Over the course of SXSWi, I’ve noticed some reoccurring questions pop up in panels regarding what developers can do to shield users from erroneous content published with Web 2.0 technologies. Most of these questions occurred in panels covering tagging and microformats, two now-and-future stalwarts of the social web.
In “Beyond Folksonomies: Knitting Tag Clouds for Grandma”, the first tagging panel, the worried questions began. How is tagging going to move from personal to social without a complete mess of duplicate tags and sloppily tagged information? Who’s going to decide on the “best” tag? How do we keep from having to look at four or more similar tags for the same information? What about Tag Spam?
In “Tagging 2.0”, more questions from the crowd seemed to allude that this tagging phenomena should be restrained, or guided in some way, despite solid explanations from Thomas Vander Wal and Rashmi Sinha that keeping tagging an individual activity can help the group in the long run.
Once again, in the Microformats panel today, the question was asked about handling bad info shared in a microformat, such as an hCalendar or an hCard. And although I believe that the question wasn’t asked in such a way to dissuade the use of microformats–in fact, devil’s advocate questions usually are geared at furthering the success of the technology under discussion–I think the answer for most of these questions is pretty simple:
Data is data, whether it’s lying to you or not.
The questions about erroneous information being spread are valid; however, I think that the conversation could be qualified a bit. We have to at least acknowledge the difference between the “human nature” issues and the technology issues.
The first time I heard about social web technologies, I was a bit skeptical that they would fall too hard on the side of factions and cult of personality problems. Sure, “Web Developer X” is smart and cool, and I really liked her CSS Zen Garden layout, but should I trust that she knows best about which Rails tutorial is the most accurate? And then, in speculation, I thought about a world where Fox News or CNN offers autocomplete tagging on their site, with a set vocabulary culled by an editorial staff.
But whether the worry is a complete democratic mess or an over-restricted or cliquish social web, the conversation could be expanded more to include the ideas of trust, context, and common sense in greater portions.
“Trust” was a keyword in all of the panels. Sure, “trust” might seem abstract and unquantifiable for our analytical coder-selves who are used to programming for EVERY exception, but we’ve been relying on this science for ages. Just the nature of markets, free press, and media consumption will bring some publishers to the forefront as trusted sources. As a consumer, you can put your trust in a source until they, at any time, break that trust.
The “Beyond Folksonomies” panel expanded on context, but I think context points to the bigger realm of “common sense”, a phrase it seems people are hesitant to throw around lest they sound reckless. For example, if you get an hCalendar or hCard from upcoming.org and it ends up be the one in one thousand that gets you to the party late or not at all, you’ll probably do your own kind of validation later by calling ahead. I mean, you got burned. Likewise, simple human pattern recognition can help someone know when they’re reading tag/content spam, sooner or later.
In the same way that our email spam filters have gotten more sophisticated, I’m sure that our Web 2.0 helper apps are going to become just as sophisticated. I’m betting that this group of people are going to be in the forefront of designing just these types of things.
So, instead of seeing the lack of restrictions on Web 2.0 for blocking erroneous data publishing, maybe we should see these things for what they are: age-old human interaction issues. Instead of being distracted by the possiblity of bad data being transferred, we can put our time and effort in making these technologies share data in the cleanest, most standard, and most open ways possible–whether or not the data is lying to us.
What do you all think? Am I too optimistic that people have the natural abilities to get by with the wild-west nature of Web 2.0 until it’s as boring as a javascript rollover?
Tuesday, March 14th, 2006