I was interested in this week's ZDNet piece, Cloud computing's real creative destruction may be the IT workforce. The piece discusses a presentation at last week's Gartner Symposium that posited cloud computing will be a net destructor of IT jobs.
I am not one to gainsay Gartner's wisdom. In fact, earlier this year I wrote about the topic of job losses and cloud computing in a post titled Cloud CIO: Yes, Your Job Is At Risk." The topic seems to be in the air: Another blog I came across this week was titled, I'll Probably Never Hire Another Pure SysAdmin"
As I say, I am not one to gainsay Gartner's wisdom, but I believe the research firm makes a common mistake regarding innovation (aka "creative destruction").
The argument is that cloud computing automates tasks that, in the past, have been performed by employees, and that after automation occurs, those people will no longer be needed. CIOs, who believe that their organizations are inefficient and are looking to cut costs, will substitute automation for headcount and start laying people off left and right.
One might be tempted to observe that this is a drum that has been beat many times in the past. Over 50 years ago, Katherine Hepburn and Spencer Tracy tussled over the same issue in the 1957 movie Desk Set. I believe one can compare total employment of so-called knowledge workers in 1957 and today and recognize that the number of these types of workers is far higher now.
The reason Gartner (and many others) are wrong about cloud computing's effect on employment is that they assume this disruption is unleashed in a static environment.
However, the field of computing has never been static, and will not be in the face of cloud computing. In fact, one may predict that cloud computing will vastly increase the amount of computation that is done and will significantly increase the number of workers in the field.
Why can this be so confidently predicted?
The posh explanation can be found in Jevon's Paradox. From Wikipedia: "Jevon's Paradox is the proposition that technological progress that increases the efficiency with which a resource is used tends to increase (rather than decrease) the rate of consumption of that resource."
More prosaically, this proposition is captured in the phrase "price elasticity of demand," which describes the responsiveness—or elasticity—of the quantity demanded of a good or service to a change in its price.
Boiled down, these two formulations amount to the same thing: As something gets cheaper, we consume more of it. Oftentimes, we begin to substitute this newly cheaper thing for another thing which formerly was less expensive but now, due to the cost change of the newly cheaper item, is more expensive.
Information technology provides many rich examples of this scenario. As the cost of computers has dropped, many manual processes have been replaced by their computerized equivalent. Word processors replaced typing pools. Payroll systems replaced time clerks.
However, I believe these two formulations fail to fully capture the effect cloud computing will have on IT consumption. Besides being less expensive, cloud computing, via its automation, makes it much easier to consume IT resources. When one can obtain virtual machines in minutes by filling out a web form rather than weeks by sending an email request to a human with other priorities, one will use more virtual machines. Taken together, the lower cost and easy access of cloud computing is going to lead to an explosion of IT resource consumption.
The net effect is that any loss of jobs caused by the move to cloud computing will be more than made up by the overall growth of computing that the cloud causes and the resulting growth of jobs associated with cloud environments and applications.
There's no question that lower-skilled IT employees who cannot develop the ability to work in a highly automated, large-scale environment are going to have severely restricted job prospects. But that makes sense, because IT is scaling up. The practices, processes and skills appropriate to smaller-scale, less dynamic IT environments cannot support the future nature of information technology.
We have worked with many IT organizations whose employees and processes are groaning under the strain of trying to apply old practices to this new world. Believing the traditional approach will be sufficient for a world in which this scale and agility issue is three, five or ten times larger is laughable.
To quote the Nixon-era economist (and Ferris Bueller actor Ben Stein's father) Herb Stein: "If something cannot go on forever, it will stop." Simply put, IT has to change and move beyond established practices and lower-level skill sets if it is to fulfill its role in the corporation of the future.
With respect to IT employees, we firmly believe that every capable IT employee, if he or she is willing to learn, will have a bright future. The coming demand for IT services will dwarf the expectations of people, even those who putatively are seers in the field.
Here are some of the jobs for which we see enormous need in the future:
Enterprise architects: These are people who can design cloud computing environments. Key skills include understanding cloud orchestration software stacks, application management frameworks, network designs, bandwidth latency, and a host of other things. The cloud computing environment of resources must operate in unison like an orchestra, and enterprise architects are the conductors.
Application architects. These employees apply the cloud computing environment designed by enterprise architects to specific solutions. Key skills include appropriate partitioning of applications into RESTful interfaces, deciding which application objects need to be distributed into content delivery networks, creating business continuity and disaster recovery strategies via data center- or cloud-spanning designs, along with many other application elements.
Storage and data architects. These workers focus on building highly-scaled, highly-distributed data and storage schemes. Calling this a NoSQL strategy is too limited. This type of architect has to account for vastly distributed data and the need for transactional events.
Software engineers. There are a host of skills that are needed to build cloud applications. From AJAX-oriented interfaces, elastic and failure-resistant components, caching tier design, to API integration and load/performance testing, software engineering skills are crucial to cloud computing applications.
Sysadmin 2.0. Critical skills for the systems administrator of the future are operations automation tools like Chef, Puppet and the like. Implementing monitoring systems appropriate for elastic, highly-scaled, geographically distributed applications is another Sysadmin 2.0 skill in high demand.
Capacity planners. In a world of explosive growth, application elasticity and unforeseen demand (aka resource self-service), capacity planning moves out from a leisurely backwater to a vital organization capability. Understanding and forecasting capacity based on one's own organization as well as industry patterns based on other cloud-oriented organizations will be a skill in short supply.
Financial analysts. The move away from quasi-monopoly, coarse-grained transfer pricing to competitive, fine-grained open market pricing associated with applications that can be migrated from one operating environment to another will require much more sophisticated financial analysis. Just as the world has many manufacturing financial analysts, we will soon have many information technology financial analysts.
In summary, evaluating employment patterns based on reduced need for manual skills with an assumption of a fixed pie of overall demand fails to grasp the coming boom in IT employment. The job titles and skills may be different, but the numbers will be far larger. Anyone who predicts a future of IT ghost towns hasn't really examined the real implications of cloud computing.
Bernard Golden is CEO of consulting firm HyperStratus, which specializes in virtualization, cloud computing and related issues. He is also the author of "Virtualization for Dummies," the best-selling book on virtualization to date.