Semantic Web - What is the Core Problem?

In his latest blog post, Mathew Ingram writes about Paul Miller's interview with Sir Tim Berners-Lee, inventor of the World Wide Web. Miller's interview writeup is very interesting - as Marshall Kirkpatrick notes on the ReadWriteWeb, Sir Tim feels that all the pieces for the Semantic Web are already in place to realize a large part of the dream and to allow us to create applications that leverage the power of structured data and the integration of that data.

[One big problem for the Semantic Web that I've written about recently is the lack of meaning-enabled authoring tools; however, in the interview, Sir Tim indicates that this need is less critical; the structured data we need can come from databases.]

Coming back to Ingram's post, he says that the biggest problem with the Semantic Web is that "it’s as boring as dry toast" - i.e. it's all about the technical side, with discussions about plumbing and widgets and standards, and there's nothing there that will make people sit up and take notice.

I have to agree, although I would express it a little bit differently. The real problem is that true user applications that leverage the power of the Semantic Web are not yet available (let alone a killer app) - although twine, Powerset and a few others are close or on the way. It's the applications that will make the users sit up and take notice; simply having this awesome power there and available is meaningless to users unless and until it can be harnessed to increase their productivity.

As the old Marketing saw goes, customers don't buy drill bits, they buy holes. They don't buy electricity, they buy light. And heat (toasters) or cold (refrigerators). In the same way, easy accessibility and integration of structured data over the web is not going to excite end users. The minute an ordinary user can ask a software agent to find the "cheapest, most convenient flight from San Jose to Seattle", she will take notice; the minute a user can go to a sports web site and ask the software to compare two teams across a variety of statistics, he will be hooked!

Having the pieces in place is great (and a tribute to Sir Tim's foresight and genius). Having this awesome power available is terrific. Now what we need are some powerful applications that put it to good use!

---

Comments

[Paul Miller]

It is indeed true that consumers buy holes rather than drill bits.

Is the Semantic Web (other than in a few unusual examples such as Twine) really aimed at consumers, though? Or is it aimed at the companies that will build the applications that consumers will buy?

Are we selling holes to consumers, or are we the steelworks selling drill bits to Black & Decker?

[Mathew Ingram]

That's a good point, Paul. In a way, I think what makes it hard for some people to wrap their heads around the Semantic Web is that it's about building the tools that underlie or enable other tools -- meta-tools.