I just noticed your response to my essay. Thank you for the critical commentary.

There is a big difference between a document search and spy search. For one, you know some certain things about the document. For example, key words, dates, etc.. You know that it exists. Second, the document is not actively trying to be not-detected. In fact, there are no certainties about finding terrorists, even whether or not there are any to be discovered. And terrorists that might exist are actively trying not to be discovered.

The secondary search of the 30,000, in fact, is already included in what NSA does. No matter how complex and multi-staged is their identification system, it is still a system that is subject to Baye’s Theorem.

There is, also, of course, the obvious fact that IF….IF…NSA could do this kind of thing well, then there is no terrorist problem. They get caught.

Part of this problem is definitional. A terrorist can only be known for certain after the fact. So, we have the case of the young man who was talking about a dirty bomb, this is a man who knows nothing about bombs and who has no access to radioactive materials. He has been tried and imprisoned as a terrorist. But the reality is that he is a young man with a big mouth, an active imagination, and a bad lawyer. We are not safer for him being imprisoned.

While resources are being wasted on this, they are being diverted from sting operations, human resource spies, etc. Plus, our intelligence agencies lose credibality with the public, who ultimately control the budgets and continuing existence of the agencies. Public employees like NSA need to take care that their employers (the rest of us) appreciate their use of our resources. Endangering that good will is another cost to monitoring all US phones and email.

In my opinion.

Floyd Rudmin