The inevitable Twitter post

I’ve been twittering for a couple of months now. I’m not actually addicted (I could give it up any time) but I’m up to a couple of hundred updates and I’ve got two separate accounts going. I suppose you’ve been reading about it all over the place so I’ll stick to what I’ve learnt:

It’s a micropublishing platform. My own mini twitter venture, Listen With Bowbrick, teaches me it’s a great way to describe/curate/link to media resources and that these resources ought to be ‘in-line’. I mean reviewing theatre performances or restaurants might work but really you want to be quickly describing and linking to stuff your followers can view or listen to or interact with right now. Listen With Bowbrick points at Real streams of radio shows, for instance.

It’s a universal command line. It’s early days but we’re already seeing people hooking up twitter IDs to web services of various kinds. A news or weather service can just passively message followers or, more interesting, followers could ‘direct message’ a twitter ID to trigger a personalised response. I don’t see any difficulty at all in grafting an economic model onto this kind of service either, especially since mobile operators are already making money from twitter via txt revenues. In fact, I wonder if twitter is the kind of service that has such universal, cross-network appeal that it might start to bust up the operators’ partial monopolies.

It’s hyper-constrained: 140 characters (the 160 characters of a txt message minus 20 for your twitter ID) and that’s it. No pictures, no sound, even URLs are difficult (hence the heavy reliance on tinyurl.com). This turns out to be a good thing: simplicity drives ubiquity, limitations promote creativity. There will be pressure to add stuff – imagine a shozu/twitter mashup or a twitter/flickr hybrid, or playlist distribution for music retailers, or a some kind of twitter plugin for MySpace. Blah blah blah. but I’m pretty sure this kind of pressure will be resisted for the time being and that new functionality will be confined to the ‘lingo’ (twitter’s command set) and to exposing more of the application via the API – that’s how you get a big network going.

It’s socially fascinating. I think the statistical social anthropologists will have a field day here: a sort of automated version of sitting watching twittervision. Parsing twitter updates should expose underlying patterns and trends (at least for participating communities) on a more-or-less real-time basis. Instant archaeology. I hope somebody’s keeping all this stuff. Mining twitter data will be a big area for grad school archaeologists in the 22nd Century.

Incidentally, it’s a pretty good place to do an informal survey or a poll (a twurvey or a twoll). I did one yesterday – five questions to Listen With Bowbrick followers that produced dozens of useful responses in ten minutes. Remarkable.