• by adam , Shelly Palmer , March 2, 2006

People who are schooled in the art of the Internet or those who understand how computers work know that anytime you interact with a computer, your actions can (and probably will) be logged. The level of detail depends upon who is doing the logging, but it is not unreasonable to expect your log on time, log off time and keystrokes to be recorded. This, of course, includes everything you type as well as the address of every Web site you have visited and what you did there--every word of every unencrypted e-mail you send and receive, every search you do, everything.

Who would or, more importantly, who should care about living in such a world?

Two-way, interactive communication over vast networks theoretically offers limitless potential for good. The more information you allow "the system" to have about you, the more the system can adapt to your needs, wants and desires. The disciplines of behavioral targeting and collaborative filtering, when combined with data mining and data warehousing technologies can, in theory, offer a world of information and entertainment in ways that test the limits of one's imagination.

That's the sales pitch. The reality is very different. Once "the system" knows you, someone must ask the question, "Who knows the system?"

This question didn't even exist in the past. (And, it doesn't really exist for people who spend most of their time offline.) In the offline world, statisticians make a very nice living aggregating data and predicting what populations might do. They don't mind using anonymous data because it is accepted dogma that statistical analysis does not attempt to predict what specific individuals within a population might do. However, in the online world, there are no populations, there are people. Populations may do things and there may be some value to trend analyses--but those aggregations are simply constructs for management. The actual data exists, and it exists on a "big brother" basis.

What makes this thought experiment interesting is the idea that while the predictive power of aggregated data can be "stolen," it has very little malevolent usefulness. On the other hand, aggregated behavioral data tied to an individual and gathered without the express permission (or knowledge) of that individual has so many malevolent uses they cannot all be listed.

We are quickly evolving into a technocracy where the expectation of privacy is unreasonable. If you sit down at a computer, you are "online." Is what you type protected? What if you need the help of your community of interest? What if you must ascribe to a collaborative filter set to enable your discovery of new content? What if you want to take advantage of the recommendation engines from retailers? What if you want to become part of a tag cluster or social network? Can you reasonably expect privacy?

Most people would say their privacy has been violated if someone obtained and distributed information about them in a way that was inconsistent with our social norms or the reasonable expectation of privacy. Obviously, there are legal definitions of personal privacy, although the current laws do not define current technology as well as they might.

In the offline world, people "do" life in private. Some people even do business in private. That's the role of printed money in our society. The local street vendor does not give you a receipt for the pretzel you bought on your way back to the office. Will he report the transaction to the government? Will he pay taxes on the income? Will he log the purchase and make the data available to your nutritionist? Your doctor? Your significant other who has put you on a low-carb diet? Of course not. One can think of literally hundreds of transactions that ordinary people complete that they would "prefer" to do anonymously.

Anonymously... an interesting word. But not one you should grow attached to. As we harness the power of technology to make life better, we will be moving toward a world where the expectation of privacy is totally unreasonable.