One of the ideas behind these things is allow machine processing to have a decent shot at truly understanding what statements are being made in the data.

I also like his post of 7/30/03:

Will somebody please keep these people away from the phone and any funding

What's troubling, at times, is the extremely narrow focus and hype angle some journalists have taken with regard to Semantic Web issues. It was only because I was willing to really, really dig into the issues that the journalist I spoke with got a grasp of the larger issues. I'm not here to criticize them as they're largely trying to do a good job. But I find when speaking with members of the press it's important for me to ask, point-blank and bluntly, what notions are them coming to this conversation with and what editorial expectations are they being constrained within?

For example, how much success will an article about versions of RSS have when the publication to which it's being submitted already uses one form or another? Especially given the disconnect between the journalist, the editor and the IT staff. So it's almost pointless to bother contributing to the article because it'd raise issues that might being the publication's choice of formats into question. How tragic is that?