That's really shallow. A nostalgia dive, not an analysis. It isn't that "tagging worked, but now it doesn't", it just never really worked for her usecases, and

That's really shallow. A nostalgia dive, not an analysis. It isn't that "tagging worked, but now it doesn't", it just never really worked for her usecases, and it still does for usecases it can be used for. Maybe I would be more sympathetic, because I also feel that "it was better back then" (who doesn't?), if it wasn't for people like her using phrases "exploring Web 2.0" and so obviously not understanding what really happened. And what really happened is this. A bunch of somewhat similar people (a tiny group, really — compared to the population of the world, anyway) were playing a cute tabletop game of "Web 2.0" at the table set up out in the open fields, pretending they are eager for more people to join. But they weren't exploring this game, weren't really thinking about all the nuances and implication its rules had, and weren't actually preparing for more people to join, instead they implicitly agreed to use their own impromptu simplified version of the rules, which wasn't that hard to do for this tiny group of likely-minded people. And this naturally broke when more people and corporations and governments joined the game, and started playing it being limited only by the actual rules of the game, which are akin to the laws of physics in that they really have to be explored and have an immense number of implications. So, what role does tagging have in all of this? What is its true nature? Sadly, its true nature is "a hack", an artificial construct that gives you a one-inch leeway to barely slide across your problem. The problem is that it's a bit too hard to search for content using natural language descriptions of the content, because there is a lot of content, and there are a lot of words in human languages, and they all intertwine in all sorts of ways, and people use the same words to mean something else than what you meant, and they might use some other synonymous words to describe what you meant, and so on. It is alright as long as there is a hundred of people and thousands of posts, and it's all text without pictures, but when it grows it becomes just unmanageable and the full-text search doesn't work anymore. And you have no ideas how to approach it other than full-text search. So you agree to use simplified version of language in addition to the original language to mark what your content really is about, so that its a bit less ambiguous and you can continue to use full-text search on all of the contents of the internet. That gives you just enough leeway to reduce the number of search results to something manageable and you are happy. And you can even automatically feed the search results to your RSS. And 15 years later, when it all grows more still, and the simplified tagging-language gets more complex and more abused (or underused) so that full-text search doesn't work for everyone again, so people stop using your tagging-language because it doesn't help them any more than the original content, so it's useless and unnecessary and they switch to other approaches to find content. Which are also imperfect, by the way, but at least they are actually another approaches, not just some hacks on top of full-text search. The real lesson here is: "tagging doesn't work out in the open fields; use it in small communities with curated content, like, for example, a blogging channel of your own, so that your readers can find your other posts on the same subject". And (surprise!) people do use it for that quite successfully.