No Longer a Philosophical Conundrum!

My grandson, in his 12th grade philosophy class, is given the following type of discussion assignments: 

1. Suppose Bill is a healthy man without family or loved ones. Would it be ok to painlessly kill him if his organs would save five people, one of whom needs a heart, another a kidney, and so on? If not, why not? 
2. You're in the cab of a runaway tram and see five people tied to the track ahead. You have the option of sending the tram on to the track, or forking off to the left on which only one person is tied. Surely you should send the tram left, killing one to save five? 

I told him that by the time you complete your graduation, this type of moral and ethical questions will no longer be topics for delving into complex intellectual armchair discussions in the comforts of your room, but you will be tasked with the responsibility of making split-second decisions in real-life cases!

As a computer science man, you will have to develop AI / ML (Artificial Intelligence / Machine Learning) models for driver-less cars, when you will be confronted with such moral dilemmas of deciding whether the car, passing through a narrow mountain road, on 'seeing' two little kids suddenly darting onto the road, should mow down the kids or should veer to the side, fall into the cliff, and kill the owner of the car sleeping on the rear seat! 

No doubt, such models would be fed with real-time data from cameras, sensors, GPS, radar, lidar, cloud services, etc. using 5G technology and are processed to produce control signals that are used to operate the car. 

And would also be supported by some of the widely used pattern recognition algorithms in self-driving cars including histograms of oriented gradients (HOG), principal component analysis (PCA), Bayes decision rule, and k-nearest neighbor (KNN). 

It is finally left to the humans to use their intuitive skills as much as their analytical abilities!