Wednesday, December 10, 2014

SUV vs Mini Cooper Situation
            Right off the bat I feel programming a car to hit into a specified vehicle or object such as an SUV or a smaller car like a Mini Cooper.  I have to say that, I do agree with the crash optimization idea, where the autonomous car would make a decision where it will crash into the car that will cause the least damage in total.  This idea is going along the path of the ethical lens of outcomes.  This idea will cause the least harm in the end.  On the other hand, programming a car to hit either the SUV or the Mini Cooper could be seen as profiling to some extent.  For example, if the autonomous car company were to program the car to crash into the Mini Cooper because it caused the least damage to the passengers in the autonomous car, many people would be scared if they owned a smaller car and would be dissuaded in buying another small car in the future.
Biker with Helmet vs Without Helmet
          Now instead of cars like above where the likelihood of living for the passengers of the cars is generally adequate, the bikers in this situation will most likely face imminent death if the car swerves into them.  In this case, going either way would be wrong.  If the car is trying to do the least damage, it would most likely swerve into the biker who has a helmet on because he has a lesser chance of dying on impact.  In the outcomes perspective this would be the obvious choice.  The problem with this though, is that this will dissuade riders from wearing helmets because they will be put in danger if they do.  OH THE IRONY!

Random Number Generations in Decision Making
         I agree with the random number generation system for autonomous cars due to the fact that randomness does not profile.  This would prevent lawsuits by proving that the car company had absolutely no say in what decision the car would make and who it would crash into.  Also, this random number generator is just like a human being's decision making.  It does not make preferences on who to hit making no one liable for injuries in the crash.  I feel that a random number generator is the safest and most effective way to stay away from lawsuits and will be the most effective in the end.

If the driver is not making control decisions, should he/she be responsible for any outcomes at all? Why or why not?
        This question above, I feel, is all dependent on the technology of the autonomous car.  If the driver is fully capable of immediately changing the direction/speed of the car and they happen to react to the oncoming car/ object of imminent crash, and hit another car causing damage, they can be held responsible by all means.  Now if the car does not have the feature to immediately switch from autonomous to the normal driving mode, it should not be the driver's fault.  Everything considered, a driver makes the choice to let the car do the work of driving while they are still capable of making choices, they should be held accountable for the outcomes.