Dave Winer dumps on the Semantic Web. He's right that Google has raised the bar for search, but that's not really the point. I spent some time at a DAML conference last week: Mike Dean's prediction is that only 10% of Semantic Web apps are likely to be about smart agents and search, the over-hyped aspect. The real value of DAML (+OIL, now OWL) will be in exposing information content and connecting resources by labeling them at a higher level. Google is a good way to find the info dots, but there's no help in connecting them. Sometimes you get lucky and it's all in one document, but it doesn't scale as the information and relationship requirements go up. Connectedness is what will drive semantic web growth: you'll want to author with richer markup to allow connections to other people's data at a much high level.

Some of the over-hype comes from the language used to describe OWL. Terms like "semantic web", "understanding", "equivalentProperty", etc. lead new-comers to think there's real understanding of content. In reality, saying SLR is a subclass of Camera doesn't mean you know in any real sense what a camera is, only that you've come to some shared agreement about how to represent knowledge. But a scalable approach to common knowledge representation is a fine goal: you don't get that from ordinary language, and that's all Google or anything else will be able to do for a long time.

This is a core area of interest to this weblog: using semantic markup to make the Bible accessible in new ways.