I'm almost caught up with my news, and I ran across this looooong article over at FT.com about the hazards of being always connected 24/7, tethered to our mobiles among other technologies. Or that's what it's generally about, it sort of rambles a bit, and quotes everyone from Mizuko Ito to Howard Rheingold to Ray Ozzie. (The article is now password protected, so bop over to bugmenot.com to get one if you want to read it all).
This bit at the end was really interesting to me, and related back to that Agents post I wrote the other day:
The current Big Idea in technology circles for handing back control is to somehow embed your personal preferences in the technology in a way that makes it respond to how you want to live your life. The phone, for instance, will be smart enough to know when you can be interrupted, and when to leave you alone. It will know, on any particular day, whether to put through a call from your mother immediately, or whether to send her straight to voice-mail.
By learning from the preferences of your close networks of family and friends, it will also have an idea of the sort of things that are likely to interest you. In future, we will all be part of â€œself-organising peer groups that provide ways to filter things,â€ says [Dick Lampman, director of Hewlett-Packardâ€™s research labs]. Only communications or media that have a place in this more closely defined social realm will be able to find their way on to your personal communicator. â€œPeople want more control: you should be able to build your own profile and use that to qualify the things that come to you,â€ says Lampman.
So I generally agree, but the question is, how do you implement that control? Having a device "smart enough" to know when you can be interrupted is an interesting notion, but some poor coder somewhere has to program that intelligence, which is hard if not impossible to do from thin air. Like I said in my Agents post, it'd be much better if we could find a way of meeting the problem half way.
We have a bazillion things that could bother us at any moment, right? If you think of just RSS feeds of news that may interest you and just one form of communication - say email - you've got several hundred opportunities to interrupt you on a daily basis. In theory, a "smart" filtering system is supposed to know when you want to be interrupted or not, right, and also make other value judgements about which information and interruptions to route to which destination, but I have my doubts about this happening any time soon. My TiVo still thinks I'm a Gay Republican, so excuse me if I'm a little gunshy about my devices doing any sort of value judgements.
So here's a thought - sort of continuing the Bayesian Filter ideas from last time - why don't we attach a point system to things that may interest us? "Interest Points", I guess for a lack of a better term. Though "points" sort of suck as a metaphor, I'd much rather have a system as simple as TiVo's thumbs up/thumbs down system, but you get the idea. You rank each thing that you want to know about during the day as to their importance, then you set a minimum point level when you will be notified. As things happen in the world, points will start to build up, but that's fine - not getting interrupted every 5 minutes is the goal - when the time comes, you'll deal with the interruptions as a whole.
It would be like having a personal assistant taking messages for you, and keeping up with news, and then coming in every once in a while and giving you a summary of all the stuff that you need to attend to now that you're free.
Let me give some more specific examples: Say your threshold is 100 points. Okay. Now, you set up an RSS Search subscription for a stock symbol you want to keep track of, but not necessarily every second of the day. You'd give every item that comes through the RSS just 1 point. Every email that comes to your personal inbox gets 5 points, and every work email gets 10 points. Then we can rank people as well: Your boss and your mom get a +25 bonus, other coworkers get +5 points, and *those* coworkers get -10 points. Maybe that's just during business hours. During the evenings, all your coworkers get 0 points except your boss. You can keep going. Every thing that could possibly interest you gets a point ranking.
Okay, so now you go about your day with no worries about what you're missing! As soon as the Interest Points pass your limit, you get pinged with a summary of what's waiting for you. Then you can change contexts once, process the pending information flow (which could be news, messages, tasks, etc.), and then go back to what you were originally doing. This gets pretty close to that Big Idea, no?
I can imagine the secret to this will be to find a way of automating the rankings so you can just have an agent watch your activities for a while and figure out the rankings by itself. But my TiVo does that and it sucks at it... So maybe instead of a machine doing the rankings, I can have people do it instead (in a way)? I guess that's where this "close network" comes into play. Though I'm not sure exactly how my friends would be able to help guage my personal preferences like that, but I'm sure there must be some way of tapping some collective intelligence there to help filter my world.