You may or may not know that yesterday was “A Day without Google,” a simple campaign started by AltSearchEngines.com to get people to try one of the alternative search engines not built or run by the big players (Google, Yahoo, MSN, etc.). I didn’t participate, honestly, but I’m struck by a couple of common themes coming up in the various reviews by folks who did.

There seems to be a shift or a surge in prevalence, if not popularity, of search engines that want to be “experts” rather than curators or guides of the Internet. For example, take the experiences of Josh Catone yesterday with two natural language engines, Lexxe (lek-si) and PowerSet. Both of these engines prefer that users ask normal, English questions that the search engine will answer. Ask.com is probably the most popular of this breed of engine, though they probably don’t match Lexxe or Powerset feature-for-feature.

Here’s what Josh, who writes for Read/WriteWeb, said about his experiences with the alt engines:

As an example, last night I caught part of a fascinating documentary about Israel’s 1967 war with Egypt, Jordan, and Syria. Later, I couldn’t remember the name of Israel’s prime minister at the time, so I fired up Lexxe and asked: Who was Israel’s prime minister during the 1967 war? Lexxe suggested that it was Yitzhak Rabin — I knew that wasn’t right. The second result, however, mentioned Levi Eshkol — which is, it turns out, the correct answer.

…

When I tried my first query in Google this morning, the first result I got was Wikipedia’s Six-Day War entry, which would lead me to Eshkol. And the fourth result highlighted text on the search results page mentioning Eshkol as prime minister.

This is a trend that’s a lot of investors and entrepreneurs are moving toward. While we focus on the power of natural language as a usability step, it’s making a couple of other leaps that I think are much more significant. These search engines are taking the next step of actually being the repository for knowledge from the Internet rather than a guide to the knowledge on the Internet. In other words, Google will do it’s best to help you find a site that can answer your question. Lexxe and Powerset will attempt to answer your question.

That’s a pretty profound difference, in my mind. A couple of things come to mind here. First, my gut instinct is that a site that tries to be an expert on everything, which answering any question implies, will likely be an expert at nothing. While I know that enough people and enough special algorithms could replace us, the fact of the matter is that there’s an enormous amount of knowledge and facts out in the world. It’s unclear to me (and, really, beyond my imagination) to believe that one search engine will be able to actually answer questions in the near future. I’m not the first to think of this, obviously, as smarter folks like Jakob Nielsen have been talking about Answer Search Engines since 2004.

Second, the tour guide-like functionality that Google and Yahoo and things like DMOZ provide is different and in certain cases better than getting instant answers. Google does a particularly good job at bridging the answer/guide gap by using special treatment in their OneBox area. It’s a smarter guide, but one that’s not trying to answer every query, just the ones that they’re particularly good at.

The third thought that comes to mind flows from Google’s OneBox. Jason Calacanis, the entrepreneur who brought you Weblogs, Inc, publisher of blogs like Engadget and TUAW, just launched his next big venture, Mahalo. Mahalo is a “human-powered search” engine, which has “guides” that create search engine result pages (SeRPs) by hand. Here’s his description:

Jason McCabe Calacanis today launched Mahalo.com, a human-powered search engine, at the Wall Street Journal’s D Conference. The site is currently being launched in Alpha with the Internet’s 4,000 most popular search terms completed. The Santa Monica-based company hopes to reach 10,000 search terms by the end of the year. At that point it will enter Beta, and launch shortly thereafter.

In other words, they are manually creating the search results pages for the top 10,000 keywords. At some point, I believed that this was an open, wiki-style project, with direct compensation, but it looks like the first set of guides are employees. They just launched their Greenhouse project, which aims to allow anyone to try creating a SeRP to earn a small fee.

The interesting thing here is that Mahalo is explicitly targeting the curator function, to create a reviewed list of results, including a fact box. Here’s the SeRP for iPhone, for example. While it’s not what you’re used to from Google, it provides a decent mix of guide-like results (here are the top sites that talk about the iPhone) and answers (when will the iPhone ship, etc.). While you can’t ask your question as a question, I actually like that because it avoids all the annoying extra typing. I suspect given another generation of net savvy users, a majority will soon get the index approach.

This was just on my mind today after reading about the Day without Google. I personally think the curator function is the most important, but I know how much people love Wikipedia and getting answers from the Internet. I guess we’ll see how it all plays out.