Metadata - Literally data about data. It could be as coarse grained as the number of emails sent from one country's domains to another, or as fine grained as a list of emails that have sent an email to a specific address. Just because a data set consists of metadata doesn't mean that it should be (1)public or (2)accesible without the equivalent of a warrant. Is the existence of communication between two parties just as private as the content of that communication? Should it be?
Datamining - Datamining is the process of observing patterns in data to gain a more fine-grained picture of that data. It's not really new but the combination of more data being collected digitally along with better computing power make datamining more important than ever.
Public data are public. Aggregation may change the usefulness of data, but we cannot make aggregation itself a crime without severely restricting freedom of expresion. In other words: because it is legal to take photographs in public places, for example, it is legal to use those photographs as a data source for mining-- through them one could track where people are going, observe where people shop etc. But, all of this information is public and done with no expectation of privacy. It senseless to change our minds and say "well if you're going to look at the data like that it's wrong!" -- the data themselves have not changed-- this is what I mean by privacy is dead.
But, what we have all gained is the ability to collect and analyze these data ourselves. That's what we have gained. Surveillance isn't just for governments and companies anymore. If you dislike this I challenge you to propose an alternative. A law against datamining? How do you enforce it? Wouldn't such a law be a greater infringement of our privacy?
The real issue are those data held by private companies that exist in something of a gray area. Google can certainly look at who you email and your patterns of use to improve their service. Can they share that same metadata with the government? Probably, it actually matters exactly what sort of metadata it is. Is it just about volumes and domains? Or is it more fine-grained? I feel as though we have not stopped to ask exactly where the line should be. I would be disappointed in the NSA if they didn't use some of this data. I would also be disappointed if they had actually read my emails. I'm not certain if who I send and email to is really private or should be for that matter. After all, the intermediate servers that deliver the email could have a record of who I email. I wonder if there is a precedent with conventional mail. Is sending a letter a public act? The post office certainly has metadata on the volume of letters sent to certain zip codes. I'd love to know about the legal precedents for this.
The point here being, before I can get upset about anything I need to understand what, if anything, was the exact violation. Just saying that they looked at data from facebook and google etc. isn't enough! They should look at such data-- and they should be mining it to. That's the whole point of intelligence.
But, it would be easy for them to cross the line, that we have a whistle blower indicates that they might have done just that. Still there are a large number of people who have not accepted the ways that privacy is "dead" and these people might believe that even coarse metadata and data mining together forms some sort of violation of our rights. I do not agree with this idea because it implies that mining coarse or even public data should be illegal. This is a very bad idea with ramifications that go way beyond just "stopping" governments and companies from spying. It would also stop ordinary people from collecting and using data... it could even stop whistleblowers! Let's think about what we are asking for.
There is a reason that corporatist libertarians are such big advocates of "privacy" --if you have a big company or a lot of money you do not want people to be able to follow what you do, monitor you actions or your movements. Privacy can allow people to hide from criticism. I have had my camera knocked out of my hand by security often enough to know that my ability to collect and analyze data is one of the few powerful things I can do as a "little person" --right now people are pressing for rules that could make it illegal to photograph license plates. (for example) I use this all the time! It's already impossible to take photos in most stores. I don't want to see this get any worse. That's why I'm so hesitant to cheer on a libertarian who says we need more privacy. Who will this protect most? Who will benefit? And how do you implement it? That's what I'm asking.
So let's talk about the specifics and not just assume that if the NSA is looking at call longs this must violate people's rights.