Archive for March 5, 2010
The NSF program in “Course, Curriculum, and Laboratory Intervention” (CCLI) has funded much of the development of new undergraduate curricula in computer science (and probably most other STEM disciplines) the last few years. The name is being phased out, in favor of “Transforming Undergraduate Education in Science, Technology, Engineering, and Mathematics” or “TUES.” Notice that the “C” (for Computing) in “STEM” is still silent.
The College of Computing has interviewed three Dean candidates over the last two weeks. All three gave us lots to think about, good advice, and plenty of blog-fodder — but we’re not supposed to name them, so the blog-potential is smaller than it might be. Still, this last one made a comment that I found so striking that I want to talk about it anonymously.
“Do you want to know the top three areas of Computer Science?
Algorithms. Algorithms. Algorithms.”
Historically, that’s an accurate view. Certainly, viewing the world in terms of its algorithms has enabled computing to change the way many disciplines think about their work. However, is that view the one that will push computing forward? Is that where the next great advances in computing will come from?
I suggest that the future of computing is people, people, and people.
- People as co-processors. Luis von Ahn’s home page says that he focuses on “human computation.” What is it that humans can do, that is hard to capture (captcha?) in a computer’s algorithms, that we can then use in concert with computation? The DARPA Network Challenge is a fascinating example of using people as probes, and technology as the networking and processing glue between them. What makes this so powerful is that we can’t understand this as algorithms, but we can use algorithms to leverage human computation.
- People as many, many users. One of our other Dean candidates emphasized the importance of multi-core processing in the future of computing. I think he missed a different massively-parallel phenomenon which is even more fundamentally deeply changing our society.People is different than persons, and social media is more than just individual users being addressed in old-style HCI terms. What emerges when we connect up millions of people through rapid telecommunications networks? Certainly, new things — I’m amazed at the number of press reports I read these days that reference gathering information through Twitter, blogs, and Facebook postings.
There are a lot of research issues to explore here. One that I’ve been thinking about lately: Based on “Nudge,” I predict that a broad range of opinions may initially appear when a new topic arises in a rapid-response social medium like Twitter or Facebook, but the majority of respondents will quickly converge on a small range of opinion. In other words, within a social group, there is no “long tail” effect — friends & followers quickly conform to a few dominant positions, and they do it more quickly than in non-Internet media. Whether or not I’m right, characterizing the behavior of these new forms of media is important, so that we can understand how they’re influencing us.
- Finally, people need to learn about computing. Our first Dean candidate spent a significant amount of time talking about computing education. A particular claim was made that I found interesting. Higher education costs are soaring. They might be capped or limited in some way, or society may expect more from higher education in the United States — like expecting Universities to play a larger role in improving the dismal state of K-12 education, especially in computing. I didn’t hear either of the other two candidates say anything about the responsibility of a College of Computing for improving the state of computing education across the society. Of course, I agree that we do have a responsibility here, to figure out what people should know about computing, to help people learn about computing, and to figure out how to improve computing learning, for both the major and the non-major.
Our past was about algorithms. Our future is about people.