Reference no: EM133318074

Autonomous vehicles hold out the possibility of eliminating many of the tens of thousands of highway deaths that occur annually in the U.S. Unlike human drivers, automated systems will not be tired or distracted, and will ordinarily be able to exercise precisely the kind of vehicular control necessary to prevent an accident. Some accidents will still be unavoidable, however, and in those cases autonomous vehicles may be programed to engage in "crash optimization," or to have the "best" kind of accident possible. Yet, there is much debate to be had about what constitutes the "best" crash result. Is it one that kills passengers over pedestrians, or vice versa? Is it one that kills the fewest people, or one that involves saving those of the greatest value to society? What do you think? Or should autonomous vehicles not engage in any sort of crash optimization that involves making qualitative judgements about people's lives? If so, though, is that not itself a choice - and are the programmers of such vehicles now responsible for the harm they fail to mitigate, and the lives they fail to save? Explain.