Apple and the Cloud: A Cautionary Tale
(Page 3 of 6)
the iPad is just an expensive screen, and the price of screens is coming down fast.
To help us navigate and interact in the coming dataverse, the devices in our lives will become simple, connected, cheap, and ubiquitous. There are data ecosystems emerging everywhere—government, education, business, manufacturing, transportation, science, etc. Imagine going to the Olympics—you wouldn’t use a single app to manage your entire experience for two weeks, would you? Or even a few apps? To navigate the Olympics, you’ll need a web of data and services that keep you up to the second on tickets, meals, transportation, events, athletes, media, people, groups, crowds, and more. To build a fully programmable and adaptable set of tools, we will need to harness the power of cloud computing in a way that doesn’t imitate the old model of apps holding the data.
Most Apps are Websites
This isn’t something Apple talks about, but the truth is that almost all the apps in the app store are really just websites. Here’s how it breaks down:
Some apps are 80 percent content and 20 percent programming. This covers most games, which provide high speed interaction with environments that change. Apps that control things in the real world—the camera on your phone, your bicycle, a baby monitor—tend to fall into this category. While some of these are true apps, most have web-based versions that work just fine in a browser. And we can improve technology to let our devices store immediate data and instructions for high performance (or, for example, if you are away from Internet access) but still reside in the cloud.
Many big desktop/enterprise apps are 50 percent content and 50 percent programming. Think of the big programs we own or use—Photoshop, Lightroom, CAD, Word processors, spreadsheets, ERP systems, etc. They are designed to move data in, process it, and ship it out. These general-purpose apps aren’t found on the lightweight app platforms we have today. In the connected, interoperable, collaborative, programmable, adaptive future they won’t play much of a role in our daily lives.
A few apps are 20 percent content and 80 percent programming. Think of an iPad drawing or note-taking app. The app and its features are more complex than what we make with it. Few apps fall into this category.
A handful of apps are 98 percent programming and 2 percent content. Examples are a chess game, translation, or voice recognition, where a small input requires a large amount of computation. This category also includes interfaces to hardware, like opening a garage door or turning on the lights in our homes. The best of these tend to run on servers, where they have access to many resources and can learn constantly, using the device only for input and output.
As you can see, very few apps really need to be apps. Most of our computing needs can be handled by Web servers, rather than having to run locally (and trap data) on stand-alone machines. We don’t really need to improve the storage capacity or CPU capability of our devices. We just need to use the cloud more effectively.
It’s the Data, Stupid
One possible solution is to replace our smart phones with a netbook or Web-based tablet, where all the apps are replaced by websites using HTML5 and other connective tissue being built today. Many people think this would be much better and more scalable than the stand-alone apps we have today. In this scenario, all the little squares on the phone aren’t apps but simply bookmarks and automated logins.
And yet even that vision comes up short when you start to think about the scale of future data ecosystems. Here, watch these two vision videos from Microsoft—yes, Microsoft!—Labs to get the idea. [Story continues below videos.]
As you walk down the street, you sport a pair of fashionable glasses that interact with the dataverse all around you. Where is the app store in this scenario? It should be obvious that a single app, or something like today’s web browser, isn’t going to be enough to … Next Page »