Clay Shirky is one of my favourite thinkers about things Internet, most of all because he slices through the bull and gets to the bone really well. When looking at all the hype surrounding Grid computing, he finds that there isn't any bone, and in the process comes up with both some interesting thoughts about how we use ICT, and a nice point on thinking about it.
This issue's essay is about Grid computing, the generalization of the [email protected] pattern of distributed computation into a sort of "supercomputing on tap" application. I was asked to be a "provocateur" on a panel about networked computing, focussing in particular on Wifi, Grid computing, and the Semantic Web. I had to write back to the panel organizer and report that I had the least provocative views possible on both Grids and the Semantic Web, namely that I thought both would be moderately successful -- not revolutionary, but not failures either. This led to a realization: people who try to think clearly about technology always run the risk of unconsciously gravitating towards technology its easy to think clearly about. It was interesting to write about the Web in the early days because its importance could hardly be overstated, or about WAP, because its wrongheadedness ditto. It's harder to think about Gird computing or the semantic web, because they are not so obviously headed for either ubiquity or uselessness. So this essay is an experiment in writing about something -- supercomputing on tap -- that is going to succeed, but will do so in a way far less important than its proponents believe.
Comments