Integration will be the main challenge," he says. "IT will be at a higher level.

This is just not so. We will never live in this magical world where every conceivable app that a individual or business needs already exists up in a 'cloud' somewhere. Someone WILL STILL NEED TO BUILD NEW SOFTWARE SYSTEMS! This is buzzword nonsense at its finest.

I agree that someone has to still build software applications, however that job moved from local IT departments (build your own) to software companies (buy instead of build) in the 80's and 90's. But local IT departments still had to install, operate and maintain those applications, and integrate them together.

Enter Y2K and outsourcing deals, which moved a lot of the operate and maintain out of the local IT shop. More IT folks now spend their day working with the business, refining IT processes, and setting future directions and strategies.

While outsourcing has its many proponents and detractors, SaaS is a major threat to their business model, much more so than the local IT shop. In particular, the entire communication and collaboration software suite (e.g. email, IM, conferencing, team spaces, etc.)is moving rapidly to SaaS to lower cost, but more importantly to enable inter-company collaboration.

IT professionals need to take a longer-term outlook and determine how they will re-skill. My advise for the last 20 years is that IT folks lose 20% of their value (i.e. relevant knowledge) every year and should constantly be looking for new challenges that meet their personal view of their career. I think that advise is still solid today.

Yes - but that "someone" will not be a business firm's employee. It will be an employee of a software firm who installs their software on a service provider's server. It is no different than any other support commodity needed by every business firm to perform its functions (floor space, office furniture, utilities, phone service, etc.). It is the old "buy vs build" business decision; if what you are considering is strategic to your business operations, you are more likely to build it (and treat it like a strategic investment); if it is a support item, you are more likely to buy it (and treat it as an expense item). Does this mean there will be fewer software vendors? How many firms still make word processing, spreadsheet, email, Web browsing and ERP software? The real question here is "Can software firms step up to the plate and invent completely new apps that greatly benefit (maybe even transform) business firms' operations?"

To call SaaS 'nonsense' is unrealistic. That statement is disqualified by the sheer amount of revenue being generated by the successful SaaS vendors (Salesforce.com etc). SaaS will continue to become an even more appealing option to midsize organisations in areas where access to bandwidth was once limited but is now on the increase. Turning to larger organisations for a moment - the increased focus on the transformation from application-centric IT departments to service-centric IT departments coupled with declining IT budgets world wide make SaaS a natural consideration (if not the obvious choice).

When will IT departments grow up, and stop looking for silver bullets, and start realizing the issue is NOT implementation, but understanding the BUSINESS, as the primary issue of the day.

SAAS and Clouds are just new names for timesharing and Fortran and Cobol callable subroutines.

Business understanding is hard in comparison to writing code, and that is why we do not want to do it - it is not writing "specs" - it is describing the business.

OK, so we're going through another platform shift. Wasn't the last one driven by business users being dissatisfied with IT (coding on the mainframe) and wanting to develop new applications more quickly? So now here we go again ... business users want more applications developed more quickly. Guess what ... in a few years we'll all be wondering why everything is so complex again.

I agree that integration is key - I for one would not like to see my data being lost in the clouds. I need business processes that add value to my business. I need different applications integrated together to deliver value. I don't need individual islands of data separated by a gulf of clouds and end up losing control.

The other thing worth thinking about is what all this fragmentation will do to my data quality. Imagine it - data in Salesforce, Rightnow, Taleo, Netsuite and Workday. All fiercely competitive, all wanting to rule the world. Hmm ... sounds like the early days of client/server!

As for making your skills irrelevant - as with the client/server shift I thnk it's more about evolving your skills. The modern IT professional focuses on helping IT to enable the business. No change here. The programming language may change (not sure about this one) but ultimately we're still trying to help the business succeed.

Take a look at Mike Kavis' post. He sheds more light on this theme.

http://it.toolbox.com/blogs/madgreek/will-it-shops-shrink-in-the-near-future-26200