Observations on safety friction and misclassification in conversational AI

2 pointsby ayumi-observer9 hours ago