and each of these is a more definite academic field. That is, even if 'autonomous vehicles' is a good goal for you, it may not be a very solid academic field for you to stand on to achieve your goal. To know more, just look at what there is.

On your past academic background, that might not mean very much. A Ph.D. is nearly all about just three things, research, research, and research, and nearly no one teaching in K-12 has even as much as a weak little hollow hoot of a tiny clue about research. Instead, in K-12, have a lot of babysitting where the teachers, nearly all women, want good little students, mostly the girls, to sit still, be nice, write neatly, be nice, be quiet, be nice, jump through little hoops, be nice, etc.

For me? In grades 1-8, all the teachers in the school agreed -- I was poor student. Apparently my standardized tests of talent said otherwise, but that didn't impress the teachers. So the teachers treated me like dirt, and I gave up on trying to please them. In the eighth grade, my handwriting just sucked (common for boys). My 'clerical accuracy' sucked -- it still does, so to get something detailed correct I have to do it one day, wait at least a day, better a week, and check it. Somehow that issue doesn't hurt my work in software; somehow the mistakes I make are ones a compiler easily catches; in all the code I've written over all the decades, I'm not sure that even once a clerical accuracy problem became an actual software bug problem. Since I have some actual talent in math, my understanding of the algorithms, etc. of eighth grade math was fast without doing the homework.

So, on tests, I didn't do very well: E.g., I didn't care enough to try very hard. I didn't even know why or how to try hard at academics. So, with my poor handwriting and poor clerical accuracy, when I had to, say, multiply two four digit numbers, in my intermediate work the columns would not line up and I would make simple errors.

So, at the end of the year, my eighth grade arithmetic teacher gave me a D and fervently advised me never to take another course in math.

My father was actually good in education, understood that actually I was learning enough, and laughed at the arithmetic teacher. Dad was correct. For the next four years, I was likely the second best math student in my grade. It's a good bet that since then I've been by a good margin the best math student from the school ever. The eighth grade arithmetic teacher knew nothing about math. Nearly none of my K-12 teachers knew anything important about academics. Likely none of them knew anything about research.

Doing well in K-8 or even K-12 is not a very good predictor of being good at research. Moreover, doing poorly in those grades doesn't mean much, either.

Don't let the K-12 system evaluate your potential for research or even academics. Why? Because for anything significant in either research or even just academics, nearly no one in K-12 has even as much as a weak little hollow hint of a tiny clue what the heck they are talking about.

For a view of some of the excitement of research and some of what in 'originality' is crucial, look at the YouTube lectures of Eric Lander on microbiology and genetics. E.g., for a course home page,

     http://ocw.mit.edu/courses/biology/7-01sc-fundamentals-of-biology-fall-2011/

Can download course text materials at

     http://ocw.mit.edu/courses/biology/7-01sc-fundamentals-of-biology-fall-2011/download-course-materials/

A TOC of the videos for the course are at

     http://www.youtube.com/playlist?list=PLF83B8D8C87426E44

See also Lander's

     http://www.princeton.edu/WebMedia/flash/lectures/20100419_publect_lander.shtml

Not all academic research is that exciting, but Lander's emphasis on the excitement and crucial role of originality is right on target quite broadly across STEM fields.

For more, look at the background and work of, say, Craig Venter. One little thing he did was take all the ideas and planning of the Human Genome project, trash and junk them, use a radically different approach, and totally knock the socks off all the NIH team. Except for Venter, the genome project might be looking to be done maybe in year 2200! As I recall, a Venter remark was that the NIH team was not looking to sequence the human genome but to set themselves up with permanent jobs!

If you want to get a Ph.D. in a STEM field, might take a fast read of, say,

     https://news.ycombinator.com/item?id=5849936

     https://news.ycombinator.com/item?id=5849938

akiselev• 13 years ago

+1 to all of this. I don't know of any easy solutions to this but I do have some fun long term ideas to add on here. The better you are at networking and seeing and grabbing opportunities, the better your outcome will be. If you can show that you learn very quickly and know what you're doing where it matters (and are clear when you don't), then the lack of a Ph.D. will only limit your choice of employers, not fields. This sounds corny and obvious but you always have to actively keep the thought in your mind.

Look into NASA space center/R&D internships and the like (JPL, Goddard, etc.) if you're interested in aerospace. Also look around for local universities with strong aerospace programs with project based graduate classes. If you're near a top aerospace program (Caltech, MIT, Cal Poly Pomona, Embry-Riddle, etc.) those classes can land you free or even paid work on satellite projects that get launched. As a concrete example: I took Aero105 class at Caltech while still a senior in high school and got to work on aRResT, a Caltech-University of Surrey formation satellite tech demo that should get launched this year or next. If you can get security clearance with NASA then getting in through a university that works with them is relatively easy (sadly the clearance is what got me).

