As you probably know by now, ex-Google employee Anna Patterson + husband Tom Costello and some guy named Russel Power launched a new search engine this morning. It’s called “Cuil” (pronounced “Cool”) from the old Irish word for knowledge - leave it to a search engine to come up with a name that’s only interesting because they say so (and gosh darnit, we believe them).

A big selling point for Cuil is it’s large index: 120 billion pages, according to their homepage; a number they also claim makes it the biggest search index in the world. Some people refute that with Google’s recent claim at 1 trillion web pages crawled. Yet it’s worth noting that Google doesn’t actually index each of those trillion web pages, as Michael Arrington recently pointed out. Apparently, Google’s true index size is only about 40 billion, since it doesn’t bother indexing pages with duplicate info. Unfortunately, statistics are often misleading, as a quick search for this blog shows:

Somehow, I doubt there are 9,373,214 pages that mention my full url (when it looks like the only page of this blog they actually have indexed is, well, the index).

And the shove-it-Google’s-face sales pitch is that is doesn’t provide results based on popularity (as Google’s PageRank setup dictates), but rather on relevance. Yet I think a simple search for “cuil.com” illustrates the basic problem with that philosophy: no computer algorithm for relevancy will ever be as accurate as actual people. That’s what Google did right.

I’ll give it a few weeks to for things to settle out, then check back and see if the results are any more relevant. It’s a good concept, and I am rooting for them.

Admittedly however, there are some things that are catching my eye with Cuil. As much as I may love being able to go back in time and see all the embarrassing Google searches I made on August 24th, 2007 (for example), having all of that information saved is a little unnerving. Cuil doesn’t save any user history.

It also uses the same predictive typing technology Firefox has in it’s search bar - a reason I still use Firefox’s QuickSearch, even if I’m already on Google’s homepage:

Unfortunately, about 40% of my page requests result in some sort of non-fatal server error, but that’s probably just first-day jitters. I’d wait until the rabid tech bloggers have stopped overloading their servers before putting it to a real test yourself.

P.S. - Looks like the entire search engine is actually offline for the moment while they add more capacity.

P.P.S. - They’re back, check it out!