Coincidentally or not, after I covered the topic of Q & A services in my last Biznology post, I’ve heard complaints from three different acquaintances about the low quality of knowledge in Yahoo! Answers, one of them mockingly calling this world where everybody is an expert “the age of disinformation.” Another friend of mine has recently complained about getting mostly useless content–with zero editorial and zero user reviews–from reputable sites whenever he Googles “<non-mainstream product> review”. Has filter failure become so prevalent that, despite all the information available to us, we are no better off than we were 20 years ago, when content was scarce, difficult to produce and difficult to access?
Image by raider3_anime via Flickr
Two months ago, my wife called me from the grocery store, asking if a product has the expiry date of “11 MA 10”, does that mean May 10, 2011 (which would be good, since it was still April), or March 10, 2011 (which would mean that the product was way past its “best before” date)?
Naturally, my first instinct was to Google it, and inevitably I ended up getting a bunch of entries in Yahoo! Answers. Here are some of the pearls of wisdom I found:
“March. May has no abbreviation”
“I think it means May 11. Unless it’s on something that lasts a long time, like toothpaste. Then it’s probably 2011”
“march” (wrong, the right answer, I found later, was “May 10, 2011”)
“most likely March cuz May is so short they can just put the full month”
“I believe it’s May… I think March would be Mar”
I finally started ignoring any result coming from Yahoo! and found the definitive right answer: the format NN AA NN is a Canadian thing–I live in Toronto–and it’s the doing of the Canadian Food Inspection Agency. You can find the whole reference here. Apparently, to get to a month abbreviation that works both in English and French, that government agency decided to use “monthly bilingual symbols.” The problem is, if you don’t know the context, and if you are not accustomed to that convention, you might mistakenly assume that MA is March, JN is June, or that the two numbers at the beginning are the day, not the year. When it comes to food safety, relying on a standard that is easily subject to misinterpretation is something that you probably would like to avoid.
On the other side of this spectrum, the product reviews posted at Amazon are typically very reliable. Amazon reveals a lot of information about the reviewers, such as “real name,” their other reviews, the “verified purchase” stamp. Also, many filtering and ranking mechanisms are provided, such as the ability for other users to comment on reviews, vote for helpfulness, and say if a comment added to the discussion, or it’s abusive, or if a given reviewer should be ignored.
Unfortunately, Amazon is the exception, not the rule, one of the few sites out there where everybody knows when you are a dog. Twitter’s verified accounts seemed promising, but since they closed the program to the regular public, unless you are a celebrity, you are out-of-luck proving that you are not the person behind that account with your name and your photo. Of course, sometimes having a verified account may play against you, like Rep. Anthony Weiner found out in the last few weeks.
Reflecting over the low quality of information generally available, I concede that skeptics have reasons to not hop into the social media bandwagon mindlessly. But what we are really observing is just an amplification phenomenon, and a moment in time that many decades from now will be seen as the infancy of social technologies.
Since the first pieces of “persistent” content started being produced as rough drawings in some pre-historic cave thousands of years ago, the bad outnumbered the good by orders of magnitude. Creating good content is the exception, and social media amplifies all kinds of content. In part, there are lots of bad Yahoo! Answers because we always had a high degree of disinformation in the world. The only difference is that that disinformation can be easily spread, but that also applies to the good content.
On top of that, the same way natural ecosystems are in a constant state of imbalance but trend towards an equilibrium, information ecosystems will find themselves in apparent disarray from time to time. The original Yahoo! Search, editorialized by real people, once dominated the Internet. It soon became inefficient, and then the PageRank-driven Google search took over. It worked really well for several years, but it’s now also showing its age. Better filters will be developed to overcome the current deficiencies, and this battle will never end. The dynamic between quality of content and quality of filters will perpetually behave like a pendulum, as they always had.
Is this the age of disinformation? Yes, but no more than any other in the past. The fact that, by producing more content in general, we also increase the quantity of good content, should make us optimistic that we are better off today than we were yesterday. If the cost of coming up with one more Mozart is to produce thousands of Salieris, so be it: we may end up finding that Salieris are not that bad after all.
