I love this part: "To make something really different, and not get drawn into the gravity well of existing solutions, you practically need an isolated monastic order of computer engineers"