As a driver, I still make a number of arbitrary decisions that avoid having another incident happen. There are algorithms in place for a number of these scenarios, but let me throw one scenario at you. Let's say I'm driving down a two-lane road on a mountainside. There's a semi truck coming at me, and a mom and her baby walking a stroller near there. I'm in an autonomous vehicle. What does the autonomous vehicle do? Does it make the decision to take me off the cliff? Do we hit the semi truck? Do we run into the mom and the baby? Now, do I get into a vehicle thinking that it would probably mean taking me off the cliff? Do I get into a vehicle thinking that there's a risk that it's going to take me off the cliff? I think these are the issues, the moral and ethical issues, that are facing the industry. I'm curious as to how we justify every algorithm.
On March 26th, 2018. See this statement in context.