The holy grail of startups is the ability to crack Google's virtual monopoly on searching. The way technology changes, one assumes it will happen eventually, but it's difficult to conceive of how any small group in their basement could challenge the Behemoth.
But there is a way... unless Google does it first. The rise of Google was matched only by the rise of the collective and open software movement. And the power of Google is matched only by the productive power of independent coders, particularly when they can get some money for their time.
Imagine an open search engine forum. Any coder could submit a bit of intelligence-- a new search algorithm, a new net-scraping bot, or even a new storage system. Like creatures in an ecosystem, these different agents compete to produce search results: Every query provides some CPU cycles to multiple algorithms, and the search results combine their answers. A click is like food or a vote in favor of the algorithms that produced it. The best algorithms float to the top, and their developers get a corresponding share of potentially immense advertising profits.
I don't have time to build it, but I would encourage anyone who does and wants to be fabulously wealthy while helping make a world where the next search-god is not just as monolithic as the last one.