The Web suffers a fundamental problem. Search is a symptom of it. Surfing is a symptom of it. Even the website itself is a symptom of it. The problem is that content is organized for you, in advance. Pre-packaged content is like ordering off the menu at a restaurant. Sometimes it’s convenient, sometimes it’s just what you want, but many times, it’s a difficult choice to make. The Web wants to become made-to-order.

Search certainly helps. If I want to order off the menu, it’s great to have access to lots of restaurants and lots of menus. User-generated content is great, too, if you like to cook. But I don’t want to access content or create content, I want to consume it to get stuff done.

No one retrieves content for the sake of retrieving content; they have a deeper purpose in mind. “I need to create a report for my boss.” “I need to plan a trip for my family.” “I want to be entertained.” We’re task-oriented. All the intervening steps amount to the bill, tax, and gratuity. And since most tasks require us to visit many different sites, the overall cost is extraordinary.

A made-to-order Web would spare us these costs. If you’re ordering a specific task, much of the legwork can be delegated to machines. Computers are becoming increasingly adept at analytical tasks. They can break down content into bite-size pieces for our consumption. They are also capable of synthetic tasks, building the content back up into new forms. These types of analysis and synthesis tasks enable made-to-order.

People grudgingly accept pre-packaged content as a necessary evil. It’s just commonsense: if you want to share information, you need to organize it. If you want someone to understand your message, you need to compose it. If you want someone to shop at your store, you need to stock the shelves. But technology often challenges our common sense. We’re witnessing that now.

The emerging technical infrastructure of the Web wants to support made-to-order. Semantic technologies provide exciting possibilities for moving content beyond search into the realm of consumption, to getting stuff done. Semantics provide a layer of meaning that is quite literally the data of content consumption. But to fulfill that promise, we need to imagine new ways for people to experience the Internet.

There is a huge difference between what the Web is now and how we might experience it. The Web is a great metaphor for a massively distributed computing platform, linked documents and data. But it’s a terrible metaphor for consumers’ experience of it. Webs catch food for spiders to eat. And anyone who’s wrestled with the pre-packaged Web, searching, surfing, meandering across dozens of websites to get simple tasks done, knows what it’s like to be caught in the Web.

We see it so often in our history, the application of new technologies to old ways of doing things. It’s like taking a steam engine and hooking it up to a water wheel. In hindsight, it seems ridiculous that we just didn’t simplify things. We need to de-couple the steam engine of semantic technologies from the water wheel of content.

My colleagues and I just attended the 2008 Semantic Technology Conference. One of the recurring messages was that semantic technologies are mostly supportive, connective tissue in the service of other activities. This has a very “steam engine attached to the water wheel” feel to it. Semantic data is consumable, the stuff of meaning. There’s no question that it can provide powerful support to existing activities, but it also opens important new vistas.

Over the next several months, we’ll be taking the covers off Primal Fusion and sharing our initial explorations in this area. We want to simplify things through a more meaningful, made-to-order Web. If you’re interested in following our activities, subscribe to this blog or check out our friends and teammates. Most importantly, please consume it: post your comments and send us your thoughts!