A model for investigating the effects of machine autonomy on human behavior

As autonomous machines become more pervasive, situations will arise when human decision-makers will receive advice from both machines and other humans. When these instructions conflict, a new social situation is defined for which we have little precedent. The authors propose a model for investigating these situations. The model synthesizes research from several different fields, including machine autonomy, affect, initial trust, individual differences, and training. The model is explained, and a set of propositions is described. The model is used to analyze the case of an air collision in which machines and humans provided conflicting advice. The model is also applied to situations in which unmanned aerial vehicles and piloted aircraft seek to avoid collisions with each other. Ways of testing the model through human subject experiments are discussed.