He mentions how Doc Searls describes Intention as 'the buy side of the Attention Economy'.
If Attention is all about sell-side, then intention is about buyers (the folks with the money) describing what it is that they want to pay attention to.
This discussion has crystallized another sponsoring thought for Touchstone that I've had floating around in the back of my mind.
Once our Attention Data is stored in a standardized way (OPML or Attention.xml or as part of what we call your 'I-AM Profile') then you need tools to start acting on it. Companies, I'm sure, will find plenty of ways to capitalize on the information for their own up-selling and cross-selling purposes, but I am more interested in personal tools that put the power back in a user's hands... in my hands.
Think of it as stored potential. The next generation of Attention or Intention tools will need to deliver value on that data by, in Touchstone's case, creating a model for what you care about to give you a filtered heads-up-display while you work.
Other value might be derived through recommendation systems, watch lists, targeted advertising etc.
So let me summarize all this as it stands now for my own purposes and see if I have it right (if not feel free to let me know!)
All these words, Intention, Gestures, Attention - they are just different parts of the same problem domain. A problem that has two separate user groups.
Group A wants your time to pitch a deal, and your money for their bottom line.
Group B wants more time and better information to make buying and living decisions.
Attention Data is a profile of what you care about (in our case based on implicit and explicit historic information that we get from the user).
Gestures are part of an Attention Profile (as far as I'm concerned anyway) but include more subtle information. I really think that gestures should somehow be stored right along with your other Attention Data and for Touchstone I think they will.
For example, whereas a click-stream only provides a list of links that a user has visited, gestures suggest the reasons for those clicks - maybe the user hated the site they visited and never wants to see that sort of information again.
Intentions, unlike Attention Data, are indications of future interest by a user. They might be determined by what a user has done (their Attention Data) and is doing (I'm at work right now) combined with some sort of implicit predictive model or explicit tuning interface.
Seth Goldstien calls 'Intentions' a 'Promise to Pay Attention' or PPA.