“Crowdsourcing” is the notion that the collective data output of many individuals can be turned into a valuable commodity, for any number of different purposes. It’s the idea, for example, that citizen journalism (in the form of CNN iReports or idle commuters reporting traffic jams via a smartphone app) can be just as informative as the work of professional journalists. And then there is the potential for the crowd – wittingly or otherwise – to become criminal apprehenders, mass data producers and/or components of a surveillance apparatus.
Earlier this year in Vancouver, rioters were brought to justice by average citizens when police asked anyone with photos or video of offenders-in-the-act to submit them to the department. The incident was a stark example of collective law enforcement, with regular citizens providing information – and by extension, evidence – that the beleaguered Vancouver police force simply could not.
On other fronts, the Department of Homeland Security is working on a technology that would turn cellphones and their owners into individual detectors of harmful chemicals, transmitting locational data back to first responders and other monitoring agencies when the phone “smells” trouble. In an almost farcical crowdsourcing venture, the Intelligence Advanced Research Projects Activity (IARPA) has launched the Aggregative Contingent Estimation (ACE) Program. ACE attempts to harness the collective predictive power of many to forecast future events. Spending a bit of time on the website of Applied Research Associates (to whom a research contract was awarded) is all one needs to get the gist of the concept.
The Vancouver incident represented a conscious decision by citizens to assist in the law enforcement efforts of the state. Less clear is the potential intelligence value and ethical propriety of crowdsourcing when the crowd is unaware of its participation in such an activity. A smartphone app like Color illustrates the surveillance possibilities of crowdsourcing.
Imagine a concert or sporting event where hundreds or thousands of photos are snapped and immediately posted to the Internet – publicly available data with potential intelligence value. It’s likely that few people’s civil liberties sensibilities would be offended if one of these photos happened to capture the face of a known terrorist who was then detected and soon after apprehended. But what about an innocent individual who, thanks to an interest in Islamic extremism and a recent purchase of chemicals that could be used for bomb making, increasingly meets the algorithmic data mining criteria of a potential terrorist? Might that photo be the proverbial tipping point for a trigger-happy intelligence analyst?
This hypothetical admittedly conceives of a scenario that is unlikely at present, but certainly within the realm of possibilities as government continues to gather data of all kinds from its citizens. We know the Department of Homeland Security reserves the right to monitor social media – though we certainly don’t know the extent of government’s social media data access and analysis. We know facial recognition of photos is already a reality. And we know smart phones are increasingly capable of documenting their surrounding environment completely, transmitting various forms of multimedia instantly while logging the moment of capture geographically.
Crowdsourcing in the age of Web 2.0 is still in its infancy, with untold prospective uses to be explored in the years to come – by entrepreneurs, the media… and data miners.