This hasn't stopped years and years of bleating about "AI safety", whatever that refers to.