The whole thing was just to illustrate the point that he thinks the Apollo API access is worth nowhere near $20 million a year in opportunity cost.