Researchers are creating digitized versions of cities from thousands of photos that people have shared online, Physorg.com reports.
The program used 150,000 photos of Rome to create a 3-D digital rendering of the city, for instance.
This capability is another step toward truly open-source intelligence: publicly available images and other information will be able to generate increasingly detailed snapshots of places, people, and particular moments. And the ability to analyze this information will steadily democratize, and grow more powerful: Google is experimenting with facial recognition software that could reveal the locations and activities of millions of people who aren’t even the primary subject of a given photo, as just one example.
This idea has historic antecedents, by the way. During WW II, the OSS collected American’s tourist photos of Europe, cataloged them and made them partially machine-accessible, and used them to plan bombing raids. (Nicholson Baker, “Deadline,” New Yorker, July 24, 2000, 47.)
(Tip from @Changeist)
(Image: Creative Commons from gruntzooki, Flickr)
A curious intelligence experiment is underway on the Web, the NYT reports.
American intelligence agencies are putting 48,000 boxes of captured Iraqi government documents online, in Arabic.
The idea, pushed by conservatives, is that this will give more attention to neglected details of Iraqis activities and intentions, possibly bolstering the case for war.
This particular scheme may be amateurish in execution, but the underlying idea — truly open-source intelligence — is profound.
Open-source intelligence traditionally refers to publicly available sources analyzed by conventional intelligence agencies.
But the newer meaning of “open source” — opening information, methods, and output in a system to broad participation — could be applied in the intelligence sphere.
Civilian analogues are already in operation:
- Amber Alerts to find missing children are a method of rapidly creating a dispersed intelligence gathering apparatus.
- Fighting child exploitation, investigators have released images with the victim digitally removed, in hopes that the public can identify something from the background or other clues. These methods have sometimes yielded results.
Applied to intelligence, open-source methods could address two problems: inadequate attention resources, and inadequate information gathering. Both are enabled by the networked information revolution that connects ever more people and devices.
Attention: networks of varying openness could be applied to answer three kinds of problems:
- Is there anything to this? Does an image or document warrant further attention?
- Do you recognize this? Where, who, or what is this?
- Is anything happening? Networks could be tasked with real-time monitoring of sensors or imaging systems.
Gathering: The Internet and mobile phone networks are forming an increasingly pervasive information network that far surpasses that of the world’s intelligence agencies. These networks can be used in several ways:
- Passive gathering: The Web and mobile networks are increasingly good for near-real time monitoring of events. After the London bombings of 2005, it was possible to get multiple images of many of the bomb sites from the photo site Flickr.
- Post-facto gathering: The library of images captured by civilians will grow more and more vast, and it is not long before any public event will be recorded by numerous devices. It will be possible to request images of an event or place from the public with near-certainty that the image is out there. (A version of this was done by the US armed forces during World War II, when they asked the public to send in all their pre-war tourist photos of Europe, in order to compile a visual intelligence database. Flickr could achieve the same results in a matter of minutes.)
- Real-time gathering: In certain circumstances and places, it will be possible to put out a request for civilians to gather data in real time.