Slashdot | Another Google Recruiting Technique:

I've answered my fair share of brain teasers, pattern recognition, cute/stupid questions. This is crap that pisses me off. I'm sure google wants smart people, but they're going to overlook all the people who just get shit done when it needs to be done (and do it well) because they're going for the people who are creative.

No slight intended, but methinks that maybe you're missing the point of these cute, stupid little brain teaser questions. It's certainly one thing to sit down and slam out something that you either know how to create or can come up with a way to create it relatively easily. It is another thing entirely to solve or approximate a solution to a fundamentally difficult problem and then implement it.

To put it in vulgar terms, Google likely doesn't need people to just get shit done, they need creative solutions to problems that don't have a straight forward answer. Do you honestly think figuring out how to retrieve highly relevant web pages from the whole net based on one or two silly little keywords, in fractions of a second no less, is something you just sit down and program? Do you even think it's something you can beat the current top players at by simply engineering it with current methods? No, it's far more akin to those little problems you hate so much. Sure, there's naive solutions that give a lackluster result, but to do it well it's all a game of tradeoffs, new and novel approaches, application of known approaches or extensions of known approaches in the right instances, a dash of brute force, and who knows what else. If it was straight forward and something just solved by getting shit done then Google and its staff of many PhDs would likely be vanishing due to the costs of keeping all those PhDs around opposed to another company running slimmer and just doing it. Instead they're number one in the search engine world because of their pioneering efforts.

The people who enjoy and excel at those questions, seemingly silly on the surface but generally with deeper implications, are the type that are typically good at doing the sort of research that needs to be done to solve the tricky steps.

Again, none of this is meant as a slight and there really is need for people who are good at architecting and implementing solutions with good work ethics. Many applications are at a point where the technical challenges lie in integration of known solutions and those certainly still need good, hard working architects. On the other hand there's still a definate need for people who like toying with silly questions because that generally translates into enjoying playing with the nuances of more research oriented stuff, simply because they're so often similar. Frankly, if you dislike those stupid little problems then you may well dislike the research experience of bashing your head into brick wall after brick wall trying to come up with a novel solution to a problem which has no real feasible solutions at the moment.