Well, lets go to the basics of the Semantic Web ideology as conceived by Tim Berners-Lee, the inventor of the World Wide Web (The Internet as we have it today), now director, World Wide Web Consortium (W3C). First, the term 'Semantic Web' tends to describe a model and technologies as proposed by W3C, these technologies which includes RDF (Resource Description Framework) are intended to provide a formal description of concepts in relationship to a given knowledge domain.
The rationale for such system, however, is on the fact that data as presented on the internet presently is hidden away in HTML files, and its often not useful in some contexts. And again, majority of data as published today are difficult to use on a large scale, owing to unavailability of a global system of publishing data that can be easily accessed by anyone.
Tim Berners-Lee defined it 'as a web of data that can be processed directly and indirectly by machines'. The application in context determines the meaning of the given data and then creates connection for the user, whereby the automated agents would enable a more intelligent access to the information.
Although many of the technologies listed by W3C for the Semantic Web are already in use on various web projects, the global application of the vision remains elusive. And there seems to be little consensus about the likely direction and characteristics of the earlier proposal.
Now, whether a global project of such a magnitude of effect as envisaged by critics as capable of dismantling the present scheme of things on the internet will be allowed a chance remains to be answered.