One way to think about the search engines is that it is their job to show you the sites that have the most activity around them (buzz) given your search terms (kind of like a buzz-o-meter). This is a very challenging task given the shear size of the Internet. Google, for example looks at all of the links coming to your site, the credibility of the site they are coming from and the anchor text used in the link to you (along with where on the page the link is etc.). By ranking and summing up all of these links, the search engine can score your web site relative to others based on these off-page factors. Search engines infer buzz (and indirectly trust) by using formulas to evaluate rank who mentions you and how often.
Social media based search is basically the same concept in that they are measuring the buzz, but rather than formulas, they add the more direct component of personal evaluation. Sites are ranked for search by how they were evaluated by people who visited them. If you can trust (or develop trust for) the people who ranked the sites, then you can be more comfortable that the site is appropriate for your search and that it is trustworthy itself.
It will be very interesting to see how these two models evolve over the next couple of years to address the challenges of real-time search, image and video search and geo-targeted search.