I'm assuming that if you're asking this question, you already have quite a bit of experience in your desired field so I'd like to emphasize universities. You can't just walk into a company and volunteer to work on their projects. Best case scenario you have to apply for an internship because the company can't legally accept free labor, fully trust you when you have nothing invested, etc. Universities, on the other hand, are free game. You'll want to do some due diligence and try to find labs that work extensively with industry (again, Universities with emphases on project based courses are more likely to have such opportunities, i.e. Franklin W. Olin College). Once you find a few labs, try to find some way to contribute to the project without getting in the way of the researchers. Try to find a github library someone in the lab wrote and contribute to it, clean it up, etc. or try to make a hardware project related to what that lab is doing. If you approach the lab with something that shows you understand what they're working on and could actually be a resource instead of a hindrance, the door is wide open. Note there are other, less expensive of getting your foot in the door but this tactic has worked best for me.

I've found a lot non-profits to be similar to corporations. Scared of liability when it comes to engineers volunteering. Unless they go out to developing countries to build stuff like Engineers without Borders, the nonprofits usually only have restrictive lab internships and the like. It's also much harder to get from "foot in the door" to paying salary. Since many professors are largely funded by grants and have tenure, they are for the most part the King in their own little world and can tolerate you hanging around the lab till they can pay you or pass you along to someone that has money for you.

If you're interested in autonomous vehicles in our atmosphere, play around with quadcopters, UAVs, and then try to make your own small, light weight and maneuverable hummingbird robot. Control theory is very approachable if you have some experience with math and computer science. Be forewarned though: I know brilliant people who have spent years at university making control software, it is a deep deep field. On the upper hand, if you manage to write software that can control a small flying robot as well as a hummingbird can fly, you'd win the hypothetical Nobel Prize in Robotics. If there are competitions that don't restrict entrants to universities, join and try to make an impressive show.

You want to have something to show to everyone that screams "I KNOW WHAT I'M DOING" because that's pretty much what everyone thinks a PhD is (ahem: "I was paid less than minimum wage for half a decade or more"). In most of my interactions with academics, I've found that the PhD club is quite like any other tribe, arrogant and full of itself. But don't worry, those doctors are _mostly_ harmless and use the almighty PhD certificate as a defense mechanism and risk management tool. They're rational people who know determination and intellect when they see it.

rdouble• 13 years ago

If you have a specific position in mind, I would suggest trying to meet with the PhDs. They often have work that needs to be done, but are unable to do it themselves. I was interested in molecular biology about 10 years ago and was able to get jobs working alongside PhDs in a lab environment because they needed a programming lackey.

lanna• 13 years ago

Step 1: Get a PhD.

Step 2: ?

Step 3: Profit!

davimack• 13 years ago

Firstly, is this about the tactical objective, or what you'd do if you managed to get there?

If it's just the tactical objective, you'd need to find someone on the inside who would be willing and able to convince everybody else that you were the person for the job - and that includes the HR end of things as well. It would be easier, actually, for that person to sell a new position than to get you into a position which has the requirement of a PhD.

If it's the long-term you're considering, though, some of the other posters have pointed it out: you don't acquire the skills simply by reading about the subject (although reading about the subject is an immense part of the PhD). You must also contribute something to the subject, in a particularly rigorous manner - that creation, in other words, will be measured not simply by its outcome but by the process used and the documentation of that process produced.

The PhD isn't about understanding a field (that's a Master's). Rather, it's about understanding how to conduct research in a particular field, on top of having that mastery of the field. What you're seeing when you read papers is only the content of the field, not the process by which that was generated.

You have essentially said, "I've looked at a lot of paintings, and I really like paintings, and I know about different styles of paintings" ... but you have not painted.

socialderp• 13 years ago

Academia has VERY different rules then the real world. Deal with it.

This whole attitude of "I'm better cause I say so hurrrrr" and "I'm gonna (keyword, aka haven't) built some amazing" will NOT cut it. Getting a PhD is no joke, takes a lot of commitment (years) and is usually very difficult in a science based field.

I await the downvotes from the fealgood social gurus who themselves do not have a degree in cs, which I do...

frou_dh• 13 years ago

Candidates for an automatic downvote:

    * Downvote preemption
    * "This."
    * Article font/background griping
    * ...

angersock• 13 years ago

I'll be a bit coarse here.

"I want to build something that goes beyond NASA's Curiosity."

Ain't nobody stopping you--build the thing. Metal doesn't care whether its machinist is a doctorate or a dullard, it responds the same way to the bite of a tool.

"However, the field fascinates me. It's getting to the point where I'm reading papers on the subject in my spare time."

There's a world of difference between being fascinated/lusting after something, and actually taking steps that will culminate in the achievement of goals. Idly reading papers and things is well and good, but you need to start actually designing things, hacking on little projects, and doing something other than intellectual voyeurism. Real artists ship, my friend, and do so without thought of whether they are qualified to do so or not.

To your beginning point, near any technical field nowadays was started by academics--and that doesn't matter a hoot in hell. The biggest advances in electronics and power generation were made by people screwing around in their free time (Edison, Marconi, etc.). The best driving work in video games was done by people that never finished college (Carmack).

Get out there and do something, credentials be damned!

