(RNS) — Basing AI’s principles on ‘the moral view of the whole world’ is both incoherent and dangerous.