One of my favourite buzz-words of the last 2 years was "tagging". It means that you assign keywords to a certain content and it was born in the "blogosphere" (another great one) to determ, what an article is all about.
In my opinion is a text article one of the contents which do NOT need a tag, it is made up out of text anyhow and the search engines should be able to assign relevance to it whatsoever.
More interessting is the tagging of files that can not be recognized automatically. Especially images, executables or ZIP archives come to mind.
I have been reading about the new Windows Vista file system, and I think Microsoft got a good idea there with the power to establish it. The whole Internet needs meta data on files asap! Personally I would prefer Google to come with something which makes sense, because they have a better base to start from in terms of "what can we find online and how can we organize it", but I take it from MS as well, if it is good!
In my opinion is Googles image search quite impressive and would be much better, if people would be able (or be forced) to add text to binary files - photo sites do that for quite some time now.