At the end of the day, most successful companies are started and staffed by a bunch of people whose works finally caught up with their promises.

pcvarmint• 13 years ago

Work for 15+ years in high-performance computing, have 3+ references, and give a nice but concise (45-minute) technical presentation.

That's how I got a job at a national lab, after being a college dropout, but self-taught and still motivated. Once you get in, Ph.D. or other labels matter less than results.

Pitarou• 13 years ago

It's getting to the point where I'm reading papers on the subject in my spare time.

Sorry to rain on your parade, but you've still got a way to go. You're an occasional jogger. A PhD is a marathon.

Which isn't to say that you shouldn't go for it, but its a hard road.

An academic paper needs:

- solid background research (think "literature review")

- a bright idea

- implementation

Guess which are the hard parts. If you can do those then go ahead, write some papers, and publish your work. Given your outsider status, people will forgive you for not publishing in a peer-reviewed journal, but you need to publish equivalent material on a website.

But would it be so much harder to just do a PhD?

JoshTriplett• 13 years ago

Exactly the same way you'd get a job that requires a B.S. when you don't have one: prove that you're already working at that level without the credential. A Ph.D. purports to demonstrate that you can do original research and effectively communicate that research. So, successfully publish papers in the peer-reviewed conferences of your target field.

infectoid• 13 years ago

Probably one of the more succinct and constructive comments here so far.

Could not agree more. In short, follow your passion and the rest will fall into place.

Here is a previous HN post that helps to illustrate the point... https://medium.com/this-happened-to-me/8f381aa6bd5e

ylem2• 13 years ago

I've seen cases of people without phDs at national labs who are doing research--but it's rare. I would suggest that you figure out who the actual researchers are who are involved with the project and contact them (personally, I would send and email and then follow it up with a phone call) and see if you could chat with them about the project and where you might fit in. It might be possible that there is a position that you might be able to step in with your current qualifications and see how you like it and if you want to pursue a phD while you work there. Do you live reasonably close to the facility? Would you be able to stop in during your free time?

zura• 13 years ago

Similar but more specific question would be - How do you get a "Postdoc" job without a PhD?

pmb• 13 years ago

Answer to the postdoc one: You can't.

People looking for postdocs almost always have a specific sort of funding and it is attached to the credentialing.

Fomite5• 13 years ago

Yeah, I'm going to go with "You can't" for this one - "Postdoc" is inherently post-PhD, and the expectations of the position will pretty much require one. For example, many require establishing yourself as an independent researcher - which means grant writing, which is next to impossible without a PhD.

nichols• 13 years ago

Try challenging your interviewer to a dance-off. If you win, they're legally required to give you the job.

ISL• 13 years ago

One of the cool things about getting a PhD is that you can get to work on the cool problem you're interested in.

If you're working hard and get lucky in your research, a PhD can take < 5 years. Masters programs can be easier to get into, and if you're good, you'll have an easy shot at getting into a PhD program. Not saying that a PhD is necessarily the best choice, but it's a way in.

Some friends worked on the PanCam rover hardware as undergrads and parlayed it into 'permanent' positions with rover operations without going to grad school. Get in touch with the people who do whatever you want to do, make it clear that you want to work for them, and do everything to get competent in the field.

blueprint• 13 years ago

PhD's are mostly devoid of value and often harmful to people's judgement in and of themselves. The certification means nothing, implicitly. If someone's PhD work is not contributive to society (I mean, let's look at quality of life in society here rather than number of papers produced) then the PhD is not a virtue of that person. On the contrary, it becomes an object of deception.

Anyone that doesn't realize this fact is cowed by an entity or knowledge-base that you probably don't want to end up slave to. If they are requiring you to have a PhD despite your ability then you may like to look into finding people like yourself and working outside of the existing employment system you reference.

goodcanadian2• 13 years ago

I have done this . . . sort of. The job ad did not explicitly require a PhD, but it was the sort of position where one would be expected (notionally, a postdoc). If the position is meant to be research, you are almost certainly going to need a PhD and a record of publication. Seriously, a PhD by itself isn't even worth much any more. However, there are plenty of job ads that ask for a PhD which are not research and do not really require a PhD. What I did was to contact the decision maker (sometimes it is easy to find out who that is, sometimes not), and then talk to him. You can ask if he is willing to consider someone without a PhD, and then give brief account as to why your skill set or experience is appropriate for the job. Keep in mind, this usually doesn't work. You may have misunderstood what the job entails, or the decision maker may simply be unwilling to consider someone without a PhD (rightly or wrongly), or there may be a whole line of better candidates who all have PhDs.

You mention building "something that goes beyond NASA's Curiosity." This is a field that interests me, too. Most of the positions that actually build things are engineering or software. Most of them do not require PhDs, explicitly or otherwise. So, in short, I am not sure what position you are trying to get. If it involves building spacecraft, and asks for a PhD, it is almost certainly research, and it almost certainly really does require a PhD.

Ask HN: How do you get a “PhD required” job without a PhD?

Related Stories