Ed Batista

A Little Feed-Cleaning

I'm not in love with Bloglines, but I still prefer it to the other feed readers (newsreaders, aggregators, whatever) I've tried in the past.  (The best reading experience, best mobile access, and decent sharing options overcome the improved-but-still-lousy feed management interface.)

But I've noticed that my list of subscriptions has grown somewhat stale lately--time for an update.  First to go is the folder structure I originally set up (1. Heavy Rotation, 2. Regular Spins, 3. Occasional Listens...) that provided a rough ranking of a) how important various authors were to me and/or b) how frequently I felt compelled to check their feeds.  That made a little more sense when I was managing my blogroll "by hand" and clicking through to visit sites "in person."  I'd visit the sites in the top folder several times a day, those in the second folder once daily, etc.

But now that I read most posts from within Bloglines and rely on the feeds themselves to tell me when there's something new, that structure's meaningless.  It's also useless to rely on such a static structure to convey each author's importance to me, because their meaning and value are so dynamic and so dependent not only on the quality, frequency and subject matter of their posts, but also on my own needs and schedule.  (What I really need is an intelligent system that will note what I'm actually reading and how often, and use that data both to tell others what I'm paying attention to and to recommend other content that I'd likely be interested in.  Maybe some people should get cracking and do something about that.)

So I'm scrapping the hierarchical folders, but I still want to group my subscriptions in clusters that correspond to my various feed-reading modes.  For now I'm going to go with AttentionTrust (for my day job), Other Technology, Management, Culture, and Personal.  Although I reserve the right to scrap the whole damn thing when a better system comes along.

UPDATE: So far, so...well, better, if not actually good.  Clustering feeds in themed folders is a much more effective way to deal with them--I can concentrate on a specific folder depending whether I'm working, focusing on personal interests, or just surfing.  The estimable Geof Morris has recommended FeedLounge (again)--I do like what I see (tag your feeds, slick interface), but I don't feel like paying $5 a month (that's not an unreasonable amount, but I'm not that unhappy with Bloglines.)  I also spent a little time getting reacquainted with Rojo and Google Reader, and they just don't work for me at all.  So Bloglines it is.

And to actually see the feeds I subscribe to, either click on "Feeds I Subscribe To" (brilliant, eh?) near the top of the left-hand sidebar, under my own subscription links, or scroll down near the bottom of the right-hand sidebar, where that list is displayed as a blogroll.  For anyone so inclined, the former link will take you to a Bloglines page where you can export these subscriptions as an OPML file.

tags: bloglines feeds rss

What's Your Audio Brand?

When there are too many books and blogs to read, too many movies to see, and--most importantly--too much music to listen to, a question of fundamental importance is "How do I find the good stuff--the writing and video and music--the content--that's going to be relevant and interesting to me?"  A lot of smart people are grappling with this problem, and at least a few of them work at Muzak.  And don't think "elevator music" when you hear that name--that's so 20th century.  Think "audio branding."

As David Owen's article in the April 10th issue of The New Yorker makes clear, Muzak is doing some very advanced thinking about matching people--and companies, for that matter--with relevant content:

Last March, at a trade show in Las Vegas, Muzak demonstrated audio branding on a large scale. The company’s simple rectangular booth had a decorative theme for each of the show’s three days: a red rose, a Martini, and an eight ball from a pool table. Dana McKelvey had designed a soundtrack for each day that was meant to evoke the theme musically. While the songs played—Etta James and Diana Krall for the rose, Frank Sinatra and dZihan & Kamien for the Martini, Blondie and Wilson Pickett for the eight ball—audio architects interviewed visitors, and used their answers to come up with a “personal audio imaging profile” for each one; later, back in Fort Mill, the audio architects used those profiles to create personalized CDs.

I went through the same imaging process during my visit to Fort Mill. Steven Pilker, a twenty-five-year-old audio architect—he had worked in a record store while in school at U.N.C. Charlotte and, when he graduated, was offered a job by a Muzak executive who had been a regular customer—asked me seven or eight questions, none of which had anything to do with music. (“When you’re not working, what do you like to do?” “If you could choose an actor / actress to star in your biographical movie, who would it be and why?”) A couple of weeks later, he sent me a six-song program, which contained nothing connected to what I think of as my main musical phenotype (“classic rock”); in fact, five of the six tracks were by artists I’d never heard of. Yet I liked all six very much, and later bought CDs by two of them (Sufjan Stevens and Jamie Lidell). Pilker’s selections aren’t definitive, of course; another audio architect surely could have had another take on my “brand.” But I was struck that Pilker, after spending very little time with me, had created an appealing musical program that was based on his sense of who I was, rather than on any direct examination of the music I actually listened to if left on my own.

