For years, semantic technology has been expressed primarily as search. The utilization of methods like Latent Semantic Analysis (LSA), Bayesian Inference and Support Vector Machines (SVM) have produced better search relevance, recall and precision for users, but have been limited to transactional search - which is dying a slow death.
Transactional search will increasingly be replaced by social filtering - fundamentally you won't 'search' - you will 'follow' content and people that match your interests and are filtered and suggested by your interest graph. At the core of this ability are semantic technologies that can dynamically derive meaning from unstructured text to provide context, the key ingredient, for the development of the Interest Graph.
The real potential and power of semantic technologies will be unleashed as semantic vendors embrace and integrate the richness of the data being generated by the social graph (Twitter, FaceBook, Quora etc...) to create networks that share more than just a relationship.