I agree with Tom Brossart that you need to explain why you think it will take us a long time to make certain decisions regarding the overlap of practical and ethical considerations in the programming. This seems to be a common fallacy: that because a problem seems ‘difficult’ it must necessarily take a long time to solve.
I rather imagine that we will very quickly come to the conclusion that autonomous vehicles should stay within their travel lanes and this whole argument of which way to swerve will disappear. In the event that an obstruction appears in its path, and autonomous vehicle will reduce speed as quickly as possible. Given that the car will a) be going the speed limit, b) be able to detect the obstruction and react almost instantaneously and c) be able to apply maximum braking force almost instantaneously, I think that the chances for our kid’s survival is pretty high compared to an inattentive and slow to react human driver going the usual 5 or 10 over in a residential area.
Of course there are other issues, but with the immense resources being invested by so many different parties, I think you will be surprised at how quickly the change occurs, at least in some areas.
Like a lot of other social change, we will no doubt see autonomous vehicles be adopted in affluent areas that can afford infrastructure upgrades, while areas of poverty that could really use a distributed transport network, will be left behind.