@tess What's an interesting view is that humans already do exactly what AI do, but we can condemn them if they make a mistake. AI will not ever be condemned, nor their creators, and that's as much a danger.