Over the next few weeks I would like to begin laying out the theme of this blog. I’ve been thinking about exactly what it is I would like to say, and what would be of value to discuss with the online community. At its core, this blog is about information retrieval in its myriad forms and applications: Text, Music, Images, Web, Video, and so forth. But the way in which these systems are designed, whose information needs they are trying to meet, and by what metrics they are evaluated, are extremely important considerations. One thing that I would like to do is investigate the assumptions underlying the retrieval systems that we are all creating and using.
As I was thinking about this topic, I was listening to a KQED Forum podcast from December 2008, entitled 40 Years of the PC. It is a discussion with Brian Cooley of CNET, Christina Engelbart (Doug Engelbart’s daughter), Jeff Rulifson of Sun, and futurist Paul Saffo. They’re discussing the historical implications of the famous 1968 demo in which Doug Engelbart and others demonstrated the mouse, hypertext, cut-and-paste, and other revolutionary aspects of personal computing.
At one point in the discussion, one of the guests — I believe it was Brian Cooley — makes the comment that it was unfortunate that Doug ended up being best known for creating the mouse, because that was one of the least important ideas that he had created. His goal, really, was to change the world through computing. He didn’t (necessarily) want to create “the mouse”. He wanted to create systems and tools that let people work together to solve the world’s difficult problems. And in a way, even though personal computers have been around for 40 years, we’ve never come close to creating the types of tools that Doug had envisioned. Doug thought that we need, paraphrased Cooley, to have “high performance tools for high performance individuals”. Cooley continued:
“It’s the difference between a 10-speed racing bike and a tricycle. A tricycle you can get on and ride around immediately, but you’re never going to go very fast. A 10-speed…takes some effort. And he [Engelbart] was trying to build a 10-speed. Everybody else built tricycles.”
This is one of the big questions I have today about the information retrieval systems that we are collectively building. It has been 40 or 45 years since some of the early computer-based IR systems were created. It has been that long since a lot of the foundational ideas in information retrieval were written about. How are we living up to that promise? Are we creating 10-speeds yet? Or are we still creating tricycles. I hope to explore this question more.
What you need depends on what you want to do. If you want tool around the parking lot, going through puddles, a tricycle is the right tool for the job. If you want to go to the grocery store in Amsterdam, a klunky black three-speed with a big basket is what you want. If you want to chase Lance Amstrong, then not even a 10-speed will do.
Most people in the world ride bicycles to get somewhere, and don’t care much about their bikes as long as the tires don’t go flat & the bike doesn’t get stolen.
Now if you want to talk about bicycle parking…
What you say is all true. I’m a firm believer in having the right tool for the right job. And depending on the situation, that tool will differ.
The point I was driving at (is that a pun? you decide) is that the distribution is akimbo. There are too many tricycles and no 10-speeds. You don’t always need a 10-speed. But you also don’t always need a tricycle. But all we seem to be building are tricycles.
Pingback: Information Retrieval Gupf » Empty Calories