"I think what must happen is that the existing enterprise systems will have to change to accommodate these new modes of discourse. As for the second point, it seems blindingly obvious to me that the existing systems will continue to have a valuable role to play for what Sig and James have taken to calling the "easily repeatable processes". Those exist, Jevon, and they generate value, and many of them are best served by letting them be. What I find interesting about what's on the horizon is the possibility of adding value to all of the "barely repeatable processes". But I suspect you're mistaken if you think that social software is going to provide some significantly better way to do the ERP's. No tool is good for all things -- hammers are only good at nails, and screwdrivers are best for screws. There is no silver bullet, and social software isn't one either."
As I said earlier, "Sexy" is such a poor word to use in this discussion. Not because I'm a prude, but I keep thinking that this keeps us focused on what's on the surface. I mean underneath Twitter, for example, is an easily repeatable process ("ERP") that supports barely repeatable processes ("BRP"). And what do people hate about Twitter - when the ERP bit goes down. So when we get down to it, both the enterprise and Web 2.0 computing environments run on ERP.
The question is, can the Web 2.0 computing cloud provide a better model for ERP than the traditional enterprise computing environment? And can enterprise computing change to make use of that cloud? In the meantime expect to see lots of Web 2.0 inspired enterprise BRP front ends appear that continue to run on unsexy enterprise ERP systems, but don't expect this to be the end of that story - after all, would Twitter be the same if it was running as a traditional enterprise system?