The ballad of John Henry, a “steel drivin’ man,” is a staple of the folk song movement. The song describes the contest between John Henry and a steam-powered hammer: To save his job and the jobs of his men, John Henry challenges the owner of steam-powered hammer to a contest: Henry will race the machine. John Henry beats it, but exhausted, he collapses and dies. The poignancy of the moment gives the song its emotional power and accounts for its long-lived popularity.
A more recent man-vs-machine contest played out on the gameshow Jeopardy. As vividly portrayed on Nova, the two greatest Jeopardy champions of all time squared off against Watson, an IBM computer. Watson is programmed to rapidly search through a knowledge database to respond to the answer displayed on the Jeopardy game board.
Nova delves into the background of IBM’s enormous preparation for the shootout, including countless project meetings, mock matches and tweaking Watson’s search algorithms. Unlike John Henry, this contest ends with machine winning, the victory being humorously acknowledged by one of the contestants, who added “I for one welcome our new computer overlords” to his final answer. The show is riveting television, as it sharply communicates the anxiety the IBM team went through in its effort to be competitive — its ultimate victory was by no means certain.
Here’s the thing, though: The power of both of these situations reflects a moment in time — a time in which the contest between human and machine is relatively equal. The fact that the machine’s capabilities are relatively equal to the human’s is what gives the situation its emotional energy.
But does anyone doubt what the competition will look like 10 years from now? Can anyone imagine that five or 10 years from now there will be any question about whether a human can defeat Watson or its descendants in Jeopardy? Given the pace of Moore’s Law, it’s obvious that a decade from now humans will be mere bystanders, completely overwhelmed by their computerized competitor.
The dynamic of a new technology arriving, competing against human capabilities, being found wanting (or barely successful), and, in a short period of time, improving far past human capabilities is, perhaps, the story of our time.
Siri and Kinect Lead the Revolution
The dynamic was brought to mind by two new technology developments that have recently come to market: Apple’s Siri and Microsoft’s Kinect. They both represent new ways of interacting with applications and it’s not an exaggeration to say that they reflect a revolution in the way humans will live in the near future.
Siri, of course, is the new voice-activated “assistant” delivered as part of the iPhone 4S. Kinect is a gesture-oriented input device was originally delivered as an add-on to the Xbox 360 game platform.
Initial reviews of Siri made it seem magical. One person queried the service “will I need an umbrella tonight?” The response: “There’s no rain in the forecast.” Not too much later, however, the reviews turned to its shortcomings.
Kinect, by contrast, hasn’t suffered much from negative reviews. What has been more amazing about it is the way people have hacked it into doing things far beyond what its creators envisioned. What was created to offer a gesture-based way to play video games has been repurposed into a device to control plumbing, perform training, and create a robotized paparazzi.
What both of these represent is a new frontier in what computerized applications can be applied to. Things that heretofore humans had to control can now be accomplished by AI-enhanced computer systems. And, notwithstanding Siri’s deficiencies, it’s clear that five or 10 years from now our world will look very different.
The rush of computerization into our society—into the very fabric of our lives—is headlong. It is occurring at a pace barely comprehensible. Imagine it’s going to take 10 years for self-driving cars to arrive? Think again. Is it going to be a decade before your health is monitored on a real-time basis, with live links to your doctor’s office? Not a chance. Do you think it will be10 years before your business gets immediate feedback from the operating characteristics of every one of its products out in the field? No way.
The Cloud and Apps of the Future
The application architecture of the future is an intelligent device communicating with an immensely capable cloud-based application that performs massive processing on behalf of the software on the device, and, by extension, for whatever purpose the device is deployed.
What will this new application type look like? Here are a few examples:
- It may be acting as a personal assistant (get a reservation at Per Se for four at 7:30 in my name).
- It may be devices tracking container location in port transshipment areas, as is implemented by one former business school colleague’s company, ContainerTrac. Its application can set alerts if a container is placed near a perimeter fence when it shouldn’t be (i.e., it will set an alert if it seems likely the container is being moved to a location convenient for burglary).
- It may be interacting with personal information and entertainment systems, as in this example, an interpretation of how Siri is going to affect the TV business.
There’s no doubt that nearly every business is going to be transformed by the digital (the cloud app) meeting the physical (voice activation, motion detection, geographic tracking).
As Mark Andreesen put it in a recent Wall Street Journal piece, “software is eating the world,” citing the dislocations in the music, book, telecom, recruiting, automobile, retail, logistics and financial services. This week, even venerable publication Forbes chimed in with an article titled “Now Every Company is a Software Company.”
What does this have to do with cloud computing? Here are some things for you to consider:
Plan for scale. The amount (and types) of data is going to explode. You may think your storage requirements have gone up a lot with your Hadoop analysis of your website’s clickstream analysis, but you haven’t seen anything yet. The number of applications you manage is also going to explode. When I speak I often predict application portfolio growth of at least an order of magnitude, and I often see disbelieving smiles on audience member’s faces. They’re judging the future by the past. With the reduced friction and lower cost of cloud computing, historic barriers to application development and deployment will fall dramatically, leading to huge jumps in every company’s application portfolio.
Plan for a much more dynamic environment. With unpredictable interaction with input devices and floods of data, application load variability will be at least one—if not two—standard deviations larger. The capability to design highly elastic and efficient application architectures will be crucial to support the operating environment of the near future.
Plan for a different class of devices. Up to now, the “app revolution” has mostly been about a new way for humans to interact with a remote application. So most mobile apps have just been cut-down interfaces put onto a cellphone. Soon the device will have significant memory and processing capability in and of itself and will perform processing locally, while interacting with remote applications.
For an insight into this, see the profoundly futuristic novel Daemon. It offers a good look into a world in which devices react and respond to the environment. Instead of just typing into devices, we’ll speak (Siri) and gesture (Kinect). For many interactions, people won’t do anything – the device will automatically obtain information from the environment and execute an action While the past has been devices controlled by humans, the future will be devices responding to humans – and other devices.
Plan for new application architectures. The Silk browser shipped on the Kindle Fire is an example of the new application architecture—the application is split between device and cloud backend. This architecture lets the device perform its role – interaction with the environment and local processing—unhindered with heavy processing; likewise, the cloud end can perform its role—processor-intensive code execution—unburdened by needing to send all bits back and forth across the network.
I know that Silk is controversial, with many people proclaiming it slower than a regular browser, but I believe Amazon will get this tuned and Silk will perform better. In any case, the design pattern is what is important, and it’s an obvious one. What this means for you is that you will need very sharp application and network architects, comfortable with distributed architectures and performance optimization, not to mention familiarity with application construction with code, new storage mechanisms and services accessed via APIs.
Plan for a different role. IT has, in the past, been mostly about automating internal business processes. And, in the past, the IT world has been full of executives asking for “a seat at the table.” The rise of these new capabilities in applications moves IT from the supporting cast to a leading role. If every company is a software company, the person in charge of software isn’t going to sit at the table, he or she is going to own the table. But, as the old saying goes, “be careful what you wish for.” When the company’s success rides on the project release date, there will be far less tolerance for missed schedules and dissatisfied users.
Noted futurist Ray Kurzweil states that the rate of change in human society is accelerating. His formulation goes something like “every decade of the 20th century contained as much change as a century 100 years previously. In the 21st century, every year is like a decade in the 20th century.” We are becoming a software-driven economy and you can see where it’s headed by looking at Siri and Kinect.
Bernard Golden is CEO of HyperStratus. He is the author Virtualization for Dummies and writes about cloud computing and virtualization for CIO.com.