Post-Docing in Computer Science

Ambivalent Academic asked about what Post-docing in my field is like, how people are viewed if they do or do not do a post-doc (talk about an awkward sentence!). This is my personal theory.

In my area (computer science) there is a very very strong industrial presence, obviously. All of Silicon Valley, and the tech sector are all ready and willing to suck talent out of the PhD pool. And pay them well. Regardless of whether they have done a post-doc or not. In fact, Bill Gates doesn't even have a Bachelor's, as everyone well knows, and the founders of Google left their PhD programs to start the company. And obviously, no one cares. So basically, if they want you, these companies will hire you regardless of whether you have certain letters after your names.

For this reason, I believe post-docs are really not required or done in my field because it's hard for people to resist the paycheck that could come from industry. If a company is willing to offer you over $100K out of your PhD, then why in the world would you go get a post-doc? Thus, academic departments would be hard pressed to require people to get post-docs in order to hire them, otherwise everyone would just go to industry. It's bad enough that academics get paid much less than their industrial counterparts, if you forced them to take 2-3 more years of getting paid very little, you'd have a mutiny and suddenly there would be no computer science professors.

At the same time, there is a lot less "methodological" training required in computer science. I don't know if that's why post-docs are important for other fields, I'm not really sure why you need to have them except maybe to get exposure to another lab/sub-field? I don't know, since I don't know anyone who has done one. But I imagine that part of the reason is to get broader methodological training. Which isn't required in computer science. Basically, you can code or you can't code. There are no "techniques" or "equipment" that you couldn't figure out quickly anyway. Yes, there might be some slightly different software tools, or maybe these machines are Suns and these are Opterons and those are Xeons, but really, it doesn't matter. So to spend 2 more years training would be pretty pointless.

So yes, some people do post-docs but for entirely person reasons - in my field no one would bat an eyelash if you didn't do one and became a professor. In fact, they might bat an eyelash and ask "why?" if you did do one.

Thus, since I thinking about the possibility of one, for personal reasons, I have to go through "connection" channels rather than advertisements, because no one advertises. We'll see how it goes.