For decades, people have imagined wildly different, futuristic-looking transportation vehicles. Although future cars will be smarter and drones will be available widely, it is unlikely that by 2030 we will have widely adopted transportation vehicles that look and function differently than the ones we have today. Our Study Panel doesn’t expect drones that can fly, swim, and drive, or flying quadcoptors to become a common means of transportation in this time horizon (although prototypes exist today).
We do expect humans to become partners to self-driving cars and drones in their training, execution, and evaluation. This partnering will happen both when humans are co-located with machines and also virtually. We predict advances in algorithms to facilitate machine learning from human input. We also expect models and algorithms for modeling of human attention, and to support communication and coordination between humans and machine. This is an integral part of the development of future vehicles.