Would you rather build a very successful & boring startup, or work on something super cool, with no clear way of turning a profit?
I'm comforted by Paul Graham's thoughts on this, namely that if you build something people like, you'll usually manage to find a profit somewhere.