What's to stop us from applying the same rationale to humans once the line between artificial intelligence and human intelligence becomes sufficiently blurred?