wanderingstan

I’m working this weekend on a little project inspired by Fred Wilson’s Blog Scrobbler. The idea is to keep a published record of blog posts that you view.

Pretty simple. But it’s got me thinking about a much bigger idea, which is starting to come into focus.

You see, even a few years ago this weekend’s project would entail setting up a dedicated server to hold the data, creating a web interface to display it, managing a user login/registration system, creating an API for accessing it, and much more. That’s hard.

But I can do the project in a weekend because Twitter has done all the work for me: it will be the “data bucket”. I don’t have to set up any servers or write any user registration systems. I just have to collect my little bits of data and ship it off to twitter where it will be stored and published for others to consume.

There is something profound about this.

Already, twitter itself is primary a storage and distribution system, NOT a website. People interact with it via numerous clients and phones. The fact that there is a web interface to view and enter tweets is not the selling point.

I see the possibility for a service that provides huge, generic, user data storage. Users could use any applications they like for adding data (like I use tinytwitter on my blackberry) or for viewing data (like I use tweetdeck on my laptop). Except that “statuses” would be only one of many types of data that could be stored and published: photos, posts, friend lists, song scrobbles, blog scrobbles, current weight, morale, classified listings, emails, etc.

Your hard drive for the web.

There could then be an ecosystem of services like gnip that continuously crawl this data and republish it for consumption by other services. New breeds of services could emerge from mining this data, in the same way the Summize creating a new type of search application by mining twitter’s data.

More importantly, this would allow a user to keep their data (or at least the gold-standard version of their data) in one place. Presently, using any web service means that your data is stored only on *their* servers. This means that your data is spread across, e.g. gmail for email, twitter for tweets, flickr for photos, zoho for docs, etc.

A geeky way to look at it is to again consider the MVC (model, view, controller) paradigm: Twitter provides a model (storage) for tweets, a simple view (the web or incoming sms), and controllers (web form or outgoing-sms). But twitter allows you to use other views and controllers (e.g. tinytwitter or tweetdeck).

If we can separate the storage of our data from the viewing and editing of it, more cool projects like this weekend’s “Blog Scrobbler” –and whole new breeds of companies– would be possible.

UPDATE: Learned from @indiekid and @w1redone over dinner that (a) others were using this convention two months earlier in California (doh!) and (b) there is a site dedicated to tracking hashtags: hashtags.org. Cool that Jason and I independently hit upon the same symbol to indicate locations.

The web has a new convention for usernames, could we use one for locations? And why do these things catch on?

In Egyptian Hieroglyphics, names were always circled–an early example of semantic tagging.

Twitter users created the ‘@username’ convention on their own almost as soon as the service started, and it has now spread beyond twitter. Usernames are prefixed with a @ symbol to indicate that you’re talking about a person, and the system can turn that name into a link.

In December Jason Lange and I hatched the idea of using hash marks to indicate locations. We settled on the # symbol because (a) @ is already taken (b) it sorta looks like grid lines and (c) it is used URL’s to indicate a location on a page. Of course, we twittered our new idea:

Brainstorming like crazy with @susqhanamj at
#trilogy.  - 09:15 PM December 07, 2007

At dinner last night, I heard the folks at Twitter HQ are talking about the this geo-hash convention.

But what is it good for? Imagine if each location had it’s own page. So in my tweet above, you could click through to a #trilogy page for the Trilogy Wine Bar in Boulder. You could see other people who have talked about it and get a sense for the sort of people who go there.

Who creates content for these pages? Twitter could allow businesses to create their own accounts (as Facebook now does), but that wouldn’t cover things like parks and geographic landmarks. So better yet: make locations into Wiki pages that anyone can edit.

No word from Egyptologists on what symbols they used for locations…

Twitter has aptly been called “microblogging”. In a previous post, I wondered what the next step in self-revelation would be. At the time I suggested “Maybe something that reads your physiological state directly.”

I overshot the mark.

Twitter answers the question, “What are you doing right now?” But if you’re online (as you are at this moment), that’s a question that can be answered with a single URL. The answer for you right now is:
http://wanderingstan.com/2007-03-27/microblogging_to_implicit_blogging

A simple listing of pages you visit constitutes a very micro and very automatic blog.

Consider these new companies and their taglines:

• atten.tv – “Watch what people are watching.”
• Cluztr – “Share your linkstream with your friends.”
• Me.dium – “Reveal the hidden world of people and activity behind your browser.

Is this form of implicit blogging (nanoblogging?) interesting to anyone? It’s hard to say. Even the 140 characters of twitter reveal some personality of the twitterer. Nothing like that in a stream of URL’s.

But as Twitter-founder Evan is quick to point out, the same “Who will care?” doubts were raised in the initial days of Blogging and are now seen on Twitter. In the long tail of attention, perhaps even a stream of URL’s from the right personality can find an audience.

Twitterment is a “search engine and buzz tracking system” released today from the University of Maryland uBiquity research group. (offical announcement)

For example, you can see that people talk about “coffee” and “sleep” at opposite times of the day, with the crosspoint coming at 5am.

Or you can ponder why people talk about “Chipotle” mostly on Fridays, and usually around 3pm EST.

I have a lot more respect for Twitter after hearing founder Evan Williams speak at ICWSM.

At the same conference I met Akshay Java, a young researcher overflowing with good ideas. (When we were talking at dinner the ideas were flowing so fast and loud that other people thought we were drunk.) Twitterment is his latest project. Be sure to also see Feeds that Matter, his previous project.

It’s amazing how quickly twitter has spawned an ecosystem of applications. A great case study of Open Data.

The ICWSM conference opened today with a talk by Evan Williams of Twitter, and formerly of Blogger fame.

My takeaways from his talk:

Functionality is becoming a commodity.

You used to use a service because it did what you need. Now there are a million services offering the same functionality. (Think of all the Delicious, Digg, YouTube, etc.. clones.)

Twitter is about Ambient Intimacy

Great phrase to describe it.

The more casual a social media system is, the closer the ratio of producers to consumers.

This is rich. So blogging made it easier for people to produce content, and a fair number of blog readers converted into blog producers. But blogging isn’t that easy–HTML, permalinks, wordpress installs…not to mention the effort of forming well structured paragraphs. Twitter is even easier.

I suppose this works in the opposite extreme as well: The ratio of TV show viewers TV producers is huge. But video sites are making it easier, thus lowering that average.

And at the other extreme? Hard to imagine anything easier than typing 140 characters. Maybe something that reads your physiological state directly (e.g. level of arousal). Or something that involves choosing nubmers between 1 and 10 like Buster McLeod’s Morale-O-Meter. I’ve been tracking my morale for a while as you can see here:

More on the conference as I find time…