Google’s Human Algorithm

Henk van Ess has been running a series of blog posts at Search Bistro about how Google uses international students to QA the relevancy of its search results. See, for example, his initial post including the advertising that Google ran through the Kelly temp agency.

This doesn’t surprise me at all; rumors have been floating for a while. However, it does reinforce that Google is far from a passive intermediary for displaying third party content. Rather, Google actively manages its content database to accomplish its goals (i.e., in this case, high relevancy of search results). This should put a spotlight on its various copyright and trademark lawsuits where Google’s defense has been partially based on the automated nature of its search engines.

The confirmation of human oversight of search results also undercuts Google’s position that technology can solve all problems. In the end, when it comes down to subjective decisions like “relevancy,” an algorithm alone isn’t sufficient. This is hardly surprising, but Google has tried to position itself as weighting technology solutions over human solutions, and that positioning simply doesn’t work across-the-board.