I like these two posts about the Semantic Web by Daniel Lemire, so I'll quote both.

One:

I honestly do not see the Semantic Web being about to take off. As Bob DuCharme pointed out, people are doing “ontologies for the sake of ontologies”. This will get old very quickly. If 8 years and millions of dollars was not enough to produce a single remotely useful application, what will it take?

Two:

I am sorry, but if the expert system debacle taught us anything, it is that, in Computer Science, it is not enough for an idea to sound intuitively useful. Ideas must be put to the test and they must provide value to users. Otherwise, users do not want to be bothered with it. In this sense, Information Technology is an experimental science. Some ideas are useful, others are not. So far, RDF and ontologies have not been shown to be useful. The burden of the proof is not on the users or on those who do not believe. The burden of the proof lies squarely on those promoting the idea. I do not have to argue against ontologies or RDF: if you disagree with me, you have to prove me wrong. That is how Information Technology works: you convince people by changing their life for the best. The reason for this is simple: there are too many good looking ideas out there for us to consider them all, and so we prune them out by whether or not they are proving useful in practice.

His point about Google getting better (in post One) is an important concept to the Datasphere: we have great tools usable by humans to quickly get to relevant data -- the trick is to bridge the HTML to the data; hence, microformats.