It’s been just over a year since the first pedestrian was hit and killed by a self-driving car in Tempe, Ariz. Since then, we’ve learned a lot about the algorithms that drive autonomous vehicles.
We’ve learned that self-driving cars are probably better than human drivers at maintaining safe speeds and distances on highways, and that researchers are working on teaching self-driving cars to make moral choices. We’ve also learned about serious problems, such as algorithms that are better at detecting light-skinned pedestrians and therefore are more likely to hit pedestrians with darker skin tones.
Nicholas Evans, an assistant professor of philosophy at the University of Massachusetts Lowell, argues that the artificial intelligence community has not done enough to correct the biases that are currently embedded in their systems.
“And the reason we know they are not is because this is not a new problem, and it's not a problem that's exclusive to autonomous vehicles,” Evans told Living Lab Radio on Tuesday. “The problem with race in algorithmic bias is very long standing.”
Evans is working with other philosophers and an engineer to write algorithms using ethical theories.
One area of his work has to do with the distribution of very small risks over millions of miles driven.
In one example, an autonomous vehicle is passing a vehicle transport truck on the highway.
“If [the autonomous car] moves around in its lane in order to keep its occupants safe, is it applying risk to someone in a third lane by getting too close to that driver?” he said.
“These aren't necessarily risks that always involve death, but they are risks that, when you add them up over the billions and even trillions of miles that Americans drive every year, could substantially affect not only the number of people who have serious injuries or die in car crashes, but also the kinds of people who can be in serious car crashes and die,” he said.
Autonomous cars will probably be expensive when they first enter the market, so only wealthy people will own them, he said. Also, poorer and less educated people already die in car crashes more often than rich and educated people, according to Evans. He is concerned that the advent of autonomous cars will make existing disparities worse.
“I think that there's a lot of promise in autonomous vehicles, but I think that we've got a long way to go to learn the right lessons to make this a safe, accessible, and fair technology,” Evans said.
