In effect we have a system which has developed a set of constraints, but those constraints in turn are modified by the behaviour of participants within the system.  As you get engaged you get to know the different editors, their quirks and capabilities.  Trust means you don't need to check on some edits, transparency allows you to track anything done by anyone at any time.  Encouragingly my experience so far has seen sense prevail even in contentious debates over time.  Overall a demonstration of the ability of a self organising system to avoid tyranny through the participation of the many.

OK I know a lot of this will already be known to many readers of this blog, but I think its worth a mention on a lazy saturday morning in Singapore.  There are some lessons there for governance procedures in organisations and in IT in particular.  I argued strongly at several sessions this week that allowing traditional IT procurement to determine social computing needs would contradict the very nature of social computing and prevent its benefits been realised.  Mandating one system for email, blogs, wikis etc. is I think a fundamental mistake.  You end up with too many compromises when a rich medley of tools that can be picked up, discarded and used in novel and interesting ways is much more likely to succeed.  Not only that it costs less.  I reckon savvy CEOs could take 20% out of their IT budgets if they brought control back to transaction systems and core databases and allowed users greater freedom and that is before you take account of the productivity improvements.  Paradoxically, security would also be enhanced by pulling in the firewalls to core data and transaction systems.  It will come, the question is when and how traumatically.

The other great benefit is the speed with which material can be updated.  Euan uses an example of this with this video showing  Wikipedia dealing with the London Bombs: 7/7