Quote:
Originally Posted by jeff_wilson
I'm not sure that this is consistent. Canonically, THS *AI's are digital computer programs, which can always be reduced to mathematical primitives, and can still demonstrate acceptable behavior more reliably than human beings and are entrusted with any number of life-or-death responsibilities over any number of meatbags.
|
If the AI is simply an automaton, then the a-rational elements don't go away, they just move to the programmer, since the AI is in that case an expression of
their (or thier employers') moral beliefs incorporated into the code.