Saturday, November 26, 2005


it's not enough to do good work as an academic, at least in computer science. in addition to doing good work, you need a good story. often, this means a lot of time spent thinking about how to find a common thread amongst related, but rather independent pieces of work.

this clearly has costs. we lose time that could have been spent on technical work, we mislead novices about how research is done, and we engender cynicism about the research process.

and the benefits? we have a nice stories to tell, but who are we fooling? ourselves, our peers, outsiders? or is this some benefit to this part of the process as well? does trying to find these common threads help us see the bigger picture? and if so, what are the concrete benefits of that seeing that picture?

Tuesday, November 15, 2005

grad school #1

the central conflict of grad school: to be pushed hard enough to produce your best work, without going insane.

mediocretes #1

it is more important to be understood than to be heard.