I would also add that VC funds have a s..t load of money to invest. The valuations that they give in comparison with a bricks and mortar business are night and day different in most cases. For example a successful local business might sell for 3-4 times net profits or 1 times revenues. A tech company in a series A might get 15 times revenues (with zero profits). A hot tech startup can justify a lot more based on a lot less or zero revenue. The hope is that the tech company is going to justify that MUCH higher valuation by spectacular growth. Obviously it is justified with a handful of companies, but it's a very high risk investment for VCs to put it bluntly. However they have the funds, and they have their criteria with which to invest, and they mostly have a mandate and must invest the money.

I think Paul Graham's point's are interesting, but they are based on a world that he lives in, the world of tech startups and VCs, which is not the same as the wider business world.