In the first scenario, we discussed whether the automated car should be programmed to hit a certain car when it has enough time to make a decision on which car to hit. In this scenario we are given two cars, an SUV or a Mini Cooper. Statistically, SUVs are safer... for the person driving the SUV, at least. James explained to us if a small car rear-ended the SUV at a good speed the car would suddenly stop. This will cause severe injury and possible fatality to the driver (and every one in the car). If you were to hit the Mini Cooper, the Mini could be completely destroyed because it is smaller. Your car would not stop so abruptly minimizing your own injury but putting the Mini Cooper driver in danger. In this situation I think hitting the Mini Cooper would be the best decision. This is because the cars are probably closer in size and you wouldn't be in as much danger hitting the Mini Cooper. With the second scenario we were given two motorcyclists, one was wearing a helmet and one was not. This is a difficult scenario. One person is wearing a helmet for extra safety but the other one isn't. Both motorcyclists, if hit, are most likely going to die. In this scenario, you're programming the car to kill where as with the car, it is more about damage and injury. The random number generator seems like the best idea to make a decision. This will be better for SUV owners and those who wear helmets on their motorcycles alike. Now those people won't be targeted and it will be completely random. I think the driver should be responsible for the outcome. This is because the person behind the wheel should be paying attention even though they're not technically driving. If you can't put the blame on anyone, this could cause difficulty with insurance purposes specifically. Also the "driver" of the car made the decision to use the feature presumably understanding that if the car is faced with this obstacle, it is going to make its own decision.