The idea for this post came to me after I had followed two recent stories on the web. The stories, while seemingly unrelated, touched upon two intrinsic elements of crowdsourcing (I’m using this term loosely here): noise and manipulation. As I was reading through the stories and the ensuing blogosphere chatter, I though that how we deal with these elements is likely to determine the trajectory of social media in the future. Sounds important, doesn’t it? So, here are a few thoughts – bear with me:
Story 1 – Noise. The first story dealt with the so-called “citizen journalism”, or in other words a crowd-sourced version of news production. Several pundits have claimed recently that traditional news media, like CNN etc., have become the thing of the past and that witness-journalists can now provide faster and more complete news reporting. As the recent tragic events in Mumbai have shown, however, this is not always the case. There is no doubt that by using mobile phones and platforms like Twitter, Youtube etc. witness-reporters can deliver the information faster. They can also deliver it in massive amounts. Yet this does necessarily translate into comprehensive and/or accurate information– qualities we all look for in news reporting. To the contrary, the Twitter reporting on Mumbai attacks was filled with noise. As blogger Tim Malborne puts it, it actually turned into an “incoherent, rumour-fueled mob …(that) was so drowned in …personal utterance, revenge and irrelvance as to be incomprehensible”. Read his full post here – it’s quite illuminating.
The problem of noise, of course, is not limited to “citizen journalism”. So far as crowdsourcing by definition involves a large number of random people contributing their input to accomplish a certain goal, this approach is bound to generate, along with legitimate contributions, a large amount of garbage. This garbage can be observed in abundance across most social media venues – just browse Youtube to get a feel.
So how can we tackle the problem of noise in crowrsourcing? One obvious solution is to introduce filters to help us separate the wheat from the chaff. Chris Anderson devotes an entire chapter in his book on Long Tail to talk about filters and their role in linking democratized production (read, crowrsourcing) with demand. But the million dollar question remains what should these filters be. Would it be better to delegate control to a select group of individuals with the proper reputation or perhaps those who are motivated to take on the task? This seems to be what Tim Malbone is leaning towards in his follow up essay on how to clamp down on rubbish chatter on Twitter. This is also in line with the approach taken by Wikipedia, which lets page patrollers oversee the process of content creation and editing. In fact, in his recent visit to Barcelona a few weeks ago Jimmy Wales, the Wikipedia founder, went as far as to suggest that he would welcome if folks from Encycolpedia Britannica start vetting Wikipedia articles to certify them as accurate and worthy of the “encyclopedia” status.
But isn’t the “power of a few” exactly what social media was supposed to help us get away from? Didn’t we have plenty of similar filters in the past, like traditional news media, recording labels etc., that determined what was deserving of our attention and what wasn’t? If so, there must be a better way to do this, a way based on the new principles and not the old ones. Can we establish filters to sift through the massive amounts of crowd-sourced content based on… well, the input from the crowd?
…And this is where the second story comes in.
Story 2 – Manipulation. Digg.com is an online community platform designed to help its members discover the most interesting news stories on the web. It works by letting users vote for, or digg, stories submitted by other users; stories that receive the most diggs are then promoted to the front page. If you think about it, this is exactly the type of filter we were talking about – the filter that does not rely on input of “the elite” but that of the community. Guess what, Digg recently banned a large number of its top users for alleged manipulation. Officially, the ban was for using scripts to exploit the algorithm used by Digg to decide which stories to promote. As a result of manipulations, Digg claims, the top 100 users were responsible for over 50% of the front-page stories. Sounds like we’re back to square one and the selected few are still calling the shots. Read David Chen’s blog for the complete story.
According to James Surowiecki the wisdom of crowds comes from putting together inputs from a large number of diverse individuals. Whereas each individual’s input may be imperfect and skewed, their sum usually produces a set of fairly sound and unbiased results. Well, as the Digg story shows this ideal “democratic” process can be distorted through manipulation by a few powerful users. …But should these users be blamed for it?
I would agree with David Chen and say no. And here is why. In the absence of material rewards, the main driver for people to get involved in crowdsourcing production is recognition. Yes, people want to be recognized for what they’ve created and they want this recognition to be explicit. That is why, as David points out, “MySpace has “Friends,” Youtube has “Number of Times Viewed,” and Twitter has “Followers”. But once you have a metric in place and a system that calculates it, it is only natural that people, in their rush for recognition, will look for ways to exploit it. And that’s exactly what happened on Digg.
Perhaps, the right solution for how to establish filters is somewhere in the middle. Perhaps, it is OK if some people participate more and have a greater say in what goes and what doesn’t. After all, subject matter expertise still counts for something, doesn’t it? What’s critical, however, is for the system to remain flexible and capable of weaning out those filters that become bottlenecks. How do we do that? Well, that’s what we need to figure out to take crowdsourcing, and social media in general, to the next level.
nice!