Yeah I understand the viewpoint here. I'd argue that the costs of building a startup have been driven down so dramatically that companies that are hard tech org

Yeah I understand the viewpoint here. I'd argue that the costs of building a startup have been driven down so dramatically that companies that are hard tech organizations need to raise ever-smaller rounds to get started but, quixotically, ever-larger rounds to scale. Something like LivingSocial costs next to nothing to start, but costs millions to scale. Those kinds of wars can't be fought with 1M Series A rounds, but burning that much cash on a mistake is also quite rough. Now LivingSocial isn't hard tech, but most DB startups don't have massive initial capital expenditures, and with things like KickStarter market validation is much simpler. Is the incubator approach better? With the exception of YCombinator and MAYBE techstars, I'd say no. It's hard to say. As always, we'll look back in 10 years and pick the winners and losers by fund growth, but the fact remains, companies can afford to put off raising until much later in their life cycle.