Rohin Shah comments on Evan Hubinger on Inner Alignment, Outer Alignment, and Proposals for Building Safe Advanced AI