Flickr just announced some really interesting new features that gleans photos for what they call “interestingness.” In short, it essentially studies newly uploaded photos and determines which ones are somewhat…interesting.
Of course, the big question is: how can software determine what is “interesting”? We can of course ask people what is interesting with features like voting and the like, but that’s disruptive. What the boys and girls at Flickr have done is effectively watch how people use their services, and thoughtfully determine what is interesting based on the actions and behaviors of their users.
A while back, I posted an entry entitled Stealing Metadata. The idea behind that article is that, while tagging and the like are great for gathering metadata, from the perspective of a user’s experience, it is disruptive. What we can do, alternatively, is carefully watch what users do, crunch that information and draw some educated conclusions. Their actions and decisions often provide insight into their motivations. Flickr seems to have done exactly that.
As an experience designer, I think this is the far superior approach to gathering important metadata. Rather than asking users to stop, “check in” with some information, and then continue, the flow of their experience is not disrupted.
On the flip side, this presents new and interesting challenges to the world of data warehousing and usage monitoring. To attain this kind of information, systems will need to do more than just log information. I think the technical capabilities are clearly there. The task at hand is on the design side: what are we interested in? what actions hint towards a certain conclusion? how much weight do we give certain actions? This is a higher level of sophistication than the typical logging capabilities that we’re accustomed to today.
In the end, the payoff is huge. To the casual user, Flickr and other services that provide this kind of information take on an illusion of intelligence and awareness that is truly captivating. Come to think of it, most great technologies are shaped this way: a simple, intuitive solution that hides away the complex work behind it. “Show me some interesting photos.” Perfectly simple.