VCs are an industry unto themselves: I agree that that implies a shift towards start-ups. My question would be: what's there to stop a shift in corporations' becoming merely comprised of much smaller modules? Google works by having many small clusters of teams working on big things, so they're able (in some cases, not all) to innovate as quickly as start-ups in their fields, but with the advantage of firm corporate backing. Not many large companies are doing the same, but what's stopping new corporations from forming specifically with modulation in mind? Or would you argue that a development like that is similarly higher-res?

Steve Yegge disagrees with your idea of Google's operations.

http://steve-yegge.blogspot.com/2007/06/that-old-marshmallow...

Maybe it's just the fact that I am a naive hacker-in-training, but can you interpret the story a little bit? It really bugged me the first time I read it, and still does.

In this podcast [1], Steve explains that this was a story of an Evil Manager who "slipped into" Google, and took a long time to be discovered. He suggests that he was fired after the story (although he doesn't say it outright).

In this interview, Steve also points out a point along the lines of the ancestor poster's - that Google encourages employees who are ambitious enough to start "startups within Google".

1. http://blog.stackoverflow.com/2008/10/podcast-25/

Forced to eat marshmallows == Suffering from overhead imposed by the management little at a time. It looks like no big deal when viewed individually... but quickly becomes nauseating in aggregate.

Also you can feel the helplessness that management creates perhaps inadvertently in the engineering team.

Actually the eating marshmallows thing is an allusion to having to eat the Google dogfood: there were new Google tools all the time that his team was forced to use.