Carl Westin

174 Discussion and recommendations pathologists in diagnosing cancer (e.g., C-Patch system), and IBM’s Watson com- puter competing and wining against humans in Jeopardy! 290 By means of methods such as machine learning, natural language processing, and computer vision, the de- velopment of artificially intelligent machines allows for automating complex tasks and processes that previously were considered “too open.” The currently most dramatic change is perhaps the development of fully au- tonomous, “self-driving,” cars. Fully autonomous cars is a good example of a do- main in which strategic conformance is particularly relevant. The experience of the human passenger (previously driver) will largely depend on the comfort from trav- eling in the car. A better experience can be expected provided the autonomous car “drives” in a way conformal to the passenger’s preferences (e.g., in terms of accel- erations, braking, turning, and distance keeping). Another example may be a fully autonomous ATC CD&R system supervised by a controller. Both acceptance and trust in the system are likely to be higher if the system solves conflicts according to the controller’s preferences, thus resulting in fewer interventions. Taken together, the introduction of fully autonomous systems may not lessen the relevance of strate- gic conformal automation, provided the human still has to interact with the system from time to time.