Bruce Schneier has a special September 30th issue of Cryptogram covering the Sept. 11 attacks (he had been publishing only once a month on the 15th - thanks to htp for the heads up.) He talks about airport security, intelligence failures, regulating cryptography, and steganography. Some things he said about intelligence failure started me thinking.
He makes the excellent distinction between data and information in the intelligence community. Data is just a bunch of unconnected, unverified bits of what might, with analysis, become information. They have a lot of data. And what we're seeing now is the reexamining of a lot of this data with the advantage of hindsight. The FBI and the CIA and the NSA are pouring over their mountains of old data looking for clues. And guess what? It looks like they should have known something was up. They had the data. But that doesn't necessarily mean there was a "massive intelligence failure" as some are suggesting. This is just a phenomenon that will always happen when your data set far outstrips your ability to analyse it.
Anyway, the interesting thing I thought (although he doesn't really take it this way) is that this should clearly point out that our intelligence community doesn't need more data! They have the data already. Expanding surveillence is not going to make any difference if it just adds to the mountain of data that then sits in a file cabinet or on a computer somewhere until after an attack happens. I don't have an answer (although more human intelligence and less electronic eavesdropping intelligence might be a start) but this might be an interesting line of defense for people in a position to try and put the breaks on police state happy reactionaries. They knew they should be watching Atta, and they still couldn't do it. How reasonable does it seem to add millions of more "potentially suspicious" people to that list? Won't that just make it all the more likely someone will slip through. What is needed is more information - and it might be the case that accumulating more data is counterproductive.
very interesting
|
He makes the excellent distinction between data and information in the intelligence community. Data is just a bunch of unconnected, unverified bits of what might, with analysis, become information. They have a lot of data. And what we're seeing now is the reexamining of a lot of this data with the advantage of hindsight. The FBI and the CIA and the NSA are pouring over their mountains of old data looking for clues. And guess what? It looks like they should have known something was up. They had the data. But that doesn't necessarily mean there was a "massive intelligence failure" as some are suggesting. This is just a phenomenon that will always happen when your data set far outstrips your ability to analyse it.
Anyway, the interesting thing I thought (although he doesn't really take it this way) is that this should clearly point out that our intelligence community doesn't need more data! They have the data already. Expanding surveillence is not going to make any difference if it just adds to the mountain of data that then sits in a file cabinet or on a computer somewhere until after an attack happens. I don't have an answer (although more human intelligence and less electronic eavesdropping intelligence might be a start) but this might be an interesting line of defense for people in a position to try and put the breaks on police state happy reactionaries. They knew they should be watching Atta, and they still couldn't do it. How reasonable does it seem to add millions of more "potentially suspicious" people to that list? Won't that just make it all the more likely someone will slip through. What is needed is more information - and it might be the case that accumulating more data is counterproductive.
- jim 10-02-2001 3:55 pm
very interesting
- Skinny 10-02-2001 5:14 pm