A smarter web is worth the investment
To be honest, I have to disagree with the author's view that the semantic web just provides standards that "no one will ever use"... On the contrary, Google, Microsoft, Yahoo, the BBC and other big online movers and shakers have all developed software that uses them - it's not just limited to government IT departments in the UK or US!
It's interesting to me that the original web proposal was called "Information Management: A Proposal" - the Semantic Web seems to be all about improving information management on the web by introducing standards that can make it more machine-readable. This seems really valuable to me - it's nice having HTML for webmail clients and the Reg, but there's more value in having more markup that can describe facts about things on the web. So perhaps it's going to be an interesting new ground for AI and data management.
Admittedly, there are massive potential pitfalls. Different groups of people will want to use different terms to represent similar data. People will find new and inventive ways of injecting spam and noise into our daily lives. And Big Brother might have more of an interest in what we do on the web if our web-based data gives a better picture of who we are and what we are doing.
But in spite of that, I'd say it sounds really exciting because the web allows us to link data with potentially far more flexibility than your traditional relational database and web-based software can take off in new directions if it becomes easier to access web-based data.
The web has transformed people's lives massively over the past twenty years, so the web deserves to be studied as a subject in its own right. Why not invest in making the web more capable, when (in spite of the time wasted distractedly surfing) it's already proved to be so useful to so many people?
I'm no apologist for the government, but it seems that they've got the right idea in this case.