The Big Switch

It's very tempting to think that change will happen quickly. That's probably because we seldom notice all the things that shift in small and sometimes hidden ways to conspire for the change that will take place.

As well, predictions are always long while time seems short. Yet, change happens and when it does in substantial ways, our lives are swept along with it as entire industries seem to disappear overnight.

My great grandmother was born in 1889. When she was young, she saw the first steam engine cars and thought that they would not take hold. She died in 1989 after witnessing two World Wars and the rest of transportation technology that filled the world she knew.

"There’s a paradox in all disruptive new technologies. The powerful lure of new ideas can make them seem inevitable. Yet it often takes very long for these ideas to have an effect. Large sums have been sunk in the existing infrastructure, behaviours change slowly and smarter new ideas may come along. That makes it hard for the layperson to distinguish between the true visionaries of the Information Age and the hype merchants who are simply riding the latest tech bandwagon."

This is the introduction on "cloud computing" in a recent article by the Financial Times. Utility computing, the main concept behind the book by Nicholas Carr, The Big Switch, is a flavor of the same idea. Carr uses the metaphor of electrification to explain the evolution of computing. Computing is turning into a utility. Cheap, utility-supplied computing will ultimately change society as profoundly as cheap electricity did, he says.

Carr uses historical analysis to build the ideas that the Internet is following the same developmental path as electric power did 100 years ago. In the second section of the book, he discusses the economic, social and other issues associated with the Internet becoming the platform and marketplace for commerce.

The book starts with the historical position of water power, the precursor to electricity, and then explains conceptually what these different technologies mean. In Burden's Wheel Carr points out the unique economic impact of General Purpose Technologies -- the few technologies that are the basis for a multitude of other economic activity.

That is followed by The Inventor and His Clerk, which is a historical account of the early days of electricity. The chapter focuses largely on the development and adoption of electric power. It points out that electric power had some false starts such as Edison's instance on local DC plants -- it needed the development of some additional technologies to take off. This chapter is well researched and was fascinating to read. I would refer back to the argument laid out by the Financial Times writer here, sometimes we can prove something just because we want to.

Digital Millwork discusses the recent history of the computer. Carr sees bandwidth as the savior of computing much in the same way that the dynamo and Tesla's AC power turned electric plants into regional power companies. Here Carr connects the history of the electricity at the turn of the 20th century with the development of computing at the turn of the 21st century.

A future of virtual computing where physical location and device based software licensing no longer exist is the central idea of Goodbye, Mr. Gates. In this chapter Carr introduces Google with the positive voice I noticed in an article he authored for Strategy+Business titled The Google Enigma.

In The White City Carr turns away from a continued development of the technical ideas of virtualization and grid computing and moves back into a historical discussion of how electricity changed people's lives and societies. He provides information to set the reader up to make a comparison to what the switch to the Internet might be. His discussions of Insull and Ford are interesting, if brief.

The central chapter of the book, titled World Wide Computer returns to the notion of what the unbridled possibilities of the programmable Internet might be. We take a peak, albeit brief, on the future of corporate computing. The IT we know today will go away and end users will be in charge. I do wonder if the individual with (potentially) infinite information and computing power available to them will enjoy that power.

One question to Carr here: why coin a new term, world wide computer?

From the Many to the Few is a discussion of the social impacts of a programmable Internet where each runs their own personal business and is a brand. The example he chooses is the video Chad Hurley and Steve Chen shot to thank the YouTube community for essentially making them rich with the Google buyout.

I could not help but think about the first bubble. Washington Post Steve Pearlstein's quote sums it up nicely,

"More powerful computers and software and the Internet have reduced demand for travel agents, retail salesmen and inventory control specialists, while making it possible for companies to outsource to India and Poland work like computer programming, tax preparation and customer service."

In this chapter, Carr also highlights that the availability of free resources, in addition to automation are eroding the economic power of individuals.

The Great Unbundling talks about the move from mass markets to markets of one. To do that, Carr uses the example of the forced unbundling of the news business online. Similar to the post I mentioned recently by Dave Morgan at AOL. For example, to gain traction in a crowded marketplace for ideas, The Times of London admits that it has already begun training its reporters to craft their stories in ways that lead to higher placements in search engines. While the financial return in print is part of the magazine or paper, online each individual story needs to earn traffic on its own. The compromise that many journalists may have to make subtracts to all of us.

The chapter also talks about the social implications of a web that connects like people creating a tribal and increasingly multi-polar world, rather than the world wide consciousness assumed to arise when education and communications levels increase. This is the same argument Stephen Baker at BusinessWeek used to counter one of the ideas I exposed in the post Artificial Intelligence agents as conversation agents to which I countered with AI Agents as Discovery Channels.

In Fighting the Net Carr discusses the weaknesses and vulnerabilities of free flowing information and the structural integrity of the net. He does not include public policy recommendations to address the points he makes. These policy recommendations are important especially in the light of an increased need for security and the desire for privacy.

Which he addressed in A Spider's Web. Here he shares the realization that as Richard Hunter says "we live in a world without secrets". This chapter is a warning about the issues of privacy and what it means to do business where everything is recorded and tracked. I've been on the soap box on privacy issues in many posts from Trading Trust for Cash to Facebook Beacon: Brands Guilty by Association?

Finally, iGod talks about the fusion of human and machine consciousness. What is possible when the human brain can immediately access infinite information and the machine gains artificial intelligence? These are the questions raised but not addressed. It indeed reminded me of Hal in 2001: A Space Odyssey as Brin (Google) is quoted mentioning the film on NewsHour.

