The Future of Digg
I was lucky enough to be at FOWA today and to hear Kevin Rose speak about the future of digg.
I was unlucky enough for my Tablet PC to crash and lose my notes from the session [my own fault]. What follows is from memory. Forgive the consequent ‘notey-ness’.
Digg will support OpenID. The emergent portable, open ID system recently given support from Microsoft, AOL and Yahoo! will also work on digg.
Digg will never pay moderators to check for false/lame stories. Rose doesn’t believe that is a workable - or rather, scalable - model for the system. The moderation, he thinks, needs to mirror the community to have any meaning. “Policemen every five steps won’t make any difference,” as Rose (probably) said.
The discussion pages will allow the upload of links and ‘other’ media to support stories. This will also be a part of the story validation system, whereby proponents and opponents of a particular claim can upload evidence to show this.
Digg can already spot spammers easily using their statistical data. 100 user accounts that all agree with each other on everything is easily spotted, for example. He showed us these rogue stories appearing on a graph available internally to the company and swiftly being buried. Rose did not state whether the managers of digg take any action in these circumstances. My impression was that he was saying that they do not, because users are more than able to detect spam stories and bury them.
Digg is working closely on making users’ data work better for them. In the not-too-distant future, it will be able to recommend friends to you on the basis of your common likes and dislikes, and also upon their geographical location. It will also be able to personalise the ‘upcoming stories’ queue to show stories you are most likely to be interested in.
On the flip side of that, digg will be able to personalise the people you interact with and do not. Rose hopes that this will counteract the effect of “900,000 people being in the same room and, of course, they aren’t going to get on”. Or, as a delegate put it: “people on digg are assholes”.
Rose confirmed that digg is actively mining users’ attention data in order to implement these changes. He professed ignorance of APML, but “there are a couple of engineers deeply into this stuff” back at digg HQ.
Digg will imminently introduce a ’smart’ digg-this button for blog and website owners. A single line of JavaScript will detect whether or not a story has been dugg already and advertise the number of diggs a post has received.
A Flash toolkit is in the works that will allow site owners considerable flexibility over displaying data from digg on their own sites.
Generally speaking, great conference so far, but lack of internet access today means I’m still catching up. More news tomorrow, I hope.