That's the thing...I [we] don't know what keywords trigger deeper searches, or how exactly our browsing/typing history is tied together in which database. If I type 'Bin Laden' in a chat, does it flag me somewhere? Or is it filtered out as useless?

If I write, 'I had a blast at a party last night' (a really simplistic example, but you get the point) or 'Check out this story about this guy who wanted to kill Obama..he was crazy!',does that do anything, or are those keywords ignored?

What's considered a false positive, and what's a true positive? The algorithms do the filtering, but the humans write the algorithms. What's the error rate?

I think I am most afraid of what I don't know about how this works.