Eliezer, the correct way to resolve your inconsistency seems to be to be less approving of novel experiments, especially when they aren't yet necessary or probably very useful, and when a bit later we will likely have more expertise with regard to them. I refer to a comment I just made in another thread.