Cloudonomics

Curator's note: Andy is the CEO of OSI (Open Software Integrators). He's also an exceptional blogger. 

Elsewhere on the Internet, I recently described my new years resolution for 2013, to pursue an “All-in Cloud Architecture”.  Some of this is driven by very tactical concerns (i.e. I don’t want to pay for fiber yet but want reliable offsite backup and high availability) and some of this is driven by my being absolutely convinced that this is a major wave of automation and economic consolidation for our industry.  Our industry has long been the means by which other industries pursued “efficiency” and “productivity” through automation.  The cloud is where our industry turns on itself to become more “efficient” and “productive” through automation.

What’s with the quotes?  The quotes are because, as developers, we typically have strictly positive connotations to these terms.  However, if you ask your average labor leader what “productivity” and “efficiency” mean, they’ll tell you “firing people and doing more work with the people that remain”.  The net effect is that the pay goes down.  Why on earth would you want that?

There are other places where we hear the term efficiency outside of labor.  One of them, “energy efficiency,” is where we do more work with less electricity, or at least don’t let it all leak out under the doors.  Often times you hear “energy efficiency” as a means to protect the environment and reduce the level of climate change.  Ironically, “energy efficiency” often has a rebound effect and actually increases the demand for energy.  This is no surprise to economists and is a pretty basic offshoot of supply and demand.

In other words, while the cloud will probably in the short term cause a net reduction in the number of system administration jobs at any one company due to automation, it will likely increase the demand for technology as a whole and cause our industry to grow further.  I truly think that this is the biggest thing since software was separated out from hardware.

The cloud is not just about technology; it is about economic consolidation.  Economic consolidation is a basic movement of society since the discovery of fire and the wheel.  Economic consolidation is a net effect of the division of labor.  Division of labor is the first sign of a civilization.  Most of you don’t grow your own food as more than a hobby, you didn’t build your own house and you certainly didn’t manufacture your own laptop or tablet.  Why?  Because you’re not cavemen.

Why don’t most companies make their own computer hardware and their own paper, etc?  That doesn’t make sense, if it isn’t part of their “core business” aka “what they do for money”.  So why do most companies have their own people dedicated to building and maintaining server stacks and writing some of the same custom software used at nearly every other company?  

The reason that companies will ultimately go to the cloud is simple, “we are not cavemen”.  
Consolidating the purchase of server hardware is a very basic effect of this.  Dell’s server hardware business will be hurt by cloud computing.  Dell sells a lot of servers that are underused and negotiates a lot of purchases at retail or near-retail pricing.  If everyone went all-in cloud, then all hardware purchases would be wholesale.  The industry and business technology consumers will benefit from this consolidation.

Consolidating the setup, maintenance and configuration of operating systems, application servers and general application infrastructure via Platform as a Service clouds will hurt application server vendors in the short term (fewer license and support contracts for underused software).  Without the rebound effect this would make a whole generation of baristas out of the least skilled system administrators.  However, in the long term, more powerful and sophisticated software will be needed.  Someone has to configure and diagnose its issues.

Consolidation of certain types of application development (i.e. no one should be writing their own internal CMS ever again, the problem is solved) will possibly stop the uncontrolled skyrocketing of developer salaries in the short term.  However, in the long term, we need that software!  The demand will rise.

This is not to say there are not negative effects.  Shipping food across the country has increased the incidents of food poisoning and often reduced the quality (I eat local food where possible).  However, it is much cheaper and we can support our ever growing population.  There will be tendencies in cloud computing to create larger more monopolistic companies.  Fortunately, there will be enough types of software with different problems to solve in order to mitigate their ability to completely corner the market on computing.  We have already seen the beginning of a remarkably competitive environment with VMWare, Microsoft, Google and Amazon all having their own cloud infrastructure.

We will go to the cloud not because it is new, not because of the annoying marketing, but because it is the fundamental progress of civilization.  We will go to the cloud because we’re not cavemen.