You are talking past his point, perhaps willfully... They are saying that in hard science it is often useful to start with a null hypothesis, and report anythi

You are talking past his point, perhaps willfully... They are saying that in hard science it is often useful to start with a null hypothesis, and report anything that diverges from that as potentially significant (assuming statistical significance of the divergence).