I agree it often seems crazy companies with no clear revenue model raising millions in funding but I'm not sure the approach of looking at each one individually and saying "it doesn't look like they could generate sufficient revenues to justify that investment" takes into account why they're getting the investment.

At such an early stage most startups are at best guessing what their main revenue source will be, the only thing they've demonstrated is that they're able to build something people want. My understanding is that the bet VC's are taking isn't that all or even most of them will work out a way to monetize this sufficiently to make a return on investment, it's that one or two of them will work it out in a big way and be successful to the point that they more than compensate for the majority of others which just about broke even or never cracked it at all.

If anything these investments are just an acceptance than no-one really knows what's going to work and one solution to this is to put money into teams which seem to be able to make something people want and hope that by allowing them what look like unrealistic costs to begin with, they'll come up with something game changing. Not suggesting that's the best approach to building companies but I don't think it's a completely unreasonable one.

The point's made a lot better than I can in this lecture (http://blakemasters.tumblr.com/post/21869934240/peter-thiels...) from Peter Thiel's startup class (with Roelof Botha and pg).