The Last Podcast

I was thinking about the ideas behind Google’s Social Graph API and considering how this could actually work:

Here’s how it works: we crawl the Web to find publicly declared relationships between people’s accounts, just like Google crawls the Web for links between pages. But instead of returning links to HTML documents, the API returns JSON data structures representing the social relationships we discovered from all the XFN and FOAF. When a user signs up for your app, you can use the API to remind them who they’ve said they’re friends with on other sites and ask them if they want to be friends on your new site.

Problem is, at this point, very little of this data is actually explicitly expressed. Even if most social networks make these relationships explicit through XFN and other mechanisms, it would seem unlikely that a lot of people will go back and tag their old links just to fit into Google’s scheme.

I agree with Jeff Jarvis that if Google could actually figure out a way to create this graph through some advanced algorithms, it would be a killer feature.

For now, though, it is going to be mostly restricted to social networks, where relationships are already explicitly defined.

However, as I have said before, social networks are just the internet with training wheels. If anybody wants to figure out my real social network, they would just have to look at my blog. If Google could make more sense out of that data, it would be a far more interesting product. As it is right now, it actually knows absolutely nothing about me.