Skip to content

Archive for

The Public Learning Layer Emerges

John Battelle recently posted a great blog essay in which he distinguished between the “Dependent Web” and the “Independent Web”—the Dependent Web being the part of web that basically has some degree of understanding of who you are and then tailors its responses to you, while the Independent Web is basically the dumb old legacy web that really is not aware of “you” and just operationally treats you like everyone else. 

John makes the point that even the Independent Web is being dragged, kicking and screaming, into what I call the new IT “era of adaptation” by virtue of the advertising delivered on these sites being personalized basis inferences of preferences and interests, even if the substantive, non-advertising functions of these sites are still mired in the legacy of the era of dumb old non-personalized systems. 

John’s take on this inexorable evolution toward a web that learns from its experiences with us is that, “it’s clear that we’re in the early phases of a major shift in the texture and the experience of the web.” Indeed, we are witnessing the plain old web transforming into various learning layers. And as I indicate in The Learning Layer, just how the web will eventually evolve into the one learning layer to rule them all is the real interesting question, and John goes into a good deal of detail on ways that this could/should occur. 

It clearly won’t play out like the original web in which a relatively simple standard was rapidly and extensively leveraged—the evolution of the learning layer will be a function of many different economic interests and objectives, as well as complex technical approaches, almost surely resulting in more of a learning layer mosaic, presenting users with a great variety of different learning textures. Which is likely ultimately a good thing. Of course, as I stress in The Learning Layer, the implementation of company or institution-specific learning layers is much more straightforward and will deliver numerous benefits all the while the long and fascinating process of the “sausage making” of the public web-based learning layer continues.


The Learning Layer: Adaptive Middleware for the Cloud

Some of us from ManyWorlds recently attended the Rackspace Software as a Service (SaaS) Summit. ManyWorlds is a long-time Rackspace customer—we are fortunate to have chosen Rackspace to power our Epiture® learning layer platform back in Rackspace’s early days, well before they became the leader in computing infrastructure and cloud computing services that they are today.   

We were particularly interested in the SaaS applications part of the conference, and it was amazing to see how the number of highly valuable enterprise applications being served from the cloud has absolutely exploded over the past few years. It is clear that delivery of applications from the cloud is inevitably becoming the standard model, even for the largest enterprises—and Rackspace is orienting its entire business around this reality. 

As a still-recovering big company CIO who spent a lot of time worrying about such things, however, perhaps the most intriguing topic at the conference for me was with regard to integration—how applications served from the cloud can be effectively integrated with existing enterprise applications. Most cloud apps have a relatively narrow scope—they do one thing, and they do it very well. But they need to fit within the context of an existing application environment, which is often the province of broad-scope applications such as enterprise resource planning (ERP) systems (e.g., SAP) and customer relationship management (CRM) systems. As the number of cloud-based applications grows, this integration issue becomes front and center for the CIO. Legacy applications generally cannot simply be ripped out and replaced—so integration approaches, whether ad hoc, or better, through standardized APIs, promises to increasingly dominate the CIO’s agenda. 

The Summit served to reinforce for me that the good news is that the learning layer can play a strong role in ameliorating this integration problem. The “ethereal network of learning” that I discuss in the book can be applied to integrate together myriad cloud applications with complex legacy application environments. Think of it as an “adaptive middleware” layer sitting between the cloud-based applications and the legacy applications, providing a single, coherent environment that includes a capability for integrative workflow, and that automatically adapts and personalizes based on usage.

When we first conceived of the learning layer, we primarily thought of it as providing an integrated layer of learning over existing applications and content—it is now clear that, perhaps even more importantly, it will serve as an adaptive bridge between the new and the old of the application world.