I tweeted about how you can view Twitter as a brain, and Yishay kindly responded with some clarifying questions.

Does Twitter learn? I need to be clear about what I mean by ‘Twitter’ here.  The software doesn’t (as far as I am aware!) learn.  But Twitter, as with all user content based systems is far more than the software.  It is more than software + data too - it includes the people.  Without Twitter users, Twitter would not really ‘be’ anything.  The people learn, and hence I would argue the system learns.  But I think the learning is also embedded in the system to some extent - frequency of signals reinforcing connections.

I don’t believe there is any necessity for a neural network to control organs. However, given the way various politicians are using Twitter, if it isn’t just a read-only practice, then Twitter is probably influencing the functions of State to some degree.  Whilst the ‘organs’ in this sense are fairly autonomous, the brain can still exercise some degree of control/influence over them.  The Mumbai experience also suggests the Twitbrain had some degree of influence over the police movements during the terrorist attacks there.

Yes, it does receive sensory input.  Only textual and temporal, as far as I know, but there are even fairly specialised nodes which take in and pass on inputs from the environment (typically the web, but also day to day life) and others which appear to generate almost random, disconnected from the real world, ‘imagination’ inputs.

Furthermore, the feedback within Twitter means that if an idea gets posited, it is fairly rapidly assessed for goodness of fit with the environmental experience of many sub-nodes (people).  This fits with my model of mind, where essentially experience builds up internal models which are used to check potential actions against in order to predict outcomes.