Overall the book reads very well, even though I am left wanting more in depth analysis of the implications. Carr could have also researched and expanded upon the role that gender equality brought to the thinking in the workplace -- particularly since women tend to think more connectively. I bring it up because the only mention of women in the book is relegated to the history of housekeeping and cleaning. 

I recommend the book to those of you who are keen on ramping up quickly to where we are today on technology, namely the announcements by Google, Microsoft and IBM among others, that they are entering The Wisdom of Clouds as Stephen Baker puts it in his in depth article at BusinessWeek.

Comments

[Mario Vellandi]

I like your big book review. It takes a lot of work to do it right though; the nature of the subject and how it's divided will affect your compilation and what you personally add to it.

Speculation and hypothesizing by themselves, especially about technology, can be a big drag unless well written. The use of historical perspective is a powerful device for the writer to bridge and posit. When I read Drucker's "Innovation and Entrepreneurship", I was delighted to learn about the early days of things like Nylon and commercial banking among others.

Although cloud-computing is very promising, I cannot easily draw mental applications beyond the R&D lab. Examples bring the world of theory and platforms to life; I hope the book provided enough of them to help escape the subjectivity that unfortunately seeps into many books nowadays. I like it best when writers combine History and Present with facts a la reportage, while gently postulating on the future with humility.

[Valeria Maltoni]

I tried to keep it as descriptive as I could adding only additional resources when I felt the book drew the parallels well.

It's well written, I like Carr's style. The reason why I started the post with my take on change and the FT article is the very one you comment on -- it is hard to draw mental applications at this stage of technology maturity. With legacy systems being so widely adopted (in this case they'd be computers;-).

[Carolyn Ann]

I'm not so sure he's right. The parallels break down, rather too rapidly.

Computing has developed in a direction that was quite unpredictable - common run-times, user interface being predominant, storage infrastructures and the like are not showing a trend to the utility business model.

Sure, services like ".Mac" are trying - desperately - to position themselves as utilities (without actually saying they are...), but the general trend is for independent computing power, playing (nicely) with utility-like systems (Facebook, MySpace, Flickr, et al).

Programming models are in a bit of a flux, right now. There's the Microsoft "server-centric" view, and the rest of the world struggling to adopt a not-quite-so-centric view; not "distributed", just not as server-oriented as Microsoft would like. (Has MS had its day? Don't hold your breath, but I'd go as far as suggesting: unless they change, maybe. Nothing like being certain and confident... :-) Okay: yes. And no. Equivocate? Moi? Sure.)

Looking at the trends, I think there's two forces acting on the marketplace: the corporate realization that storage is a strategic issue, and the consumer, who is beginning to realize that... Storage is important. The consumer is also driving another trend: communication.

Forward-thinking corporations like Fox's Speed network, are integrating real-time information and delivering real-time, individualized feeds to a purchasing public. Their coverage of last year's NASCAR Nextel Cup Championship was exemplary; it was also light years ahead of the Formula-1 coverage. (Although F1 still draw in about 300 million per race, worldwide, according to the stats I read not that long ago.)

But for the consumer, like myself, I want to store the race, and watch it when I have time. I don't want to be held to getting up before I've gone to bed! (Shades of a Monty Python sketch await those who dare go there...) That, requires storage and it requires at least a modicum of computing power!

The other issue I have is that utility-like computing power is simply not much in demand! Billing is an issue - easily solved, to be sure, but not so easily justified in the face of grid computing. When it costs just a few thousand dollars to build a supercomputer - what's the point of hiring that service out? (If it cost a few hundreds of thousands of dollars - maybe. This week, at least.)

If you need that sort of computing power now - you'll need it again (and again, and again, and so on), and it becomes cheaper and better to simply hire the right people to build it for you. Besides, it's a fantastic tax-write off!

There might be a market in utility-based routing, but only in the game market. Applications like Second Life and Microsoft's Xbox network require massive investment in a global infrastructure. If someone came along and said "I can route your stuff cheaper, and with more reliability" companies like Microsoft might pay attention. More likely, they'll simply buy you out of the market. But you have to deliver something they might not think they can: an enhancement to their reputation. Fail once - and you're out of business.

The analogy of electricity doesn't work; the technical problems Edison faced with his DC systems precluded its acceptance. Alternating current is simply easier to send down a wire! I remember a period where lots of economists seem to be discussing "natural monopolies"; these do exist (just ask the people in California!), but I don't think computing is one. The cost of computing is declining so rapidly that any utility is likely to outstripped by its customers!

Really good review, though! :-)

Carolyn Ann

[Valeria Maltoni]

The main problem with the utility model to me is users' data. In a fit of ignorance, and thinking it would be convenient, I bought a subscription to .Mac. Only to find out that I can do better without -- and that I really, really do not trust my private data to be online (esp. my network contacts info).

There is no evidence, as you point out, that companies are indeed leveraging the scale and resources they have so far and distributing content (and value) ad hoc. I used to be a huge F1 fan when I lived in Europe, closer to all of that. The new Ferrari is a beauty!

Perhaps the best argument is that sooner rather than later we will run out of space -- data storage. Along those same lines, I see the lack of complete fail over capabilities and non interruption of service in addition to back up a compelling argument. It may be more efficient to outsource that portion (disclaimer: I do live in the IT infrastructure and services space now).

The other big issue is security. Yes, these are all problems companies are facing, esp. those with high levels of consumer traffic and interaction. Social media did increase demand, expectations, and impatience bubbling to the surface those same demands of regular service providers -- like banks, Amazon, and the supply chain of our favorite online retail store. Reputation and trust play a large role here.

The comparison with electricity stops short when we consider content, what is transmitted through data pipes. After all, we would not want the "Hals" of computing to take over for us, would we?

Thank you for the opportunity to drill down a little. You would enjoy the book.