--- Thinking out loud ---<p>There is yet a thing to come of twitter:<p>I design big systems; HQ campus networks, Hospital networks and systems, RFID/RTLS tracking systems etc...<p>Back at Lockheed in 2002 we talked about this untapped M2M information layer that was just becoming accessible.<p>We imagined that there was a ton of data, contextual and revealing about what human actions were, that was happening and going to exponentially happen at a machine to machine level...<p>this device is here, in this state, with this sensor data and tethered to this owner... etc...<p>Twitter is kind of this sub-meta layer of information that has emerged - but not about machines -- about the human psyche...<p>there is a lot to be said for the 140 limit -- think of it as a protocol spec for the machine enabled telepathy that is the internet.<p>take that paradigm and apply it to the M2M idea previously stated: use twitter as a platform (middle ware, much like Emergin) to deliver short context messages between social systems, in an automated way) and then mine the fuck out of it for intent, interest, trend, etc...<p>This is like using it to data(sift) mine the information users put into the twitdome to determine such things as stock prices, national satisfaction on [issue], etc...<p>the beauty is that much can be gleaned DUE to the fact that the inputs are of a known, limited size.<p>I theorize THIS is exactly why the library of congress has chosen/forced? the archiving of all tweets -- as it is immensely valuable for "opinion mining" a populous due to the fact that all entrants are limited to post in a small, set frame. (better be more informed than the populous you wish to control)<p>Thus the issues about mining are much more manageable.