This is really one of the promises of attention: Aggregating our interests and preferences and applying the right algorithms to that data in such a way that we can find the good stuff (and avoid the crap) on the basis of what we pay attention to (and what we ignore).  And it's not just (or even primarily) about efficiency--it's about discovery, it's about broadening our horizons, it's about a richer life.

(By the way, if you like the Owens piece, check out Barbara Hagenbaugh's USA Today article, which covered much the same ground 18 months ago.  I do believe Owens owes Hagenbaugh a debt of gratitude, to say the least.)

At PC Forum

Just a note to let folks know that I'm blogging from PC Forum, pimping AttentionTrust among the Masters of the Universe.  It's been a fascinating day-and-a-half so far, and you can keep up with me at the AttentionTrust blog.

tag: pcforum attentiontrust

I'm an Attention Pimp

According to Valleywag.  I suppose it's better than being an "attention whore," but it's the same business in the end.

PBwiki (Or, I'm Easy)

I got the following email today:

Just a quick reminder that tomorrow (Wednesday, February 15) is the last day to double your wiki's storage space for free. All you have to do is write about us online. More details at http://www.pbwiki.com/postonblog.

Now, I have no idea how much storage space a PBwiki has, but I seriously doubt I'm ever going to bump up against my current limits.  So I'm not really interested in pimping them just to get the space.

However, I do like these guys.  I like their name, I like that their service is easy to use and lives up to its promise, and I like the innovative, scrappy nature of this promotion.  So, here ya go:

Mmmmmm.  Peanut butter and jelly.

Are You (Meta)Experienced?

Even non-geeks like me are increasingly aware of and involved with our personal metadata.  The personal metadata I interact with most frequently is the information I share with Last.fm derived from the music I play on this laptop.  For the unitiated, anytime I play a song on iTunes (or stream a song from Last on their player), the song title, artist, and album information is uploaded to my Last profile.  The system also notes when a particular song was played and how often I play it.  All of this metadata is compiled by Last, and most of it is made available in a series of charts on their site, as well as via the Audioscrobbler Web Services API (for examples of the latter, here are my Top 50 Artists in plain text and XML.  Put that in your aggregator and smoke it.)

I'm focusing on Last because it's such a great example of what happens when user-friendly systems make it easy for us to record, store and analyze our metadata: We have meta-experiences.

Just as metadata is "data about data," a meta-experience is "an experience about an experience."  When I look at my Last charts, I'm having a meta-experience; the experience of comprehending and assessing that information is derived from a previous (or concurrent) experience--in this case, listening to a song.

And as systems like Last proliferate (and they will), and we're able to capture more and more of our personal metadata, our meta-experiences will become more explicit and meaningful to us.  We'll be more conscious of the presence (or absence) of the meta-experience at the moment of experience.

I notice this now when I'm playing music on a device other than this laptop--an iPod or my car stereo, if I'm out of the house, or the CD player hooked up to my stereo if the laptop's off and I don't feel like booting it up.  I'm oddly conscious of the fact that the music  isn't going anywhere but in my ear.  The metadata's simply disappearing, and I'm being deprived of a potential meta-experience.

Why do I care?  What impulses make my Last meta-experience (and potentially other meta-experiences) meaningful?  Meta-experiences document our experiences, and allow us to, well, experience them in entirely different ways.  They essentially add a whole new dimension to whatever we're doing.  For example, by sharing my musical metadata with Last, I have a much better understanding of which artists I listen to and when, and how those preferences can evolve over time.  I can exchange my metadata with other people through explicit and implicit gestures, both by inviting them (my "Friends"), as well as by letting Last match me with people whose tastes are similar to mine (my "Neighbors").

I could go on, but my point is simply this: Now that we've begun to capture and make use of our personal metadata, we're going to become increasingly conscious of the resulting meta-experiences.  And we're going to want more.  (Of course, all this has huge implications for the field of attention.)  More to come, I'm sure.

