Full Idea
If there is to be total identification with others, then if another's preferences are mistaken, the preferences I imagine myself into are equally mistaken, and if 'identification' is the point, they should remain mistaken.
Gist of Idea
If morality is to be built on identification with the preferences of others, I must agree with their errors
Source
comment on Richard M. Hare (Moral Thinking: Its Levels,Method and Point [1981]) by Bernard Williams - Ethics and the Limits of Philosophy Ch.5
Book Reference
Williams,Bernard: 'Ethics and the Limits of Philosophy' [Fontana 1985], p.89
A Reaction
Yes. The core of morality must be judgement. Robots can implement universal utilitarian rules, but they could end up promoting persecutions of minorities.