Here’s a good description of intelligent infrastructure – they’ve defined it clearly, and that’s been one of my major problems with the whole autonomic computing baloney. The interesting part is that with a focus on these characteristics, any process (programming group, ops group, etc) could really gain perspective on the existing systems. The differing vantage of each point they bring up should allow some triangulation on where systems ought to be…

Hmmm. Something to try out…