How to build a better PhD | Hacker News Reader

How to build a better PhD

nature.com

41 pointstokenadult17 comments

The problem is simple, and easily solved: fund science at greater levels.

I trust everyone knows this story - the funding rate has dropped dramatically in the last 15 years, and as a result the average age for an RO1 has gone up steadily (see [here](http://nexus.od.nih.gov/all/wp-content/uploads/2012/02/age-o...).

Just give scientists more money, and a lot of these problems will vanish. There will be more academic positions; there will be less competition for grants.

This is entirely a problem of money. I have a PhD and have somehow managed to shoehorn myself into academia without pursuing a tenure track. There is a tremendous scope for people like me. I could hire three people today and put them to serious full-time work on difficult projects that would make a substantial impact in my field (cancer research) - if I had the resources.

It's easy to fit people into non-tenure track positions in academia, there is no structural issue. The reason people go to industry is only because there is no room made for them on the public dole. If the public steps up and funds more science, this problem will abate.

This is a completely layman's analysis, but I think that science funding has gone down because we are less and less able to understand what it is that is being studied and how it can benefit humanity. Go back 40 or 50 years and the things scientists were discovering were tangible and real. They produced things that people could touch, consume, and understand.

The biggest science project we have going, the LHC, is impossible for the lay man to understand. "A big circle which smashes tiny things together? You're just watching it? I don't get it!". Compare that to the discovery of penicillin; "You're growing fungus under your fridge and it's going to cure my sick child? Take all my money!".

That is because this generation has lot of diversions. Focus is the key for research.

Focus is easy when there's a strong value proposition. We got nuclear done very quickly because the value was near infinity. The science being done today seems to have weak value.

I agree.

I work at a large national research institute, 50 % of our budget is financed through grants. Just last week we were told by my institute's director (and I'm rephrasing only slightly): "If you really want to help our finances, do something that makes it into the evening news. Then it's easy for politicians to send money our way, they love to finance things that appear on the evening news. Because that is their currency for the next election. But if we don't succeed at it, it will be even harder in the future."

When these articles about people leaving academia after doing a PhD, I'm utterly baffled.

I thought it was completely normal and expected that people would do a PhD and then go into industry or do something else entirely? Where do they think all the industry researchers, school science teachers, etc come from? Where do they think all the banks were getting their PhDs from? I don't think many of my peers when I started my PhD were set on an academic career.

But these articles talk about it like it's some bizarre unexpected thing, and they seem to say that when people leave for academia they are never heard from again. This article literally illustrates them as floating off into the sky http://www.nature.com/news/life-outside-the-lab-the-ones-who...

Do they actually want to limit the number of PhDs to the number of academic positions? What will industry do in that case? Where will schools get science teachers?

I have never seen the picture they try to paint of fool-hardy young students all thinking they'll get a faculty position.

People often go into PhD, like I did, to do research and contribute to the public good. Most people like me dreamed of an academic job with tenure. Tenure so that you can actually do radical work without retribution.

That environment no longer exists. Tenure is highly fought for and there are very few positions. And even if you get tenure, you really don't have the luxury to work on anything you wish because of funding. A good portion of your days as an academic consist of teaching, another good portion is writing grants. If you don't do the popular thing, you don't get grants. If you don't get grants you don't get students.

And sadly, without students that means that you don't get to do much research because the majority of your time is teaching and dealing with grants and other bureaucracy. As my former supervisor put it: you get time to do research yourself when you get home after work.

So you have time, but not much. Most profs use this time, but it's just not enough without grants and students.

Ok, so let's go to industry. You get paid a lot more. You get to do primarily research. But it's not public. It's mostly locked up and directed to ultimately lame purposes like targeting ads or making trades. These help no one and don't help society, so it's really not very enticing to an academic hopeful.

And yes, these researchers do fall off the radar because they stop publishing. They may file patents, but the only people who read those are lawyers. This is because patents are not papers. Their primary purpose is law, and they don't contain any of the information an academic would be interested in. You generally can't reproduce a work from a patent, and patents don't include any requirements of proving your invention works, which is the primary requirement of science.

To make matters worse, the patent office typically patents extremely simple things that aren't even worth the attention of someone trying to achieve state of the art. Essentially patents are a huge joke, so nobody looks at them, either in academia nor industry engineers.

TLDR: people go into academia to make the world a better place and do novel research that gets published to the world for free. Academia is increasingly a bad way to achieve this and industry doesn't achieve this at all. Though industry does get you paid so ends up being the fall back, even if you fall off the map.

Industry researchers don't do things in public and don't publish? Lame purposes and helps no one?

Again I just can't relate any of that to what I know at all!

