Challenge Propagation: Towards a theory of distributed intelligence and the global brain

Francis Heylighen, Spanda Journal, Jan 05, 2015
Commentary by Stephen Downes

This is a really good paper that outlines many of the major elements underlying connectivism. It begins by recognizing that "Contemporary science sees societies, organisms and brains as complex adaptive systems. This means that they consist of a vast number of relatively autonomous agents (such as cells, neurons or individuals) that interact locally via a variety of channels." These systems (or as I prefer, networks) are self-organizing and exhibit distributed intelligence. The author moves through various approaches to intelligence (p.6) en route to the distributed "information processing" model, which is explicitly network-based. "The distributed character of neural networks means that its information and “knowledge” are not localized in a single component: they are spread out across all the nodes and links, which together contribute to the final solution." Such a system, he argues, is not focused merely on problems - rather, much of the intelligence-generating activity is productive. There's a lot more here; it's definitely worth the read. Via Jon Dron.


Views: 0 today, 168 total (since January 1, 2017).[Direct Link]