Social search engines such as Mahalo have very recently come to use techniques that are not so much algorithmically excellent but rely on human intervention. There are other models of leveraging social power with algorithmic power as seen in social bookmarking sites. While these represent the dominant approaches in new information retrieval techniques, its hard to say whether any one of them will replace modern day search engines.

I, for one, was once very optimistic about the social bookmarking approach. The simple reason being that social bookmarking on a global scale would be better at categorizing content based on relevance than algorithms that power the major search engines.

One simple scenario to support this point is that suppose a great archaeologist decides to throw some light on some arcane topic, the process he would use is to perhaps open a blog account and make a decent post. This may be picked up by the community that is better at assessing the information in the article than a search bot that uses algorithms to rank on keywords and link juice. Also, bots may be foxed by spam but that cannot be effective with human bookmarkers.

However, what has emerged on major bookmarking sites is that a relatively smaller group gets to decide the importance of content that goes online. It is the bias factor where the search bots score. Bookmarking sites are taking steps to reign in this kind of bias but it is to be seen if anything significant comes out of this.