In the fevered world of cloud computing, much is possible certainly. After all, the concept was first discussed as a grandiose parallel to the rise of the electric grid and the replacement of locally created power by a commoditized utility. Internet worrywart Nick Carr chronicled this in his book "The Big Switch" and went so far as to view it as a coming cause of major employment disruptions, as labor-intensive local computing operations were widely replaced by mega-datacenters with just a handful of operators.
There's certainly a kernel of truth here; more computing happens in other places, standardization and commoditzation continue apace, and so forth. But it's become clear to myself and others that (as shouldn't be surprising), there's huge inertia to things as they are; evolutions are the norm. And, indeed, there are very good reasons not to yield control of sensitive data or service levels in many cases.
For these reasons and others, there is no big switch--at least in any interesting planning horizon. Rather, there are a lot of different approaches aiming to make IT better and imbue it with the flexibility, efficiency, and faster time-to-new-services that cloud computing promises. And many of these are happening within the firewalls of enterprises. I admit to once having been a "private cloud" skeptic myself--at least by that name. But the level of interest in private clouds is so overwhelming that to fight the idea seems almost quixotic at this point.
So we're going to have private clouds. And even if they aren't going to completely usurp all computing anytime soon, we're going to have public clouds like Amazon too.
One of the interesting questions to me is: what else are we going to have?
One option is the appropriately named hybrid cloud. The idea here is that you can own a datacenter capable of handling routine levels of work, but you "cloudburst" out to public clouds to handle spikes in capacity. This raises all sort of issues around portability across many dimensions of computing and applications that are important for moving between clouds even absent an explicit hybrid model. It's an important topic, but one I'll save for another day.
The other model that I've seen pop up in a few discussions recently is perhaps less obvious. It appears in the National Institute of Standards and Technology cloud definition as a "community cloud." To quote N.I.S.T.: "The cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be managed by the organizations or a third party and may exist on premise or off premise."
The notion isn't necessarily intuitive. However, as we've seen with software-as-a-service applications such as Salesforce.com or even outsourcing in general, the concern among customers isn't necessarily about sharing infrastructure per se. It isn't necessarily about generalized privacy or security concerns per se. Rather, it's about being able to meet specific regulatory compliance regimes, satisfying audit requirements, and meeting required service level objectives such as response time.
In other words, it's about meeting a set of requirements that may very well be shared among the companies in an industry or the departments in a governmental entity. There is precedent here. Applications and online services are already widely shared within industry verticals.
A community cloud could well be a logical next step. This approach doesn't require that each company or organization have its own IT infrastructure--although those in the most technically sophisticated businesses like financial services may well do so nonetheless. At the same time, it's shared infrastructure that isn't a generic public cloud designed to appeal to the needs of the broadest set of users.
In many cases, a community cloud could take the form of a service provider that caters to a specific industry, providing some combination of infrastructure and applications to subscriber companies. In the case of governments, it may be a more internally driven organic effort. But in both cases, it's essentially an outgrowth of the common vendors and shared services that we see today in such cases that don't require some radical newfound spirit of cooperation among sometimes-fierce competitors.