Sunday's New York Times featured a disturbing story on the IT culture clash between Google and the FBI:
Data is compartmentalized so that case information compiled in Phoenix might not be accessible to agents in Minneapolis, and retrieval of the full text of case reports is not possible. Devised for the quick retrieval of the names of known suspects, the network can be searched for terms like "aviation" or "schools, " but not "aviation schools" -- in other words, precisely the kinds of phrases that may have made it easier for law enforcement agents to connect the dots and discern the patterns of activity leading up to Sept. 11 attacks.
Mr. Schmidt of Google said that government had characteristically been slower than industry to adopt new information technology and to link its multitudinous information networks. This leads to a condition that the industry calls "stovepiped" information, which means that data is warehoused in separate, unconnected silos. That is partly by design, Mr. Schmidt said, as a precaution against wandering hackers. "They don't want a network interloper to come in and do a lot of damage to other computers." [New York Times]
I'm sure it's true, though no-one can come out and say so, that the FBI are among Google's most intense users. I hope a private network of weblogs will be the next step. Valdis Krebs has a new paper that suggests how social network mapping can be used to thwart terrorists. He writes:
To gather the data for mapping these networks, individually and as a group, requires much cooperation between departments, agencies and countries. This requires vertical, horizontal, and diagonal links between all of the investigators on the case -- in other words, our network needs to be as good or better than enemy's! [ Valdis Krebs ]
Maybe I've just got blogs on the brain. But like all stovepiped IT organizations, the FBI's will not be rebuilt anytime soon. The way forward is a human awareness network layered on top of those stovepipes and connecting them.
Such an overlay network needn't, of course, intersect with public blogspace. But purely internal use of existing low-tech weblog software could reproduce the same effect: a knowledge network with human routers. Would it be perfectly secure? Of course not. But in the end, what's the greater risk? That the enemy might discover we had connected the dots and have to change its plans? Or that we have no hope of connecting the dots at all?
Former URL: http://weblog.infoworld.com/udell/2002/06/12.html#a298