Rohin Shah comments on Updated Deference is not a strong argument against the utility uncertainty approach to alignment

[Rohin Shah]

I think this comment is a better representation of my views and roughly agrees with your perspective (though the comment you link is also accurate and I still endorse it).