The MIT technology review recently published the findings of a thought experiment: Given only two choices should a self driving car run over a grandma, or a baby?
Thought experiments are fun. Heck, they can be an important philosophical exercise to help us uncover what we really value. This MIT ethics study showed that who one kills in this scenario, largely comes down to one’s culture.
This whole concept (and the accompanying graphics) are absurd. Human error, including intoxication and distraction causes the absolute majority of accidents. This proposperous scenario is massively outweighed by the lives the will be saved by #selfdrivingcars.
— Nathan Corliss (@oakfive) November 15, 2018
Unfortunately, this thought experiment is a distraction from the real promise of self driving cars. Right now around 40,000 people die per year in the US in vehicle accidents. Many of these deaths are due to distraction, and intoxication – self driving cars will essentially eliminate these types of accidents in turn saving 1,000s of lives.
“Each day in the United States, approximately 9 people are killed and more than 1,000 injured in crashes that are reported to involve a distracted driver.” CDC.gov
“Every day, almost 29 people in the United States die in alcohol-impaired vehicle crashes—that’s one person every 50 minutes…” NHSTA.gov
Self driving cars will have fast reaction times – reaction times dramatically faster than humans.
What’s the scenario where a grandma and baby are in the road and the driverless car would be driving at a speed it can’t stop or do a safe evasive maneuver? This scenario will be so rare that the thought experiment is moot.
The real question that this thought experiment should create is: what are we doing to ramp up driverless car sensors to ensure they never hit a baby, let alone a child, grandma, or dog? – The question of protecting a child or dog that runs into the street after a ball is the one that needs to be urgently answered, or if it is answered, publicized widely.