I guess alot of people, including this guy, are openly questioning "validity" of YComb's startups. Since everything they do is so trivial, can their creations be considered "products with value"?

My personal issue with YComb's approach is that they actively (via this site, for instance) promote engineering mediocrity. Remember posts "PG made me drop out of college today!" ? I have seen several of those. Web 2.0 is all about this: cheap&quick high-school level programming project done over a weekend, polluted with google ads with mandatory "blog" attached to it.

To me that is sad, There is nothing "glamorous" about it. My major in school was AI, and our professor used to dream alot about more powerful hardware that was supposed to become a reality soon. Well, it became. And for what? To run 8 Mongrels with some quick&dirty Ruby scripts converting contents of 4 DB tables into nicely looking HTML with roundeed corners? What a waste...

Business does not equal mastery. if you complain about MBA BS then it should be obvious enough. The iPod was nothing new in terms of technical mastery. But it was innovative in terms of interface design, and a damn good example of successful marketing and integration into a whole product family; as such, it's also an innovation of process.

But does an iPod engineer understand any AI? Probably not. Steve Jobs made stupid comments in his presentations (still remember the "branch prediction... it predicts branches!"). But he has an eye for opportunities that not even your professor can boast having. If there's a market, and you know what it takes to supply it, supply it. In my view, YC is addressing this; it may saturate quickly, but the need obviously exists. If your expertise targets the extremely educated, by all means make something that satisfies them.

But don't forget that there are a lot of people to please. Just please the ones you understand, or think you do. I'm completely taken aback by twitter's success, but hey, people use it, what can I say.

It's all about opportunity costs. Neither the world, nor Linus would probably be better off had he started a company rather than hacking on Linux. The world, and PG, are probably better off that he pursued Viaweb rather than some earlier version of Arc. But it's ultimately up the individual to choose what they want most. This site and PG certainly promote one route, but sensible people realize it's not the only thing out there.