>> I don't really see that as a bad thing. This is a policy decision and as such needs to be based on a rational calculation and values, not the feelings known

>> I don't really see that as a bad thing. This is a policy decision and as such needs to be based on a rational calculation and values, not the feelings known to be hopelessly biased and generally unreliable. > I hope future AGI doesn't behave this way or we're all fucked. So, given what you're responding to, you prefer AGI to not use some values (e.g. DALYs or whatever) and rational tradeoffs, but instead to be based on fuzzy, biased, unreliable heuristics? I agree this argument is generally not going anywhere, but this kind of response in particular makes me wonder, because I have hard time finding a a charitable interpretation I could respond to - it just sounds like vacuous platitudes nearly without context. Is there a good argument for empathy that isn't?