How does the morality work when a car-ai crashes and the human driver could've prevented it? What if the AI makes a "difficult" decision to run someone over in order to prevent a worse crash, and what if, or not, the human prevents it from doing so and then something worse happens? Is he morally responsible?