Computer: there is a small group of unpopular people causing social problems. Liquidating them humanely would increase net happiness at the least cost.
Shall we proceed?
Shall we proceed?
Who picks the frameworks used? What about more contested topics which have a closer social split like abortions? Who decides what is and isn't ethical, and how does this cope with social divides or in places with toally different moral frameworks?
Honestly I think this problem is probably harder than the AGI part!