Reposted from old site – original date: Saturday 11 September 2010
I am seeing a lot of tweets and social site updates that I saw a few days or even weeks back going around my primary network of people, coming through now on my secondary network(s). This got me thinking.
I should probably start with explaining what I consider “primary” and “secondary” networks. My primary network I define as the other programming geeks and Free Software people that I interact with on almost a daily basis on the various networking sites, while my secondary network (OK I use social networks mainly for business/programming purposes (think big mailing list)) is the more social bit – i.e. the people that don’t post about programming etc all the time.
What I have observed in many situations with, often the more humerous posts floating around my primary network, is that they often end up in the non-geek world a few days later. Now the problem here is that all those folks see that as “new” and “realtime” whereas the actual post is already pretty dated. The only way to ascertain whether it is truly “realtime” is to track down the primary source of the post and check the date and time, which takes time and negates the realtime-ness of it all!
Because of all this, we can start thinking of realtime as contextual realtime. That is the only true realtime that any one person can get. I do realise that some applications (such as financial data) is a lot more realtime than most things, so please don’t flame me there, this post is more about how people percieve social networking applications as realtime!
Contextual realtime is the context in which you percieve information flow and the time that it takes to get to you. For any one person, the stuff that you see now, is realtime, the stuff that other people see in their realtime universe is different, which means that any and all social networks are just big contextual datastores that can be queried for the state of a person’s (or other noun) state at any one time in actual time. This is a spatial, contextual, social and temporal thing all at once. The idea of each of us (or our online personas) co-existing in parallel universes is realised.
Another thing that we can start thinking about is how do predictive algorithms applied to these contextual realtime circumstances work? Each noun has a different predictive context that may be historical context for another noun or person. This creates massive headaches for marketing people especially who are looking at realtime local or hyperlocal advertising systems.
As usual, with enough computing power, this could probably be achieved, but that is probably not a reality, so we will have to code for “buest guess” situations.
If you have some thoughts about this, please do leave a comment or two!