UPDATE: Putting theory into practice, here's a little graph of the number of songs I've played by the top 300 artists in my Last profile as of today.  (Look familiar?)  Congratulations--you are now  meta-experienced.

tags: attention attentiontrust last.fm metadata metaexperience meta-experience

Google Analytics on TypePad

I began running Google Analytics here this morning, and even with just a tiny amount of data to work with, I can tell it's going to be a useful tool.  I signed up for an account as soon as they were available last year (before they shut down open enrollment), but I heard there were problems caused by user demand overwhelming the service, so I never bothered to set it up.  I finally got around to it today, and I thought I'd describe how I set up Google Analytics on a TypePad site like this one, since it's an easy but not necessarily intuitive process:

1. Copy the relevant javascript from Google Analytics.
2. Create a new "Link" TypeList and configure the Notes to display as Text.
3. Add an item to the new TypeList and paste the javascript into the notes.
4. Add the TypeList to your site at the very bottom of the right-hand sidebar.  (I'm not certain, but it appears that this is the only location that will work, since it puts the javascript just before your site's </body> tag, as per Google's instructions.)

I should also note that it's important to filter out internal traffic (from your domain or IP address), particularly on low-traffic sites (uh, like this one).  Setting up these filters is pretty straightforward in Google Analytics, but you have to remember to format the filtered IP address properly.  According to their instructions:

Remember to use regular expressions when entering any IP address. For example, if the IP address to filter is:

192.168.1.1

then the IP address value will be

192\.168\.1\.1

That's it.  More to come when there's some data worth taking a look at.

tags: google analytics typepad

Last.fm: One of Us, One of Us

Thanks to a great suggestion from friend and colleague Ruby Sinreich, I've started an AttentionTrust group over at Last.fm.  Here's my hastily-scribbled raison d'etre:

This is a group for Last.fm users who are interested in making more effective use of their "attention data" (including, but not limited to, all the data we're sharing with Last). We're big fans of Last, and we love their service, but we'd also love to know what Last is planning to do with this data. We hope this group will prompt some interesting discussions between Last and its users.

I really do love Last, and I hope the folks running it realize that eager, engaged, heavy-duty users like us make the best evangelists.

I also anted up and subscribed, figuring that $3 a month was a very cheap way to show my support for the best working attention service out there (and that's what they are, whether they realize it or not.)

I'm obviously a fan, but a few nagging questions remain: What's happening to all of that attention data?  Can I get a copy?  Can I manage it in any way?  Who else has access to it?  What will happen to it if Last tanks?  I have no idea--and that bothers me.  I'd love to see Last join AttentionTrust and give users the ability to exert more control over our data.  What about it, Last?

If you're not a Last user, here's the bullet: 1) It's a plugin that allows you to record your iTunes metadata (which songs you play and when, favorite songs and artists, etc.)  2) It's a web service that allows you to analyze your own metadata as well as others', matches you up with "neighbors" who share your musical tastes, and recommends music you'd like.  3) It's an online community that allows you to create groups with common interests, communicate with fellow members, and keep music-related journals.  You can use the basic service for free, and if you get addicted like me, you can opt for the paid service and get a few extra goodies.

And if you are a Last user, join the AttentionTrust group and become...one of us, one us.

UPDATE: Answering some of my own questions above...

• What's happening to all of that attention data?  Almost all of Last's data is available via the Audioscrobbler Web Services API.  (Audioscrobbler was the original name for the Last plugin.  The two entities merged under the Last name a while back, and the Audioscrobbler site now serves to support the Last development community.)  The data is available under Creative Commons' Attribution - Non-Commercial - Share-Alike License.  The data is also available for commercial purposes, but you have to contact rj@last.fm to obtain a separate license.  (There's no info available on the site about commercial licensing terms, or who's obtained them.)
• Can I get a copy?  Kinda sorta.  You and anyone else adhering to the terms of the CC license can grab a range of feeds in .txt, .xml and .xspf formats.  Some feeds are related to a specific user profile, and others are related to specific songs, artists, etc.  But you can't grab all of your user data--for example, you can only get the top 50 artists, albums and tracks from a user profile, not the full lists.
• Can I manage it in any way?  By this I meant, "Can I delete any shitty music I don't like anymore, so that it doesn't screw up my recommendations?"  The answer appears to be no.
• Who else has access to it?  Anyone who adheres to the CC license, and (I'm assuming) anyone who successfully negotiates a commercial license with Last.  I want to emphasize that for me at least, this is really about transparency, not privacy.  I assume that everything I do online is being tracked, recorded, analyzed and monetized by someone, and that's the price I pay to insure that my actions, my preferences, my interests are attributed to me.  It's a way of expressing myself, really.  Anonymity has its uses, and like everyone I have personal data that I want to keep private, but I'm not worried about keeping my Last data private.  But I would like to know more about who else has access, and under what terms, so that I can make an informed decision about using Last.
• What will happen to it if Last tanks?  Still no idea.  And this is a big concern of mine.  I've invested a lot of time and energy amassing attention data in Last's system, and it's become a valuable resource to me.  I'd feel a lot better if I were capturing and storing my data on my own (in an open, portable format, of course), and then opting to share it with Last.

