05-14-2019 01:57:15 PM -0400
05-09-2019 05:01:30 PM -0400
05-09-2019 01:41:48 PM -0400
04-18-2019 10:46:35 AM -0400
04-18-2019 10:18:40 AM -0400
It looks like you've previously blocked notifications. If you'd like to receive them, please update your browser permissions.
Desktop Notifications are  | 
Get instant alerts on your desktop.
Turn on desktop notifications?
Remind me later.

Wide Angle Scan

Students at a geography class in UCLA correctly guessed that Abbottabad was one of the likely hiding places for Osama Bin Laden back in 2009. Their deductions came entirely from open source informed by common sense. Well, maybe not so common.

Science reports that UCLA geographer Thomas Gillespie, together with colleague John Agnew and a class of undergraduate students, developed a probabilistic model about the whereabouts of bin laden. Gillespie was teaching a class on remote sensing, and he and the students used publically available information on bin Laden’s movements before he disappeared to predict where he was likely to be. Science notes that the students used a theory called “island biogeography.” The theory says that species on large islands are much more likely to survive a catastrophic event than species on small islands. “The theory was basically that if you’re going to try and survive, you’re going to a region with a low extinction rate: a large town,” Gillespie says. “We hypothesized he wouldn’t be in a small town where people could report on him.”

“It’s not my thing to do this type of [terrorism] stuff,” he says. “But the same theories we use to study endangered birds can be used to do this.”

Based on the information on bin Laden’s movements and on the theory, the group predicted that bin Laden was hiding in the Pakistani city of Parachinar, a town not unlike Abbotabbad (the model predicted that there was a 100 percent likelihood that bin Laden was in Parachinar, and 80.9 percent that he was in Abbottabad). Gillespie said that one of the problems they faced was the fact that there was no reliable information about bin Laden’s movements after 2001. Still, Gillespie was not surprised that bin Laden was in a town. “Caves are cold, and you can’t see people walking up to them,” he says. The late al Qaeda leader made a bad choice of real estate, in Gillespie’s opinion. “An inconspicuous house would have suited him better.”

The paper itself gives some examples of the types of other deductions they made.  They guessed, for example, that he'd be in a multi-storey building surrounded by high walls. Parachinar was favored because it had historically been a hiding place for mujahedeen.  Abbottabad was not named specifically, but was within a high probability zone given is distance and characteristics. Still, the actual hiding place of OBL was roughly within the parameters they guessed it would be in.

Life History CharacteristicsPhysical Structure Attribute
Is 6’4" tallTall building
Requires a dialysis machine that uses electricityElectric grid hookup or generator
Prefers physical protectionWalls over three meters high
Enjoys personal privacySpace between structures
Retains a small number of body guardsMore than three rooms
Prefers to remain protected from aerial viewTrees for cover when outside

One might argue that the paper made a lucky guess; and that hindsight always confirms a past guess, conferring upon it wisdom it never truly had. Nevertheless, the Gillespie and Agnew paper was on the right track. And being on the right track is important. In general, when looking for objects, the first clues often come from the wide scan. Anyone who has poked around in the woods knows that the first indication of something out there often comes from the naked eye.

One the slight movement or disturbance in the pattern is found in the wide scan, the next step is often to glass in, using binoculars or simply by just staring at the location of the anomaly. The world is a big place, and it is often impractical to examine each square inch through a viewport the size of a straw. Normally the procedure is to bound the area of search and progressively reduce it. Once you get down to the nitty gritty then the glass comes out for the closer look.

But the problem of constructing the wide scan in the first place is often a nontrivial one. Because first of all you have to be prepared to see. It's a shock to realize that every human being sees with his mind, not really through his eyes. The image as physically transmitted to our retinas are upside down. The brain flips it over by image processing. Just how powerful this routine is was demonstrated in an experiment by George Stratton, who wore glasses which which inverted the normal image of the eye. At first, he saw them as upside down. But "on day five, images appeared upright until he concentrated on them; then they became inverted again. By having to concentrate on his vision to turn it upside down again, especially when he knew images were hitting his retinas in the opposite orientation as normal, Stratton deduced his brain had reprocessed his vision and adapted to the changes in vision."

What bureaucracies see in the "wide scan" -- the pattern their analytical cells see -- literally depends on what they are prepared to see. Gillespie and Andrew knew what to look for. And so did the bureaucracies, apparently.  But if the bureaucracies don't know how to do the wide-scan, they'll miss what they don't want to see.  There "are none so blind as they who will not see". The contribution of "intelligence fusion" is that it creates the possibility of the wide scan. It provides intelligence agencies with a framework within which to apply their nifty ground pentrating radars, the secret agents, the long-range lenses. It is the brain side of seeing.

But intelligence fusion does have some drawbacks. One of those problems is the reification of the system itself. An intelligence fusion system cannot exist in the abstract. It must be implemented and the users are very often unaware of the blind spots created by the architecture in practice. The software designers no doubt know more about the trade-offs they've made but have no easy way to estimate their effect on the effectiveness of the system. The result may be users who are happy with "the system" though they don't know what it does, together with designers who are happy that the users are happy, although they don't know why the users are happy.

This can happen for a long time. A Kiwi biologist attempted to understand why the vertebrate eye had all the cabling in front of the sensors while the invertebrate eye had all the cabling behind the sensors. The answer he gave was that the power supply requirements -- the need to nourish the huge pipeline -- dictated the front to back vertebrate design. The result was a "blind spot" where the nerve system were, but nature needed the bandwidth more than the optics. Only nature didn't tell us until we learned to listen to anatomy. We were happy with the design before we knew how it worked.

Although the Gillespie and Agnew paper will probably be remembered for guessing the location of OBL fairly accurately, maybe its real importance is to demonstrate how critically dependent our sight is on what we are prepared to look for. The words of Auden come back to mind, slightly changed.

For knowing makes nothing happen: it survives

In the valley of its making where executives

Would never want to tamper, flows on south

From ranches of isolation and the busy griefs,

Raw towns that we believe and die in; it survives,

A way of seeing, a mouth.

"No Way In" print edition at Amazon

Tip Jar or Subscribe for $5