Moral Dilemma for Google's Driverless Vehicle Control Software

Moral dilemmas involving driverless vehicles are best solved by programming the intelligent agent governance to simulate an ethical human driver. In this fashion, the agent reuses the existing human-applicable laws and conventions.

To be sure, laws will have to be revised to accomodate certain super-human powers of vehicle-controlling intelligent agents. For example, I suppose that such agents will be much more aware of various environmental aspects due to their mesh-networking with other vehicles, roadside features, etc.

Suppose that blind pedestrians have smartphones that mesh-network with nearby driverless cars. Existing driving laws could then be revised to require driverless cars to yield right of way, even if the blind pedestrian is not holding a cane, or is not otherwise visible.