tags: last.fm attention attentiontrust

Wikipedia Edited, Dog Bites Man

Verne Kopytoff has a "What's Web 2.0?" article in today's SF Chronicle, and in general it's a fine primer on the concept, neither too credulous nor too cranky.  I particularly liked this quote from Dave Winer, which Kopytoff pulled from Winer's Scripting News:

Web 2.0 is a way for certain marketing people to claim they invented stuff that they didn't invent, without actually claiming they invented it.

Although Kopytoff doesn't include the full quote in the Chron, Winer's original post goes on to say:

[Web 2.0 is] the kind of double-talk marketing guys love.

In a sense people are right when they say it's another bubble. It's dishonest like the bubble was. Yet the technologies they're hyping are honest.

Winer's full quote gets at the essence of Web 2.0: a marketing buzzword that nevertheless represents something real. I understand Winer's resentment of marketing bullshit and appreciate his ability to cut to the heart of the matter, but some buzzwords which start out as pure marketing-driven hype gain traction because they effectively translate technical issues for non-geeks like me.  Of course, they're subject to abuse by unscrupulous flacks, but that doesn't make the terms themselves illegitimate.  (For more on buzzwords, Tim Oren had some interesting thoughts on the useful purposes they serve, which I excerpted last year.)

But Kopytoff goes astray when he backhands Wikipedia:

The reputation of Web 2.0 was recently sullied by Wikipedia, the online encyclopedia written by volunteers. Held up as a shining example of Web 2.0, it suffered a scandal after one of those amateur contributors rewrote history by accusing a former government official of being involved in the Kennedy assassinations. [emphasis mine]

Scandal? Really?  That strikes me as absurd. Crackpots screw around with Wikipedia entries every day.  That's the price of having an open system which anyone can edit--it's not called Wikipedia for nothing.  But errors are typically corrected within minutes. And debates will rage over controversial and highly politicized topics where a definitive answer can't be found, but that's as true within Wikipedia and other user-supported systems as it is anywhere.

This isn't a scandal--it isn't even a story.  It's the way Web 2.0 (sorry, Dave) works.  It's the wisdom of crowds. By allowing everyone--no matter what their authority or credentials (or lack of same)--to contribute to the intellectual marketplace that is the Web, we guarantee the system's fluidity and maximize accuracy at the macro level.  Of course we sacrifice accuracy at the micro level--with no editorial apparatus constraining the system, anything you read at any given moment could be total bullshit.  But in a free, open and fluid intellectual market, bad ideas will eventually be driven out by better ones--and sooner rather than later.

Perhaps it's not surprising that a newspaper article completely misses this point, but it is sad.  I love newspapers, and I don't want to see them die--I want to see them adapt and thrive.  I'm hopeful that publishers and journalists--the smart ones, anyway--will find a way to do that.

tags: web 2.0 wikipedia verne kopytoff dave winer

Wayfaring

Wayfaring is a neat little mapping mashup that allows you to use Google Maps data to create your own web maps, with routes, notes, and "waypoints," i.e. those upside-down teardrop shapes.  Great concept, slick Web 2.0-ish site, but the execution is clunkier than I expected.  Or maybe I'm just stupid.  But I do love the idea and hope to get some real use out of it.  You can make your maps public or private, allow others to edit them or lock them down.  One really nice feature is that it'll tell you the mileage of your route--very helpful when Google Maps directions just won't cut it.  The pic above?  My sorry-ass 3-mile jog in Golden Gate Park--and even that feels like a long ways these days.

tags: maps mapping mashups wayfaring