When I finished my PhD I moved into industry and basically kept working on the same project. I work for a megacorp, and almost everything I do is developed fully in the open (public GitHub repo, not a mirror), we publish several times a year at the same venues I did in academia, and everyone is free to benefit from what I work on from using the code, reading the code, reading the papers etc.

And this isn't unusual. I know many many people in the academic community who do almost exactly the same thing. Look at everyone in Microsoft Research, Oracle Labs, IBM TJ Watson, etc.

There are a few industry options for the sciences but less so for the humanities, where tenure is intended to protect scholars who investigate and question things that the wealthy and powerful might find dangerous to their interests (e.g. Marxism).

Also, it sounds like you are talking about (software) engineering, rather than basic science (e.g., astronomy). There are plenty of questions humanity would like for science to answer for which there is no financial incentive to a corporation.

Most people like me dreamed of an academic job with tenure.

Not in my experience. It might be true in subjects like english, history and the social sciences where jobs in industry are rare, but and certainly not in the technical engineering fields. Most people I know getting a PhD did so with the intention of moving on to the private sector.

And I don't know what industry you are in, but lots of companies publish papers and present their research at conferences.

Why not work part-time in industry (it pays well), and do science in the remaining time?

Because I want my work to be meaningful, and I don't want to be a shill. I want my contributions to go to mankind (the model of science) not some rich investor (the model of industry). Plus the work in industry is almost universally boring compared to the work in academia.

I got my PhD in physics in the early 90s. Already by that time, it was widely known among students that we weren't all going to get tenure track jobs. We understood the numbers. It was called the "birth control problem."

I told my dad about it. He's a PhD scientist, and spent his career in industry. He told me that the "birth control problem" was already known in the 1950s, and that it was never a slam dunk that a PhD would get into a decent academic research career.

A couple comments.

There are pervasive claims both by prominent scientists and politicians in both parties that there is shortage of scientists (Google STEM shortage) going back at least to the 1950's. Many people even to this day starting a Ph.D. program are unaware of the long odds against getting a permanent research job.

Most Ph.D.'s in STEM (Science, Technology, Engineering, and Mathematics) fields who leave academia find jobs as some sort of software developer, an often high paying but highly unstable and often short lived career. There are not a lot of jobs for scientific research in private industry; many research labs such as Bell Labs or Xerox PARC established in the post Sputnik 1960's period have been closed or substantially shrunk. Microsoft Research and similar outfits are not nearly as large as the Bell Labs, HP Labs, Xerox PARC of the past.

The vast majority of STEM Ph.D. students are funded through grants and contracts from funding agencies such as the National Institutes of Health (NIH) which by policy chooses to fund far more Ph.D. graduate students than there are long term positions. In most cases, the funding agencies promote the false idea that there is a current or future shortage of scientists. The funding agencies could by policy limit the number of funded graduate students to some reasonable multiplier of retirements of senior researchers each year, taking into account that some graduate students will reasonably decide to pursue other careers after completing their Ph.D.

The extremely long odds of a permanent position, most of which pay no more than a software developer with a few years of experience or an associate at a top tier corporate law firm ensure that a high fraction of the "best and brightest" will not pursue a career in research, tackling problems like curing cancer or better energy sources. The present system hopes that quantity and regimentation -- an army of easily abused indentured servants from India and other "third world" nations -- will win over quality and individual creativity. In many fields, like cancer research, over forty years of this approach has failed unequivocally.

As the article notes in its introduction, too many PhD students misunderstand where their PhD will lead. That's a clear problem, and robs many of our brighter kids of years spent in prep for jobs that don't exist. Doing that deliberately is malignant and should end.

How to correct the problem? Start by requiring universities to inform incoming students AT ALL LEVELS of two things:

1) Given the forthcoming degree, what jobs await them? Report job trends for their graduates in the past 20 years, including what job titles they take at what employers.

2) What will this education cost them? Report the range of student debt for students over that same 20 years, including repayment horizons.

Much needs to change in academia, and it starts with a full accounting of the cost and market value of a degree, especially when a significant market value is expected.

(It's worth noting that nowhere would the benefit of doing this be greater than countering the for-profit degree mills that abound in recent years. But in the end, the damage done by teaching useless skills at a good university is little different than failing to teach useful ones at a bad university.)

>“And a master's is not considered a failure for those who can't make it to a PhD.”

I was in a physics PhD program straight out of undergraduate school. In that program, at least, the culture very much was that the master's degree was the "failure option".

I ended up leaving after one year for personal reasons, and I felt like even more of a failure because I couldn't even get to the "failure option" point. Five years later and even though I'm relatively successful working as a programmer it still bothers me.

Reply on news.ycombinator.com