Steve Gillmor’s GestureLab

I always thought the day would come where I finally had enough bullshit from the various companies trying to game the attention model. This afternoon I had a long (2.5) hour chat with Loren Feldman. As some of you may have noticed, Loren is not a big fan of the non-profit AttentionTrust, which he called a joke in a recent video post. In the chat today, I found some common ground around the notion that even though the Trust was in Loren’s opinion hopelessly conflicted by co-founder Seth Goldstein’s participation in a number of attention startups, the basic idea of attention was in fact a powerful trigger for what I believe is the successor to the page view model.

What’s wrong with the page view model, you might ask? Nothing, if you have unlimited time for cruising the Net for actionable information. Attention for me has never been about getting attention, selling attention, or any of the Attention Economy constructs put forth by Goldstein, Goldhaber, Dyson, O’Reilly, and others. The last two years of conferences on the subject have provided precisely nothing of interest for me beyond my original impulse: harnessing the power of each person’s “karass” to reduce noise and prioritize information beyond the algorithms of page rank, techmeme, and RSS rivers or mail UI interfaces.

That was, and is, the goal of my work in this arena. Getting there brings politics into play, and here I’ve tried to emulate the strategies and innovative obstinacy of Dave Winer as practiced in the RSS wars. Attention.xml, the Attention Recorder, GestureBank — each phase of this campaign has been designed and supported by me only as a vehicle for abstracting out the notion that users can, and will, take charge of their implicit behavior, or gestures as I mandate them. As Loren says in today’s chat, the signals of an endowed and hand-raising few are inherently more interesting to me (us) than the mass culture of enticed reactors.

At Rafat Ali and Staci D. Kramer’s recent conference, I met and sat next to Jason Calacanis’ Weblogs-Inc. partner Brian Alvey. At dinner one night he told me about a SXSW panel he attended that dealt with a mashup of attention and identity themes. From his description of the event and a little research, I deduced that the thorough summary of my thinking on gestures was delivered by AttentionTrust board member Mary Hodder.

A few days ago Brian sent me a pointer to a New York Times article on identity that quoted Kaliya Hamlin sounding a similar note. These two datapoints, plus a long conversation I had with Kim Cameron at the last Internet Identity Workshop (and a follow-up this week), convince me that we have internalized sufficiently the fundamentals of what I find compelling about attention. Brian wonders why I am not cited in these discussions; I wonder why it took me so long to get over the embarrassment of being shunted aside and get back on track.

Rather than pointing out the deficiencies of the various attention startups surfacing these days, here is the lens through which I view the credibility from a user-in-charge perspective:

• Who owns the data?
• How is the data distributed should the service-offering owner be bought?
• Is data being collected in return for one service and distributed for another?
• Can data be resold or laundered to remove user control over distribution?

I’m sure there are more granular delineations of data flow, but in practice I can usually spot questionable terms of service in About statements or privacy policies from these bullet points. I don’t mean to suggest that these are legal violations, as most attention services make reasonably clear what you get as service in return for harvesting your data. Much of this is common sense: you sign up for Gmail, you understand you’re getting a so-called free service in return for sharing your behavior with Google. Twitter: a soapbox for a social media map of your cosmos.

But I’m less concerned with the legal trail of these clicks than the utility moving forward of the accumulating data and what directional data it holds within its social map. In the wake of the collapse of the operating/Office system, I am looking for the network effects that derive from orchestrating attention signals to accelerate discovery, incent affinity-derived content creation, and indemnify contributed data from pollution by laundering and then merging with formerly clean user-managed data. Whether the attention service has a legal responsibility to handle collected data with the user’s interest in mind is not crucial; how the attention service appears credible to users is.

Put simply, GestureBank was conceived and implemented as a mechanism to insure initial and continued contribution of anonymous data to a user-controlled aggregated pool. Such a data pool survives the most rigorous tests of data flow and integrity. Data queries and services produced by affinity services aligned with GestureBank must not be corrupted by co-mingling with less rigorous attention collection strategies.

To mandate this “clean” requirement, GestureBank will release a GBX2 Firefox add-in that adds header data to each request (click) that specifies the user’s issuance of a specific license to receivers of that request, allowing use of that data as long as the principles of user control remain in effect. To reiterate, this is not a legal requirement, although it may ultimately pass that test, but a direct communication of the user’s intent and a lens through which the user subsequently can view the credibility and integrity of entities who seek to use that data subsequently.

With this stamp of user-control in place, I can now go back to working with like-minded affinity groups without concerning myself with conflicts of interest or attacks from those who would obfuscate the attention opportunity with complexity. If you do your own homework and examine attention service offerings, you can quickly assess the validity of the user contract. Then it is up to you to decide how, or if, to trust the integrity of the service and its shared data.

If GBX2 is in the chain, you can then go to the successors and query them as to how they support the terms of the user license. If they ignore the data, that presents an opportunity to ask why? If they refuse to provide their services to GBX2 users, that speaks even louder. Conversely, those services who respect the user license will be rewarded in turn.

Wednesday, May 16th, 2007 at 9:06 pm.
Filed under Uncategorized.
Subscribe to RSS 2.0.
Leave a comment or trackback.