Privacy: All about control

I have been thinking about a theory of privacy online lately, and came across a fairly thought provoking article which will appear in the latest Newsweek. It points out that companies like Microsoft, Google, and AOL are gathering all sorts of personal information, and that there’s a good chance that we won’t complain because of all the benefits of personalization and customization. However, it notes:

The result would be an erosion of privacy very different from the ones privacy advocates have long warned about—coming not at the hand of a big-brother government, nor of stalkers, hackers, telemarketers, spammers or abusive employers, but rather of companies bribing us with services.

They sort of get the point: that privacy in Web 2.0 is all about control. I don’t mind letting the world know what music tracks I listen to every day through Last.fm (recommendations through groups and profiles are a good thing), but when iTunes included a recommendation service based on listening I was offended. I don’t mind posting information about myself on facebook, or on this blog, but when others gather that information to advertise or to even make assumptions about who I am, again I get offended. Sharing is a good thing, but I (and would argue others) want to have intimate control over exactly what I share or others do with my data.

Perhaps I’m thinking of personal information as something that can be owned by an individual. As someone in the article says, “Wouldn’t it be nice if the network knew you liked to visit warm places, and put together a $3,000 package to Cancún?”

No, I may like warm places, but when I’m ready for a trip and have the money, I’ll come looking for you. It might be different if one was willing to give some sort of additional compensation for the information…why should we agree to be advertised to for mere “convenience.”
Interesting new terms in the article:

• sentient network: “The network magically figures out what you’re looking for and provides it at the right time, anywhere,”
• Personal data distortion (de-identification): tools which let one slightly morph personal data so that things are still useful, but not personally identifiable.