Two things regarding the above exchange.
On the one hand, Danika makes a fair point. The USA seems to have a history of 'collecting data', intelligence 'gathering' and yet, being rather in the dark about the things they gathered intelligence ON. I am not talking about the WMD-thing (I think that was a lie/excuse rather than a mistake). I am talking about, say, the warnings the USA intelligence received from French intelligence agencies about a few of the people that went on to commit the 911 attacks. I am not an expert, but this seems to be a bit of a tradition, in the US, going back to WW II, Vietnam, etc. Even if their spies were the most well-funded and equiped, there just may be a REASON why 007 is a Brit.
And it would make sense too. The USA seems to have many competing agencies that often seem to feud with one another. How is effective filtering and 'adding up' of information possible, let alone information actually leading to action, in such a scattered landscape of agencies?
And yet, I think Treat has a point as well. The plans of the NSA obviously went way beyond just collecting stuff. The whole prism thing seems to have been about creating semi-intelligent software that would be connecting the dots, that would be able to do advanced data-mining from meta-data (though I do believe Snowden when he claimed that the stuff that the NSA collected went well beyond meta-data). I read something about the NSA asking/demanding information from google about the AI systems they use to sort their queries and create their translations (I think that Google, of all major private companies is probably the one that devotes most energy to developing processes that might be said to border on AI). I think the ambitions of the NSA in that direction were pretty clear.
This is one of the reasons why the whole NSA thing worries me more than it seems to do most people: I think that such a smart system 'connecting the dots' is entirely possible. And I do think that such a thing may well allow such a system to do far more than just find terrorists, in the future. If you add up all the stuff people do online (even if it is just what sites they visit and what they type in in google), you can probably form a pretty clear picture of a person, their interests, proclivities, sexual preferences, politics, financial situation, desires, aspirations, and so forth.
For example, if I could see every site that Treat or Danika have looked out, if I can see when, I can probably see how long they have been looking at the site as well. And what link they clicked. Should Treat have a tendency to linger longer on, say, sites with pictures of pretty women, I could examine what the pictures he looked at where exactly, which link he clicked, or which picture he paused at. Most likely, that would tell me whether he prefers blondes or not, his preference for but and breast-size, and so forth.
Harmless, you may think. But the fact is that most of us here do not post using our real names. Not because what we do is a crime, but because it is no one's business. I'd rather that there WAS no one, no system that would allow anyone to make it their business, regardless of assurances about responsible use. Especially not from a country that thinks it is fine to spy on people if they are not Americans, with my own government (of Smurfistan) offering our great friends from the USA total access to the private lives of it's citizens, perhaps in the hope they will get some scraps of the data collected back in return.
Yes, I am ranting. Stopping now.