Mashable's Stan Schroeder expounds an interesting theory: that Google's current (and expanding) dominance in web search, at least in the English-speaking world, has trained websites to do all they can to show up high on Google - to such an extent that no upstart search engine can hope to do better than Google for broad horizontal search.
Stan's review of recently launched, well-funded search startup Cuil, as also the TechCrunch review, are lukewarm at best and support Stan's thesis - but this thesis assumes that Google has the best possible knowledge of a user's intentions, and since websites will compete to match Google's algorithm that expresses that intention in a query, Google is by default the winner in any search competition.
The flaw in this argument is that Google has very little information about a user's intention - indeed, Google doesn't even seem to use all the information it could have about the user, because of privacy and latency considerations (for instance, I doubt if Google looks up my interests in my Facebook profile when I search as a logged-in Google user, to discover that by typing the search term "kayak" I probably mean a watercraft and not a travel search engine). Google is constantly refining its approaches to divining intention, of course ... and the masses of data generated by user searches help it get better every day.
But web search is far from perfect today, and it stands to reason that Google's own momentum - and success - will lock it into the innovator's dilemma of doing little more than tweaking its existing algorithms. And some enterprising startup will bring a refreshing new take to searching the web. Wonder what it could be?