The Trade-off Between Useful and Interesting

I don’t think I’m cut out for graduate school.

I say this as someone who just spent half an hour hiding in the bathroom in tears, so take what follows with a pinch of salt. I’m not giving up quite yet.

Today, my supervisor told me my work made no contribution. I admit, that I haven’t exactly defined in my own head what my contribution is, but given the interest in my work from a couple of companies, the people I talk to, the requests for talks, and the traffic stats for my website I thought it was clear that I’d made something that was potentially useful. I thought that would imply a contribution.

Perhaps the reality is that I have not made a contribution to computer science. I’ve taken stuff that already existed, and arranged it in a different way. My innovation lies in the combination, not the creation, of technology.

However, if we look at the people who are innovating they’re innovating at the edge. Facebook innovates in technology, but their greatest innovations are in the social aspect. Apple innovates where hardware meets great design (fascinating post on the design of the iPhone and the upcoming Apple Tablet). Google innovates where technology meets utility. Google docs, for example, innovates in Javascript but the biggest game changer is being able to work on your documents online, from anywhere, with anyone. Whilst Google came from academic research, the thing that made them the big player they are in technology today was AdSense, not the content of the seminal paper that started the search engine. Google is a great example of what I’m talking about. Everything they innovate seems to bring some new innovation that’s not just technical. Even speeding up searches – whilst I don’t notice that my single search takes less time, the search API is infrastructure underling things like Google Squared, so I will notice when a square takes much less time to generate.

The people whose work I admire most are working at the intersection of tech and art (Sep Kamvar, Jonathan Harris, Gilad  Lotan), or organization in creating tech (Joel on software). The academic work I admire is coming from places where technology enables, but are not necessarily technology-focused, like danah boyd at Microsoft Research, Clay Shirky (NYU prof and author of “Here Comes Everybody” – Amazon) and the MIT Media Lab (I particularly like Mycrocosm).

This  disconnect between real world and university is frustrating me. In design, there’s the ideal that something can be both beautiful and functional (see Don Norman’s TED talk and his post about teapots). In creating software, my ideal is something that is both interesting and useful. As a compromise, I’ll take useful. The university ideal is interesting.

I was confident because I thought I’d hit the interesting and useful jackpot. I’ve read so many papers about Twitter, the bar seemed low, I thought it would be easy even. Papers proposing the addition of semantics (I never saw the point of this, the power of Twitter is it’s simplicity), making simplistic errors like saying you could only @ someone if you followed them. Papers proposing a system of vast complexity in order to facilitate how people used instant messenger asynchronously or their status to send a message… and then Facebook came and blew that idea away by just inviting people to set their status and then displaying it in a stream friends can dip in and out of.

This search for interesting above utility has the potential to spawn research that’s like trying to find the fastest way to ski on old-fashioned straight skis. What’s the point in that? If you goal is to ski as fast as possible, you need to get a pair of parabolics and learn how to master the parabolic technique. This research has a very limited audience. It is not where I want to be.

In this analogy, University is like the ESF (Ecole du Ski Français), an institution so steeped and stifled by unionization and protectionism they will take you, on your parabolic skis, and teach you to ski upright with your legs jammed together, rendering it impossible to take advantage of the parabolic edges.

I need to be skiing parabolically. I want to be heading to what I think is important – usefulness, and interesting. But if I have to compromise, I’d sooner be working on making better hairbands (useful, but not that interesting), than making better straight skis (interesting, but not useful). This means that I want to be at the edge of Computer Science, not in the middle of it. Because that’s where I think the innovation is. Perhaps grad school is not the place for me as a result of this.

Yes, what a terrible time to realize this.

In any situation, there are options. What are mine?

• Give up, drop out.
• Work harder to determine a “contribution” I can claim to make.
• Stop trying to publish and just write a thesis.
• Something else I’ve yet to think of.

Advice welcome.