To an extent, as with all big data systems, massive collection of data has both opportunities and risks. The risk in overcollection (e.g., the NSA type programs) is that you collect so much damn information that you are either behind in analyzing it or flat out never get there [1-2]. Combine that with (as others have noted) the lack of translators and you get what you got [3]. The old maxim* that not all data is information (DIKW) heavily applies here [4]. This whole phenomenon becomes especially problematic when you are spending all that time energy and money largely trying to predict 'black swan' events using what has happened before as what to look for in the future (e.g., the TSA as a whole or the NSA still focusing on cold war languages pre-9/11 [3]).
To an extent, as with all big data systems, massive collection of data has both opportunities and risks. The risk in overcollection (e.g., the NSA type programs) is that you collect so much damn information that you are either behind in analyzing it or flat out never get there [1-2]. Combine that with (as others have noted) the lack of translators and you get what you got [3]. The old maxim* that not all data is information (DIKW) heavily applies here [4]. This whole phenomenon becomes especially problematic when you are spending all that time energy and money largely trying to predict 'black swan' events using what has happened before as what to look for in the future (e.g., the TSA as a whole or the NSA still focusing on cold war languages pre-9/11 [3]).
[1] http://www.zdnet.com/article/nsa-whistleblower-overwhelmed-w...
[2] http://apps.washingtonpost.com/g/page/world/how-the-nsa-trie...
[3] http://www.washingtontimes.com/news/2009/aug/31/lack-of-tran...
[4] https://en.wikipedia.org/wiki/DIKW_Pyramid
Edit: *fixed typo of maximum to maxim