When it comes to autonomous cars, get used to hearing about ‘the trolley problem.’
A thought experiment in ethics, the basis of the problem is as follows:
A runaway train is about to mow down five people tied to the tracks. You, standing in the distance, have the option to pull a lever to make it switch to a second set of tracks. On this set of tracks, there is one person tied down. You have two choices:
- Do nothing, and watch as the train kills the five people on the main track.
- Pull the lever to divert the train to a new track where it’d kill one person.
Which is the most ethical choice?
The problem is the perfect logical aid for attempting to program artificial intelligence met with similar quandaries.
Luckily, this two-year-old seems to have solved it for us. In a rather simplistic solution that I’m almost certain you didn’t think of, the toddler moves the single person to the train track where the other five people are tied down — and proceeds to kill all of them.
A simple solution, but I’m not certain it does much to help programmers design safer cars.