I think another important lesson to draw from this is that you should build things for domains and problems in which you have a legitimate interest/curiosity. That was something I took away from this that's difficult to accept, because it means you're probably going to be limited if you want to work solo.