Chad told me that Tim O'Reilly posted about interestingness today. I've been contemplating another post about interestingness for a while, and I was glad Tim beat me to the punch! Some of this I hope to discuss at the Adaptive Path UX conference Wednesday too.
I've been starting out most talks that I've given lately by showing two examples of "user-generated content" back-to-back. First I show the numa-numa kid:
Then I say something like, "As amusing as this is... does anyone else find this kinda depressing? If stupid human tricks, pratfalls, fratboy pranks and skateboarding dogs are the future of media... let me off the bus!"
Then I say, "But fear not. This is also 'user-generated content'":
And I fire up a slideshow of the 100 most interesting photos on Flickr. It's hard to describe the unfailing impact that these photos have... they are alternately moving, funny, disturbing, provocative... I go on, "What's cool about these is that they are not only user-generated... They are also implicitly 'user-discovered'... It's not as if I spent a couple hours finding the 'good stuff' myself. The Flickr interestingness metric percolated the 'cream' to the top of the pile. By 'implicitly' I mean that there's no explicit 'rating system'. [I talk more about the value of implicit v. explicit means of deriving value here...] To be clear, Flickr is filled with plenty of junk. In fact, we like it that way. There's not just a low barrier to entry, there's virtually no barrier to entry. Got a camera? Bam! You're a 'photographer!'"
"So Flickr is a system that accommodates taking a 'worthless' picture of a hangnail, or a breathtaking Ansel Adams-like landscape. The cool thing is that while creating a frictionless environment that serves both scenarios, we can also determine which of the two is likely more 'interesting' to the community at large."
The ability to seperate wheat from chaff, or more accurately personally interesting from collectively interesting, is subtle but huge. And it does so without the use of link flux (i.e. PageRank) but rather uses 'in system' heuristics.
Usually after invoking the Flickr example, I transition to Y! Answers. If there's a complaint I hear about Y! Answers is that there's a lot of noise in the system. Admittedly, "Umm.. my boyfriend caught me sleeping with one of his best friends?", or "Why is the sky blue?", or "What's up?" do not necessarily resonate with the "expand all human knowledge" meme. But what's cool is that we can create a system that accommodates everything from the ridiculous to the sublime... but knows the difference between the two! (Or perhaps more accurately is taught the difference by millions of users.) This is the power of interestingness!
At this point I usually drop in a dry remark, "At Yahoo we have spent a fair amount of time and energy focusing on systems that are noisy, where anyone can say anything at anytime, etc. One of the most popular datasets and testbeds for these kinds of conditions is popularly known as... [prepare for punchline] the web... and we've been working on it for about a decade..." ;-)
I'm not sure why this post took on the flavor of a running commentary on my own talk, but that's how it came out!
I want to also remind folks that my relationship to the products I often invoke in this blog is best characterized as awed bystander. All hail Serguei, Yumio, Stewart, Tomi, etc!
