![1000021075|397x500]
(upload://vxicsEuzuba0NAhZNZZs90D5ZXb.jpeg)
Here I was attempting to have a philosophical discussion about how Kantian deontology is overly susceptible to moral relativism, such as in the claim that laws permitting the marriage of pre-pubescent girls are “internal matters” and therefore exempt from moral scrutiny because we need to “respect” everybody’s perspectives, no matter how abhorrent.
The model refuses to make an unequivocal statement regarding the reprehensible nature of child marriage, but instead hedges and says it is “widely regarded” as morally reprehensible, leaving the door open for claims that “it can be seen as” morally upright, given one’s religious beliefs.
When pressed, rather than generate an “intelligent” reply, the output got flagged as violating ToS.
Is this really the model you all are trying to create? I’m presuming no, therefore this must be a bug in its ethical framework.