(March 09, 2005) - You might have noticed our references to A9 in yesterday's column. A project from the Amazon search people, A9 is a more intrusive search engine. It wants to help you make those leaps that turn search into find. It does so by monitoring your web behavior.

If you have a large enough screen, A9 can show you (as a set of search results) images, movies, books,  Yellow pages and the note you've made in the past about a website. The juxtaposed ideas in the search results (we're particularly fond of seeing images mixed in) help launch the mental leaps required to move into innovative discovery.

On some levels, A9 is Amazon's attempt to extend its recommendation engine into general search results. It's a powerful notion. The system ties your Amazon behavior, internet search history and browsing history into a engine of Discovery. Nicely, the whole process is wrapped around a small discount on Amazon purchases.

The results of a search will tell you whether you've seen a particular site and how long it's been since you've visited it.

Although it's unclear what the actual biases are in the search result filtering process, we really like the sense that search becomes somewhat non-linear when using A9. Usually, our searches (whether for people or ideas) start as a quest for insight, not prioritized answers (and the inherent bias of this or that search engine). We want to ponder awhile, mulling over the possible directions before selecting a path.

There are often several rounds of mulling before what we seek becomes clear. Sometimes, it's just about the mulling.

The paranoid will certainly note that A9 involves a level of trust in Amazon and its intentions. We've always assumed, on the other hand, that well tailored results require unprecedented levels of intimate knowledge.