I’ve just spent a short while playing with Cuil, the new search engine that’s supposed to be both bigger and better than Google. The main thing they tout is a bigger web page index than Google. This may well be true, but it would appear that the index may contain many more pages which have little relevance, and fewer real pages. For example, searching for ‘imaginary roots’ turns up no pages on Cuil. On Google, you not only get this site, but hundreds and hundreds of maths sites with people talking about complex numbers.Similarly, searching for ‘the sombrero principle’ returned no results.
How can they have not indexed any of these pages? It may be that they have indexed them, but their search algorithm isn’t quite refined enough to fetch those results. As TechCrunch points out, Google has had a lot longer to refine it’s searching algorithm, and credit should be given to Cuil for being really very good for one that is only a few hours old. And it does have some nice features such as the related categories – but is this not something Google could build very rapidly, and make a very good job of? Probably. Whether they want to or not is another matter.