Maybe There Really Will Only Be Five Computers…

Thomas J. Watson, legendary chief of IBM during its early decades and the Bill Gates of his time, has oft been quoted – and derided – for stating, in 1943, that "I think there is a world market for maybe five computers." Whether he actually said this quote is…


Thomas J. Watson, legendary chief of IBM during its early decades and the Bill Gates of his time, has oft been quoted – and derided – for stating, in 1943, that “I think there is a world market for maybe five computers.” Whether he actually said this quote is in dispute, but it’s been used in hundreds of articles and books as proof that even the richest men in the world (which is what Watson was for a spell) can get things utterly wrong.

After all, there are now hundreds of millions of computers, thanks to Bill Gates and Andy Grove.

But staring at how things are shaping up in our marketplace, maybe Watson was right, in a way. The march to cloud computing and the rush of companies building brands and services where both enterprises and consumers can park their compute needs is palpable. And over the next ten or so years, I wonder if perhaps the market won’t shake out in such a way that we have just a handful of “computers” – brands we trust to manage our personal and our work storage, processing, and creation tasks. We may access these brands through any number of interfaces, but the computation, in the manner Watson would have understood it, happens on massively parallel grids which are managed, competitively, by just a few companies.*

It seems that is how Watson, or others like him, saw it back in the 1950s. According to sources quoted from Wikipedia, Professor Douglas Hartee, a Cambridge mathematician, estimated that all the calculations required to run in England would take about three “computers,” each distributed in distinct geographical locations around the country. The reasoning was pretty defensible: computers were maddeningly complex, extraordinarily expensive, and nearly impossible to run.

Now, that’s not true for a Mac, an iPhone, or even a PC. But very few of us would want to own and operate EC2 or S3.

Right now, I’d wager that the handful of brands leading the charge to win in this market might be Google, Amazon, Microsoft, Apple, and….IBM. About five or so. Maybe Watson will be proven right, even if he never was wrong in the first place.

* Among other things, it is this move to the cloud, with its attendant consequences of loss of generativity and control at the edges, which worries Zittrain, Lanier, and others. But more on that later.

6 thoughts on “Maybe There Really Will Only Be Five Computers…”

  1. I see where you’re going with this, but the computational toothpaste will never go back into the tube. Smartphones are computers, and there’s no reason to centralize processing that can be done faster at a local level.

    Cloud computing holds a lot of promise, but I don’t think we’ll ever see a complete shift

  2. I agree with the general trend you outline. It started with the advent of the Internet, accelerated with “blank-as-a-service” companies using the web as a distribution channel, and is expanding outward to encompass more and more computational needs as more people and companies get comfortable with “cloud computing.” The Google Chromebooks are a great example of taking this idea to the extreme, in which the devices in front of the users are only an interface.

    To simplify the debate, it really comes down to whether computing is best done in centralized hubs (which can leverage economies of scale, virtualization of hardware, etc) or at the end nodes (or “terminals” as they were called in Watson’s day). This assumes that traffic between the hub and nodes is smooth, efficient, and cost effective, which has only happened recently for mobile devices.

    There will always be some need for computing at the node, at the very least when a user needs to be offline or when extreme security measures are necessary. Imagine trying to tell your boss that you couldn’t complete the spreadsheet analysis because your 4G signal dropped! I think there will also continue to be some edge content creation cases which are best done at the node, such as audio recording (although the editing process may go to the cloud).

    Computing at the hub really excels for meeting the needs of a) lots of computational horsepower b) unlimited and always-available content storage and c) real-time data feeds/updates. Amazon’s EC2,, and Twitter are examples of cloud products that are tackling those specific needs. But I see the general trend of users wanting products that are convenient, collaborative, and heavily networked to be leveraging the most relevant and real-time information. This suggests a continued push to the cloud, and centralized computing, for the majority of consumer and business needs.

    Two caveats: there may be a backlash against cloud computing if security breaches of cloud companies continue to be an issue, and not everyone will be willing to give up edge control to a handful of extremely powerful companies (as you alluded to).

    John, as a side-note, I just saw your George Harding post. Since the job opening is still up on the Federated Media site, I assume you’re still waiting for your lightning.

    I’m currently in career transition and am extremely intrigued by this opportunity. I graduated from Stanford 4 years ago, worked for McKinsey & Company in NYC, and most recently started (and just closed) my own company, I love reading paperbacks, debating interesting topics, and writing for my new blog (

    Let’s talk. You can reach me at everitt [dot] rory [at] gmail [dot] com.

  3. I derided that quote for a long time and then I changed my mind. The line between each computer and the cloud started to blur and in 2006 I put the following on my blog, because the line was not anymore meaningful to me.

    “I think there is a world market for maybe five computers.”

    Thomas Watson, IBM, 1943. (misquote)

    “Why the four others?”

    JFNO, 2006.

    It’s like insisting on talking about the numbers of CPU in a computer. After feeling all smug about myself, I thought about a short story where I saw the notion of one computer for everyone. “The Last Question” by Isaac Asimov was written in 1956 and the balance between a multitude of computers and one mega-computer is nicely presented. The throughput and latency of the connections are the factor that make the distinction disappear.

  4. Rory, there’s a contact there to use to apply, think it’s still open. JFN, as I said to a colleague in an email about this post: …maybe there’s only one computer (the Internet), but you can get it in five flavors, at least…

  5. John, I just scoured the post again and only came across these instructions:

    “If you’re the right person, contact me. You’re smart enough to figure out how, and what to say when you do. (And tell me whose picture that is up at the top).”

    In the comments section, there’s a link from a Ric Steinberger, but I’m pretty sure it’s a virus rather than an application form 🙂

    Please let me know how you would like to take things from here. Or you can privately email me at everitt [dot] rory [at] gmail [dot] com.

Leave a Reply

Your email address will not be published. Required fields are marked *