Once you know who people are, you can start making claims against that.

The wonky thing about the social graph is that it’s a reciprocal claim between two nodes.

This is opposed to more typical identify claims, where one node makes claims about another, like when the State of California says “This is Jeff Nolan, and according to the state of california, he’s allowed to drive (claim), he’s older than 21 (claim), and he lives in XYZ (claim).”

What’s going to be amazing, is when Facebook finally starts syndicating identity out to other parts of the web (in a way eBay could have, but never did), and then starts selling user-centric advertising (i.e., whether I’m on MySpace or Jeff’s blog, or NYTimes.com, the advertising enginer will know it’s useless to pitch me ads about lingerie or cruise ship vacations, but will rather put the same kind of high-CPM ads I see on Economist.com and in Harvard Busienss Review) against that identity in a way that Tacoda, DoubleClick, and Yahoo could only dream of.

Then it’ll be some amazing stuff.

Of course, that part doesn’t even really need the social graph component to work, but Jeff’s point about “rich get richer” with the news feed is a great one. In that respsect, the algorithm that governs what makes it into my FB feed, will soon govern as much attention day to day as Google’s PageRank did four years ago.

As i’ve pointed out to some friends of mine, Facebook commands as much user attention every three weeks as a Superbowl (based on some rough numbers). How they direct that attention is going to make serious waves in the online and offline world.