This doesn’t appear to have any filter for generating based on racist / extremist language