Should driverless cars kill their own passengers to save a pedestrian?

[Aractus]

---

Rather than derail the baby Hitler thread (albeit already so) I thought I would start a new one.  This deals with a similar ethical situation.

I had mentioned the "trolley problem" there and a couple other places and there seemed to be a complete lack of acknowledgement.  Perhaps it was too complex, strayed too far from the OP or just plain ignored on the hypothetical basis it was set in.

So, here is a real life, non-hypothetical "trolley problem" that will affect society immensely.  With the advent of driverless cars, how should the cars be programmed?

A car breaks down around a curve on a steep hill.  A child runs out in the street or the driver steps out to grab the tire that got away.  The driverless car comes around the curve and cannot stop in time.  Should the car swerve and end up going over the edge and kill the passengers or should it sacrifice the pedestrian in favor of the passengers?



This is not about "The pedestrian should not be there", that is a moot point, the pedestrian is there, what now?

What?  You would not get a driverless car so it does not matter?  Except that it was your manual car (or a loved one's) that broke down around the corner as the kid got a wild hair up their ass and ran out into the street (albeit only for a moment).

---

The answer is no it shouldn't. And my solution to the "trolley problem" is no I do not by default redirect the trolley, but I would think about it before making a decision. It's true that 1 life is worth less than 2 or more lives, but it also doesn't mean that one life in particular is worth less than 2 or more other lives in particular. Also my life is worth more to me than the lives of several other people, thus I would prefer the driverless car to kill the pedestrians and allow me to live. However there is the predicament that one could argue that it's equally likely that I would be the pedestrian - however I find it very unlikely that in the real world that several pedestrians would be killed all at once, thus my best probability of survival is to have the driverless car kill the pedestrian, with the possibility that as the pedestrian I could escape fatality and only become injured/paralysed, etc.

Again though in a specific instance this might be different:

[url=http://www.canberratimes.com.au/nsw/skydiving-instructor-tony-rokov-killed-in-accident-at-goulburn-airport-20151121-gl4raf.html][/url]

Tony Rokov died after he plunged to the ground while tandem skydiving on Saturday afternoon at Goulburn Airport.
...
A 14-year-old who was strapped to him, and was believed to have been skydiving for the first time, was critically injured and remains in hospital.

It is believed Mr Rokov wrapped himself around the boy to cushion him from the impact.

It now looks like Elijah Arranz (the 14 year old) will survive the accident (which was a 10 meter free-fall to the ground). The philosophical question here would be could the 44-year old have instead survived by using the boy to cushion the impact? The answer is perhaps - however the probability of survival would be considerably less than the other way around since he is the one who weighs more, he has the parachute pack on his back (which itself provides some cushioning as it presumably still contains the reserve parachute), and he would have had to have hit the ground face-first (and since he is taller it probably means hitting his head anyway). On the other hand, one could also make the philosophical argument that a face-first dive into the ground would have provided the boy the certainty of a merciful death. So there really is no clear-cut right or wrong answer as to whether to hit the ground facing forward or facing back in that situation.