So, there this guy, Matt Welsh. And he left Harvard to go to Google. OK.
Now he's baaaack, to point out that academia isn't that rosy.
Yep. He's not wrong.
Were I in a sarcastic mood, I would say something like "ohmigod, Matt Welsh is pointing out that academia is a social endeavor that is constrained by funds!!" and proceed to point out that you are, frankly, rather unworldly if you don't understand that this is true of just about everything. In fact, unless you are not constrained by funds (because you are, say, independently wealthy), you should probably expect to have to engage in convincing other people that you are doing interesting work and they should fund you, not others. Cue politicking.
(Of course, grad students are almost by definition unworldly. But.)
Now, I did spend some time, unsuccessfully, in startupland. It turns out that I am uncomfortable bullshitting about commercial viability. I am far more comfortable in the academic world, where people will fund me based largely on whether or not I can convince them my work is interesting, not whether it will make money in the real world. And I've been more successful as an open source hacker and as a professor than as an entrepeneur -- remember, kids, if you set the bar low enough, even you can be successful :).
So I think I know a little bit of where Matt Welsh is coming from, and I will close out this digression by pointing out that someone who has the skillz to become a tenured faculty member at Harvard is probably going to do well in many intellectual endeavors. Good for him, but take his (and my) advice with a grain of salt -- credentials do open doors, for better or for worse, and advice from a former tenured prof at Harvard (him) or a Caltech PhD (me) may not translate. (This is obvious, right?)
On to the advice
If you are a gradate student,
and whether or not you are definitely planning to go into academia, or toying with going into biotech or a startup, or completely uncertain,
write a professional blog.
But seriously, it gives someone something to look at beyond your (no doubt very limited) CV/resume.
I have recently been engaged in cold-e-mailing industry people on behalf of friends-with-PhDs (FWPs), and invariably I wish to send out the blog, Twitter account, and github account for the student in question. Only rarely can I do this, because most grad students are stuck in an academic silo and only poke their head out when they realize they need to find a job. By then it's too late.
Or, to put it another way: invariably, the advice I wish I had given these students two years ago is this: create an online presence for yourself well before you're looking for a job.
What should I blog about?
Whatever you find interesting. My blog is mostly professional, with some satire and occasional sci-fi/fantasy references. The point is, it's me.
But, what should I say?
Something not obviously stupid, racist, or sexist. (Hopefully this, too is obvious, if only in hindsight.)
What if no one reads my blog?
That's OK. It can take a while even if you're saying fantastically interesting things. The important thing is to have some sort of online presence showing that you are an awake, interested, and occasionally articulate person.
But what if no one likes what I have to say?
What do you care what other people think? Seriously, see "articulate, awake" point above. That's more important than being particularly likable.
What if I suck at blogging?
Practice makes better, if not perfect. Improve thyself.
Could I tweet instead of blog?
Absolutely. The important thing is to get out there. I think blogs are more permanent, though.
Is it a competitive advantage if everyone is doing it?
Good question. Maybe not, but if you're the only one not doing it, then it's definitely a disadvantage, eh?
Are you, a tenure-track professor at an R1, seriously telling my students to waste their time on blogging instead of generating first-author pubs for themselves?
Yep. Dear professor, at this point you need to look at yourself in the mirror and admit that there is more to life than a tenure track career, and moreover that you cannot guarantee your graduate students such a career. It would be a good idea to start helping them prepare for something else.
Besides, you can put their blogs down as evidence of your outreach component for your NSF grants. Yes, this works.
Why should I listen to you? You're not at Harvard!
Neither is Matt Welsh.
My advice really boils down to one thing: build your online presence so that you can point prospective employers at it. This will help when employers want to know something about you; I think you'd be surprised at how difficult it is to tell good applicants from bad applicants via only a CV or interview, so every bit helps.
(Also, if you're a programmer, I'd strongly suggest working on open source projects in your spare time. But not everyone's a programmer.)