I’m fascinated by today’s effort by The Guardian to crowdsource investigation of the MPs expenses.
My interest isn’t in the topic investigated, really. While I agree that MPs need to be brought down from that Lala land where they can expense duck houses for our hard earned tax money to pay, I do feel that it has somewhat turned into a witch hunt. The time and energy spent by auditors and journalists to establish who should be burnt at the stake first could have been used in much better ways.
So The Guardian came up with a solution; use those idle hours we spent faffing around on the web when we should be working and crowdsource the investigation. Built by Simon Willison and a few others, it is a giant repository of the scanned expenses documents for us to browse.
The process is simple:
- Visit the “Investigate your MP’s expenses” site
- Hit “Start Reviewing” to see the first expense document (Bonus, they provide a progress bar telling us how many pages we’ve looked at)
- Decide what kind of document it is and whether it’s interesting
- Make observations to help the journalists investigate the right entries
So how did The Guardian manage to make it such that we have collectively crunched through 20,000 pages in the past four hours, when we procrastinate for weeks before doing our own 12 receipts worth of expenses at work? By rewarding us and feeding our voyeuristic streak.
The rewards are simple; we get satisfaction in knowing we’re taking part in “justice” being made. So far, I’ve marked two items as “Investigate this!” and I feel like a mini Sherlock Holmes.
In addition to this, the use of a big fat progress bar makes us feel the power of the crowd by showing us how quickly it’s progressing. At one point, I refreshed every 60 seconds to see over 100 documents knocked off every minute. For a generation used to racing games’ lap times and Flight Control high scores, it’s just another little buzz.
Feeding our voyeuristic streak
Admit it, you’ve always wondered what these people spend their allowances. Being able to snoop around feeds that urge. Well… somewhat does. There’s enough black tape redacting out claim details to hold together the hockey sticks of an entire team for a couple of seasons, but we can make out enough to shake a finger and tut at our MPs.
How crowdsourcing could (and should) be used
This kind of manual work that cannot be handled by a computer is already widespread on the web.
Spammers use a clever tactic through which they republish a captcha they want to solve from any given site to a porn site, let an “innocent” porn site visitor solve the captcha by telling them they must fill it in to access the site. Use the solution to access the first site. The poor porn site visitor has not only killed kittens, but also helped a spammer fulfill its dirty deed.
Amazon’s Mechanical Turk is also using the power of crowds by enabling companies to outsource manual tasks to workers at a lesser cost than hiring staff to do colour comparisons or other tasks computers struggle with.
However, the potential for this type of crowdsourcing is amazing. Thousands of people, passionate to get something done, can achieve seemingly impossible tasks when shown a way to direct their efforts.
It makes me wonder how we can unleash our own communities’ potential; beta testing, idea shaping and customer cross-support… The possibilities stretch forever.