Back in the days when the web was young, and I was younger (as were Steve and Ivan as this front cover of Wired will attest), and I had the rather excellent email address firstname.lastname@example.org, I got rather het up about something Nicholas Negroponte wrote in Wired. Unbelievably they published my rather precocious and slightly wanky letter (I had just left academia at the time and was using my recently ex-scientist nom de plume, Christopher J. Thorpe).
Fast forward about 14 and a bit years and I was lucky enough to sit in on a lunchtime talk at The Guardian by Clay Shirky. It was totally fascinating, pretty much everything he was saying rang true, but at the same time my old alarm bells about serendipity were clanging away.
As you atomise news content as the internet does and APIs do more declaratively, the bundle of serendipity that is a printed newspaper disappears. He was talking about streams of content from friends being serendipitous but that didn’t quite sit right so I tried to work in my mind the sort of question which would sum up what I felt and here I’ll paraphrase it.
"The serendipity we see in content from friends isn’t pure serendipity in the way that a physical newspaper is, which contains things you know you want, things you don’t know you want yet but now you do, things you don’t want and things you just read as they’re there. How do you rebuild this without just doing chatroulette for news"
Cue mirth from room and Clay suggesting we should do that, and probably would be coding it right away, before describing another model where you look at the content shared by friends of friends of friends and people between you and them as another mechanism for generating a curated serendipity. This would generate the sort of signal to noise ratio of a newspaper, you read The Guardian as it has a tone and a window on the world you like and agree with, it brings you things you like and some other things.
Well fast forward a few hours and my rather clever colleague at The Guardian, Daniel Vydra had made the simpler one; The Random Guardian. And it’s lovely and brings you things in an almost John Peel like curated randomness and mayhem type way.
I loved Paul Carvill’s comment…
@jaggeree it’s no chatroulette; i haven’t seen any penises yet
Give it time. In the archive there are only 2888 search items out of over a million-and-a-bit which match that term, therefore the statistical likelihood of an article on penises appearing in a random selection of today’s stories is quite small. And also as Tom Hume so sagely pointed out in his analysis of The Guardian’s swear word patterns using the API - Cock is flat. (excellent discussion here)
I also really liked what Tim Davies said:
http://random-guardian.appspot.com/ really is strangely addictive. Random button for the iPhone app? Then I’d never get anything done.
For me one of the real reasons why the iPhone app works so well is that it leads you into parts of the newspaper the more directed browsing and search don’t. Likewise Dan Catt and Meg Pickard's wonderful Zeitgeist does similar things.
For people who think that paywalls or the iPad are the things that will be the magic bullet that saves the newspaper industry from falling over a cliff I think there’s an easier answer; more random/serendipity please. If you take me to unknown places I’ll read more and I’ll spend more time, be more engaged, you can target me better and I’ll love you and buy things.
Bring me wonder and magic and I’ll love you forever. I still buy Wired for that very reason, although for me the jury is still out on the CD-ROM reinvented iPad demo they did the other day. I won’t link to it, you’ll only be disappointed, it’s not the future, this is.