Jamie's Blog

Lessons from a life of startups, coding, countryside, and kids

Post-MLE: A PhD in Southampton

So I guess I’d better quit alluding to it and declare my intentions to the world: I’ve accepted a PhD at Southampton University which I’ll start in January.

It was a really tough decision to make. On the one hand, I have the things that I’m beginning to want from life: money, a house (or at least the ability to buy one if I want to), a pension and generally a stable, more settled, life. But I’d also like to do something interesting during the day. Over the course of the last 5 years since I left uni I have reached various stages of applying for a postgraduate course but the money factor always got in the way. A quote from my undergrad tutor continues to haunt me: “if you get a real job you’ll never come back to university”. And it’s true, once you’ve tasted the fine food and pleasant life a salaried job can provide it doesn’t seem likely that you’ll opt for a meagre university handout. Nor will the huge debts you’ll accumulate inspire you to go back to college.

So why am I doing it? Well, that /. discussion I mentioned yesterday was basically drawing the following conclusions:

  • A PhD is theoretical and you won’t have any practical experience of the “real-world”
  • If you do a PhD you’ll have priced yourself out of the monkey-boy coder jobs
  • A PhD does nothing for your career prospects in the computer industry (in contrast to career opportunities for PhDs in the sciences)
  • A PhD is a waste of 4-8 years of your life
These are all pretty good points if you are a recent graduate thinking of doing a PhD, but I’m not. I don’t think I’d recommend anyone doing a PhD straight after their BSc, but it is also worth bearing in mind my old tutors quote above.

I’m not a recent grad. I have 5 years of commercial R&D experience and I’ve never considered myself a computer scientist in the academic sense. I’m a research engineer - I like building the things that researchers build but I like building it the way a software engineer would. That means I use version control, build scripts, unit testing (where possible), my code is well-documented and there are even design documents. Of course the difference is that I’m usually working on my own and there is no specific end-user to be interrogated interviewed. In my opinion, far too many projects building the latest Really Cool Thing™ never make it out of the lab, or even within the lab because they are poorly implemented. Other researchers can’t figure out how to use the code and so prefer to reinvent rather than reuse.

There is some confusion about the relationship between a post-doc and the salary they’d expect. If I finally become Dr. Jamie and apply for a monkey-boy coder job then you can be sure that I’m smart enough to have a) read the job description, b) read the potential salary scale and c) decided that I want that job for that wage.

As for the other points, the PhD I undertake will not be theoretical (well, I might have to throw some in) but I learn by doing and I do by building and I build by first imagining. And I’m giving myself 3 years to complete the thing with absolutely no intention of using it within an academic context. I won’t do the postdoc circuit, I won’t lecture, and I won’t be aiming for a tenure track position.

Today, I have two choices; get a “real” job or do a PhD. If I was to take a real job then I would never go back for a PhD. This is the last-chance saloon. After the PhD I’ll have three choices; start my own company based upon my research, get a research job in a commercial lab or get a “real” job. At least that’s an improvement on my choices today.

There are other reasons that I’d like to spend 3 more years focused on my research ideas. MLE has not been a productive environment for me. In the 2 years I’ve been here I haven’t managed to achieve the vision I arrived with - which is very disappointing. I won’t go into the reasons now (maybe some other time) but what’s done is done. Unfortunately what is not done are some of my ideas in the intersection between agents, p2p, semantic web and pervasive computing. I need to get these ideas off my chest.

Southampton is the only place I could consider doing this PhD around in UK/Ireland. The group has a long track record of producing interesting and useful research and it contains some of the brightest people in the field. And I’m sick of working for morons.

Oh, and did I mention that there aren’t any interesting jobs out there anymore? Particularly not in Ireland. At least with the PhD I know that there’s some plan for January, after the trip. Some plan is better than no plan I guess. And in 3 years time I should have alternatives to the monkey-boy coder jobs - after all, “You can’t be king of the world, If you’re slave to the grind”.

I obviously have some concerns. Money is one. Being in England again is another - I imagine that I’m going to miss being able to take the train down to Inch and away from real life. Hilary is another - should I move over there by myself or would she be able to get a job there? Will I want to become a hermit in a remote part of NZ after the trip? Maybe but we can’t guarantee the future, we can only make a prediction with the information we have available.

I’m starting a PhD at Southampton University in January.