Euan Semple points out two articles that examine the state of knowledge management. In Whence goeth KM? Dave Snowden concludes that knowledge management is on its way out because it has changed so much since it first appeared in the early 1990s.

It came to prominence at the same time that technology was beginning to emerge and dominate organizational thinking. People were trying to figure out how to use knowledge management technology, but were applying the rules of hierarchical, centralized business. As a result, the technology didn’t focus on connecting people naturally. Instead it tried to manage knowledge as a workflow, i.e. one person authors a document and others read & approve it in a linear fashion. This didn't make sense because it's too rigid — for example, what if you finished a document and sent it along the review & approval chain, then remembered something critical that you wanted to add? Would you add it and start the whole review process again, or just forget it?

Furthermore, those early KM systems tried to treat tacit knowledge (stored in peoples’ heads), as something that could be extracted and turned into explicit knowledge (written down), then turned back into tacit knowledge simply by another person reading it. The flaw here is that tacit knowledge is intimately connected to the person in whose head it is stored, and people need to directly communicate with each other to transfer tacit knowledge.

Something wiki this way comes
Today, social tools like wiki focus completely on letting people work together online the same way they'd work in person. The fundamental difference here is they approach knowledge as the product of that organic, non-linear human connection and collaboration. Knowledge management as a concept deserves some credit for getting people together to work on the issues that have resulted in the more mature social tools we’re developing and using today, but what stopped it in its tracks was when, “people tried to create standards and certify competence in the subject.� I was going to be nice and only criticize knowledge management for having the name knowledge management, but sheesh...certifying people? in basic human behavior? This is where, in my view, social computing took over and has dominated as the overarching concept ever since.

For an example of an article that’s still stuck in the old mentality, Semple points to Modernizing Knowledge Management, which is pretty much a recipe for failure because it’s a how-to guide for the old way of thinking.