Crowds may have wisdom, but some crowdsourcing concepts could be turned to evil purposes, writes blogger Steve Kelman.
Earlier this week, at our weekly faculty research lunch seminar at the Kennedy School (one of the few "free lunches" that has survived our budget crunch-driven cost savings), Jonathan Zittrain, the Internet law and policy guru at Harvard Law School, gave a fascinating presentation on various developments in the burgeoning world of crowdsourcing, broadly conceived.
Zittrain began by discussing sites that try to bring inventors to problems -- someone will advertise a problem and a prize for the first innovator to solve it. The government, I have argued in the past, should use such sites more often. It's a subset of the concept of using prizes as a procurement tool -- something, to my understanding, of which ederal chief technology officer Aneesh Chopra is a devotee. (There is a great Wikipedia entry that explains this further.)
Zittran then noted the growth of applications (this one from the U.K.) where people, for very small amounts of money, are apparently willing, from the comfort of their couches, to monitor crime surveillance cameras to look for suspicious activity and report it. Some companies are also getting people, again for micro-payments, to report in if they recognize photos of people participating in a mass marijuana smoke-in.
Zittrain's purpose was to raise questions about the downsides of all this. Terrorists could request help on having people develop different chemicals that could be put together to create some horrendous weapon of mass destruction, without the innovators being able to connect the dots on the little part of the project they were working on.
The Iranian government could have people match photos of protesters against huge photo rosters of the residents of Teheran. There are already a number of sites that give people micro-payments to vote on Amazon-style consumer feedback applications in favor (or against!) a product, a movie, a book, or whatever -- including in some cases getting people to click that all reviews that argued the opposite were "unhelpful."
Zittrain's presentation was fascinating, but my reaction was partly that lots of good things have problems associated with them (think of cars as a great example), and that some of these problems -- such as inventors working on a narrow part of an evil chemical without knowing the broader picture -- could occur in a non-crowdsourcing world as well.
Zittrain seemed to be a fan of using crowdsourcing techniques themselves to reduce the problem, by various kinds of opportunities for alerts or warnings (though obviously these too could be subject to voting manipulation of the kind he described). Nothing's perfect, but I came away from Zittrain's thought-provoking talk probably with the opposite conclusion from the one he intended us to draw, namely that there is a lot of cool stuff going on here.