Semweb solution is to add annotations to everything (tags), so its obvious to a machine what the information means (ie, the content providers/authors do more wo

Semweb solution is to add annotations to everything (tags), so its obvious to a machine what the information means (ie, the content providers/authors do more work up front to provide more automation later). Versus the crawl the web and index approach with unstructured data (which google are very very good at, clearly). I see it a bit like the static/dynamic typing "disagreements" people have (heck I disagree with myself, frequently).