Reading for October 11
“Alessandro Acquisti and Jens Grossklags, Privacy and Rationality: Preliminary Evidence from Pilot Data, , The Third Annual Workshop on Economics and Information Security (WEIS04), May 2004″ provided a nice overview of literature hinting at stuff that I have suspected for a long time: that humans are not rational. However, their experiment was too small of a sample size for me to feel like it was worth my effort to read about it.
“George Danezis and Ross Anderson, The Economics of Censorship Resistance, The Third Annual Workshop on Economics and Information Security (WEIS04), May 2004″ seems like a horribly contrived model as my effort to share resources is not so related to my ability to fight censorship as the author suggests. Likewise with the cost of fighting censorship. It does, however, offer the interesting observation (apparently, just from looking at the equations) that censorship is only worth fighting when fighting it could be twice as good as not. They seem to be arguing for known heterogeneity in systems that are out to fight censorship, and point to music sharing systems as an example. However, this seems at odds with the desire to combat free-riding. In a sense, I suppose free-riding is a form of censorship - it is a node censoring itself from speaking at all.
Our particular censorship model provides some further insights. For both random distribution and discretionary distribution, the censor will meet resistance from a node once his activities halve its utility. This is because, for the particular utility function we have chosen, a node will react to mild censorship by investing in other resources rather than engaging in combat. So mild censorship may attract little reaction, but at some point there will be nodes that start to fight back, starting with those nodes whose preferences are most different from the censor’s. This is consistent with intuition, and real-life experience.