Paul O’Rorke

Nicholas Carr’s 2008 book “The Big Switch:  Rewiring the World, from Edison to Google” follows up after his earlier book “Does IT Matter?” claimed that information technology is becoming a commodity like electricity and IT doesn’t matter anymore:  it no longer provides a competitive advantage.  ”The Big Switch” claims that computing is turning into a utility and the use of computers is undergoing a transformation similar to what happened when the production of electricity was centralized.  Currently most companies have their own data centers just as factories used to have their own electrical power generators but in the future nearly all computing will be supplied by utilities like the data centers that have emerged at Amazon, Google, and other next-generation internet service providers.

The term “Utility Computing” seems to go back to 1960.  John McCarthy (an Emeritus Professor of Computer Science at Stanford best known for co-founding the field of “Artificial Intelligence”) is quoted:

“… computing may someday be organized as a public utility, just as the telephone system is organized as a public utility.”  MIT Centennial, 1961

Jack Dennis, an Emeritus Prof. of Computer Science and Engineering at MIT attributes the notion of Utility Computing to MIT Prof. Robert Fano in 1960 in a note on his career:  “Towards the Computer Utility.”  However, an interview with Fano never mentions “utility computing” and credits a related concept, time-shared computing, to a January 1, 1959 memo by McCarthy before he left MIT for Stanford.

The book goes back in history from muscle power thru water and steam power but focuses on electrical power.  The historical content includes familiar stories for example about Edison and Tesla and the change from DC to AC for power transmission.  It also includes less well known information about people behind the development of large electrical utilities, for example Samuel Insull and Commonwealth Chicago Edison.  It is interesting to see how Insull went about growing his customer base and his utility.  I wonder whether Insull’s strategies are being used by current cloud computing providers possibly even without changing the terms, like “load balancing,” although the meanings may change by analogy.

Carr lists things that utility computing will have to do to become dominant.  He identifies companies and people driving change including Amazon’s Jeff Bezos, Eric Schmidt at Google, and Marc Benioff at Salesforce.  He mentions Bill Gates’ 2005 memo sounding the alarm at Microsoft about the effects of the internet and utility computing on Microsoft’s business model.  Interestingly, Ray Ozzie, an Illinois grad who took over as Microsoft’s Chief Software Architect from Bill Gates, wrote the attachment containing a plan for shifting from the PC era to the internet-based utility computing era.  Many companies that fail to make this transition will vanish like the ice industry melted away when refrigeration came along.

One of the best chapters of the book gets into some details of the economic changes that are happening as new technology makes rapid growth of startups and accumulation of wealth easier.  In contrast to what happened when Insull and others created the great electrical utilities promoting the growth of the “middle class,” we currently seem to be seeing accelerated concentration of new wealth in the hands of a few.  Carr points out that the trends in computing are contributing.  Users are contributing huge amounts of labor to many online services often without even realizing the value of what they’re providing while the owners reap the profits.  Unable to compete, older industries are being hollowed out and people are losing their jobs.  In addition to reducing the number of jobs, information technology facilitates the outsourcing of jobs to cheaper workers overseas.

The most interesting chapters talked about how the internet is tending to “unbundle” everything from news to society.  Fewer people read newspapers or bundled collections of news articles and advertisements, for example.  Instead people read isolated selected articles and so on.  Personalization-based recommendation and search algorithms can lead people to things they “like.”  Carr summarizes some studies including an experiment by Thomas Schelling that won a Nobel Prize in Economics that seem to suggest that the internet could become balkanized or segregated instead of becoming a utopian global village.  Even more ominously, he points out that businesses and governments are using data-mining and personalization to control and monitor us.

The final chapter contains some amusing vignettes from interviews with Sergey Brin and Larry Page, Google’s founders, describing their efforts toward and opinions about Artificial Intelligence.  The rest of the chapter contains various random bits about AI and the Semantic Web suggesting that the computers in the clouds and the web will become intelligent soon.

The problems that I have with the book are:

• It reads like an expanded magazine article.  It is a couple of hundred double spaced pages.  The content is fluffy and frequently seems disorganized, random, and superficial.  Much of it is highly oversimplified and much of it will become dated quickly.  It seems odd to mention companies and examples like Linden Labs and Second Life almost in passing.  Those of us living in the here-and-now who know of them already don’t learn much if anything and future readers probably won’t know or remember a large fraction of the things mentioned in this way.
• Some of the author’s arguments and conclusions are non-sequiturs, for example his belief that economics and “the free market” and technological developments necessarily determine our future.  That view fails to take into account the fact that we make the rules of the games we play and laws, regulations, and social conventions can be as important as other forces, biological, economic, or whatever.  Arguing that some outcome or trend is inevitable so “why fight it?” is a form of resignation that is inconsistent with – among other things – the kind of revolutionary can-do spirit that drove the technological changes the book chronicles!  One wonders whether global warming is one of supposedly inevitable outcomes of this kind of argument.

I recommend this book in spite of reservations.  The book’s main claim and topics are important and timely, highlighting a historic shift and providing useful historical analogies and ways of thinking about the shift to “cloud computing.”