A few ex-Google employees just launched a new search engine called CUIL (pronounced cool). The UI is sexy, black with white text and laid out in three columns, allowing you to quickly scan the results. It’s got some nifty AJAX sliding menus on the searches that allows you to view sub categories in your search.
According to their website, CUIL “Cuil searches for and ranks pages based on their content and relevance. When we find a page with your keywords, we stay on that page and analyze the rest of its content, its concepts, their inter-relationships and the page’s coherency.”
CUIL ranks pages based on relevancy of the content so I did a search for my own name and got links to content I’d written or mentions of me or the company, but several of the images that came up with the search were totally unrelated to the content on the page or my name. (There’s no mention of the potomac on the page and we’re in California.)
I did another search for the Exploratorium and discovered some nice features. A box pops up with related searches such as location, other museums in the area, local information and maps. Sweet!
They may need more time to index as fully as they intend to. A search for Twitter on Google yielded 73,400,000 results and on CUIL 800,000. I didn’t go through them all to see if the 72,600,000 missing links were irrelevant. Maybe I’ll try a smaller search.
All in all I think CUIL needs more time to develop, but I like the direction. For the moment I prefer Chunkit which returns more digestible chunks of information, let’s you email the “Chunk” to a friend, and shows you an indexed version of any site you visit. More on that one later. 
