In the search-and-rescue domain, an autonomous system (AS) may be tasked with conveying information or delivering instructions to people who have never encountered that system before. In such cases, willingness to comply demands a “leap of faith”.
This project explores the conditions under which a human subject can be persuaded to take that leap, focusing in particular on three factors: the AS’s relatability, the degree to which its instructions can be verified as correct (guarantee), and it’s level of authority – that is, the guarantor or what the human believes to be the originator of the instructions (e.g., the police, the fire department, the AS’s own calculations).