Game Theory and Autonomous Vehicles
How would you react knowing a car would definitely stop when you walk in front of it?
The autonomous vehicle (AV) technology has been growing extremely fast and a lot of money has been put into researching it. AVs are thought to be the solution to road accidents as they are designed to learn to identify potential crashes. The problem is how would pedestrians adopt knowing that a car will stop when they go ahead of the car. It is highly likely for pedestrians to step right in front of a self-driving car because the car would not injure them. This is definitely not optimal for transportation within a dense city.
Game theory could be applied to analyze the behavior of a pedestrian with a manually driven car and an autonomous car. When there’s a pedestrian and a regular car crossing a road, the strategies that are the best response to each other is for the car to go and the pedestrian to stay to avoid injury. In the situation with a pedestrian and an autonomous car, to go becomes the pedestrian’s dominant response. The autonomous car’s best response to the pedestrian going is to stay. The Nash equilibrium in this situation is for the pedestrian to go and the car to stay. If that is the case, this is great in terms of safety. However, it could incentivize people to always go in front of the car and easily cause a lot more traffic on the road. This leaves the question: would autonomous cars really make the transport system more efficient?