So when the invite landed in my inbox today I had to think hard - what the hell was Boxxet again? A down-market Porsche knock-off? Something in undergarments? A smaller species of low, leafy evergreen shrub for the border?
It turns out, Boxxet (pronounced "box set") is something entirely different - but strangely familiar. And fun.
Boxxet takes a few familiar concepts and repackages them in a clever new way. The concepts are social networking, news aggregation, learning engines and - my personal favorite - the vote button. Throw in a little wikipedia "specialist magnet" flavor, and a functional, "work in progress" UI, and you're about there.
When I finally logged in to Boxxet I immediately found an assortment of existing "box sets" - collections of digital "stuff" (news, pictures, locations, etc) arranged around a central topic, be that a sports team (I spend a blissful 15 minutes making the NY Yankess box set better today), a TV show, a company, whatever. The default page has a few of these box sets neatly laid out with a clear eye to print design principles - you feel like you're looking at a magazine cover.
Dig a little deeper and you discover that box sets are defined through a process of users both qualifying and disqualifying machine-searched data and explicitly including new content on their own. For a box set in a state of "becoming", a real person (comme moi) can click through a series of questions, in which you vote whether a news item, place, piece of gear, picture has anything to do with the topic of the box set.
For a "mature" box set, you can add new entries, blogs, and whatnot, and in principle (because I haven't fully experienced it yet) the learning engine must get "smarter" about what it pulls into the specific box set - because after a while, the NY Yankees box set I was working on seemed to look better than it did when I started.
In a nice touch, if you spend any time working on a box set, your user icon is shown as a "contributor", along with the most recent item you purged.
A few observations:
1. The UI is clearly a work-in-progress. Finding your way around isn't entirely intuitive, but at this point, that's not a bad thing. The forced-exploration that an early UI demands of you is a great way to learn.
2. Site performance is spotty. Sometimes you'll fly from screen to screen, other times you. . . just. . .wait. Success is a bitch.
3. The current roster of box sets is, shall we say, limited. Which wouldn't be bad were it not for the fact that. . .
4. You can't create your own box sets. Yet. The operators claim this is coming, and I'll take their word for it.
Tim O'Reilly had a lot of nice things to say about Boxxet at his conference back in early March - and I have to agree. To quote a comment I left on the founder's blog at Boxxet:
The exciting element I see behind the Boxxet concept is being able to capture - in a broad-based way - the combination of passion and knowledge that so many individuals carry around for their favorite topics. Then share it. It's a "benevolent cycle" - more readers create more specificity create more impressions drive better content which draws more readers. . .lather, rinse, repeat.
I hate to compare it to anything, but it's almost a live wikipedia, a snapshot of the digital gestalt around a particular topic, distilled and presented in a tight, relevant way. Charming and terribly addicting.
BTW - I have two invitations left - if you want in, drop me a private note.