The Net Delusion - Evgeny Morozov [86]
Tremendous cost savings introduced by digital surveillance technologies have also made it possible to shift surveillance personnel to more burning tasks. In a 2009 interview with Financial Times, a marketing manager for TRS Solutions, a Chinese data-mining firm that offers an Internet-monitoring service to the Chinese authorities, boasted that China’s Internet police—thanks in part to the innovations developed by TRS Solutions—now only need one person where ten were required previously. But it’s too early to celebrate; it’s unlikely that the other nine were laid off. Most probably they were shifted to perform more analytical tasks, connecting the dots between hundreds of digital snippets gathered by automated computer systems. As the TRS manager pointed out, business is booming: “[The Chinese authorities have] many different demands—early warning, policy support, competitive spying between government departments. In the end, this will create a whole industry.” Perhaps, this is not the kind of Internet-friendly industry celebrated by the proponents of wikinomics, who rarely acknowledge that, while the Internet has indeed helped to cut the unnecessary slack from many an institution, it has also inadvertently boosted the productivity of the secret police and their contractors in the private sector. A book on “wikiethics” is long overdue.
Say Hi. You’re on Camera!
It’s not just text that has become easier to search, organize, and act on; video footage is moving in that direction as well, thus paving the way for even more video surveillance. This explains why the Chinese government keeps installing video cameras in its most troubling cities. Not only do such cameras remind passersby about the panopticon they inhabit, they also supply the secret police with useful clues (in 2010 47,000 cameras were already scanning Urumqi, the capital of China’s restive Xinjiang Province, and that number was projected to rise to 60,000 by the end of the year). Such revolution in video surveillance did not happen without some involvement from Western partners.
Researchers at the University of California at Los Angeles, funded in part by the Chinese government, have managed to build surveillance software that can automatically annotate and comment on what it sees, generating text files that can later be searched by humans, obviating the need to watch hours of video footage in search of one particular frame. (To make that possible, the researchers had to recruit twenty graduates of local art colleges in China to annotate and classify a library of more than two million images.) Such automation systems help surveillance to achieve the much needed scale, for as long as the content produced by surveillance cameras can be indexed and searched, one can continue installing new surveillance cameras.
But as the maddening pace of innovation in data analysis expands the range of what is possible, surveillance is poised to become more sophisticated as well, taking on many new features that only seemed like science fiction in the not-so-distant past. Digital surveillance is poised to get a significant boost as techniques of face-recognition improve and enter the consumer market. The face-recognition industry is so lucrative that even giants like Google can’t resist getting into the game, feeling the growing pressure from smaller players like Face.com, a popular tool that allows users to find and automatically annotate unique faces that appear throughout their photo collections. In 2009 Face.com launched a Facebook application that first asks users to identify a Facebook friend of theirs in