How to leave academia (for science PhDs)

I recently did a blog post about a similar issue, the tension between doing software work in academia vs in industry. In a way a kind of prequel to this post. http://elazungu.wordpress.com/2012/03/26/the-researcher-prog...

I recently left my PhD program too (I'm "mastering out" in local parlance ;)) and what this post doesn't address is what I found to be the hardest part: making a decision to leave a world where people that leave are construed as a failures (note: contrary to my expectation, I've only received support in my decision from my awesome and anomalous department.) (Also, I should note that I left a software industry job to do the PhD so I knew I shouldn't have difficulty finding a job.)

For anyone who wants to take the leap but is afraid or unsure, I offer some words that were incredibly helpful for me from some of fantastic friends. To quote my amazing advisor from his response to my "I'm leaving" email:

'We had a Head of Department at Lancaster who used to stomp around the corridors moaning - "I've just lost another student to industry. He's got a great job, has a starting salary bigger than mine, is working on a fabulous project with better resources than we have. In what mad world is this judged as a failure?"'

And another colleague, who's currently a junior professor: 'You know, most Ph.D.ers are smart and successful people. Hence they have a difficulty in saying “This is not for me”. They instead say “I’ve been successful all my life, and I finished everything I started, so I should finish this as well”. By saying that, they choose to hang in there for many years in a depressed state.

Sometimes, the most courageous thing and the best thing to do is to quit when you know you would rather work in another capacity, or when you know you don’t want to work in academia. I congratulate you on your decision and I hope the best for you. ( In case you later decide to come back to academia, it will be waiting for you, so I would not worry about it.)'

Best of luck, those of you that are struggling with the decision. If you're anything like me, if you decide to leave you'll feel better than you have in years =)

"to leave a world where people that leave are construed as a failures"

That was very much the view when I left - is it a view that is particularly common in the UK?

The post is awesome. Having just "graduated" (dropped out of my PhD program) and also interviewed a bunch of really smart folks who are making the transition to the "real world", I cannot emphasize the importance of the following for a good machine learning/engineering interview:

1. It doesn't matter if you have done research in a topic, do brush up on simple machine learning algorithms. The math after going through a PhD should be really easy for you. However, it is hard for an interviewer to gauge how good your math background is when you have done most of your quantitative work in a totally different field and you don't have enough of a common intersection.

2. Write code! If you are joining a small to middling start up, even if you will eventually be doing quantitative stuff, your peers will be people from engineering and they will evaluate you as a programmer. Brushing through something like the Algorithm Design Manual is going to be amazingly useful. Back in school, there was a linguistics professor who wrote most of his code in Scala, used version control and pivotal tracker. If someone who is working towards tenure can do it, there is no excuse for you. :-)

3. Try to work on tiny projects that involve some data analysis and some programming. If you are an R guy, do it in python and vice versa. Most importantly, think of the whole pipeline: taking the data (Infochimps/ AWS public datasets), cleaning it up, processing it using a random machine learning algorithm (pick any, its an easy and awesome learning opportunity), and then visualizing it. Repeat with cooler tools (E.g. d3.js for the visualization etc). The importance of this process is that it moves you away from the traditional anonymized datasets that you see in academia and helps you encounter real world data and gain intuition for that. This is incredibly helpful in the interview process because it allows you to get good ideas about other people's data and how to think about it.

I received my PhD (Physics) in 1999 and went directly to industry. At that time I saw industry as a well paid post doc. My intention was to work a few years and then seek an academic position.

In 2001 I had an offer from a top engineering school in Canada to join their faculty. And I said no.

Academia attracted me with the promise of intellectual freedom -- the ability to work on what ever problems I wanted to. This is a lie. To be successful in Academia, one needs to pick a narrow field and become the biggest expert in the smallest space. Neils Bohr described this as knowing more and more about less an less until you know absolutely everything about nothing.

In Industry I can move around to whatever problems fascinate me most. I've worked in Optical Componetns, designing video games, and now what I call computational sociology on social networks. My pattern has been to work in a field of about five years until I've built up the expertise I desire to have, then move on to a completely different field.

For those recently dropping out of academia or considering dropping out, you know in your heart it is the right choice. Don't worry about what people in that little club think. A decade from now they will be looking at you. With envy because you are free to work on the most exciting and interesting problems while they are stuck in the same shrinking field for the last ten years.