Ian Smith sent me these notes a few month ago, I think he is onto something:

Given that:

• The economics of the computing/storage cloud are overwhelming and it will be adopted broadly. It is/will be like power grids.
• Few organizations will have concerns that will justify the use of “local” servers/services. These will be organizations that have the same relation to computing that hospitals have to the current electrical grid.
• Mobile phone companies will be forced to support mobile computing “to the cloud” and it will be “always on” for almost every citizen of the developed world (same as electricity, except not wired).
• Many service providers will emerge that make it easier for users to exploit the cloud. Backups, applications ala GMail, etc.
• Users will begin to think of their own data the way they think of web data now. It will not have a “place.”
• Computing has been and will continue to be commoditized and homogenized. The importance of platform, OS, and CPU are shrinking and will continue to do so.

Thus:

• We have to be concerned about the way the cloud’s infrastructure evolves and be sure that it is a “public good” not a “private good.”

Example of kind of solution we want: The internet, mostly by accident, evolved as a public good. Efforts to stifle that have failed. It is important to be aware that there is no alternative to this network for most individuals and companies. There is no “competition” for the internet largely because it is unnecessary.

Example of the kind of solution we don’t want: the US phone system, monopoly for private profit that hurt consumers for generations. Similar in Western European telephony as well as airlines.