ReadWriteweb has a post on the new trends that have been shaping up at the search front and those that could possibly challenge Google's dominance in the future.

A summary of the list

1. Outsourcing of crawlers
2. Targeting single proven sources like Wikipedia
3. Peer to Peer search
4. Core search is still getting funded

Over all the article lambastes the approach to using natural language as users have more or less come to expect the machine to respond to queries where the difference in the terms is stark. Among these the use of P2P for search does look very interesting. It does solve one big problem with search - scaling to the size of the web.

The more fundamental aspect is the inertia to change. Even if there are engines that provide distinctly better search than Google, perhaps there will be no en-mass switching over unless Google itself deteriorates in quality.