The Interdependence of Shared Context: Autonomous Human-Machine Teams

Abstract

Continuing to report on our work-in-progress with our research on interdependence, the key phenomenon in social interaction, we have proposed a mathematical model of interdependence for human-machine teams to develop optimum and efficient performance. Humans will have difficulty with their understanding of this model since it directly involves them; since it provides no grand sketch of meaning and even finds the notion of meaning from an engineering perspective meaningless; and since its results are counterintuitive but highly predictive. To review the progress of our research until now, in our first major study, we found no relationship between education and the performance of physical skills (namely, the book knowledge and air-to-air combat skills of USAF pilots); however, we also found a significant relationship between training and air-to-air combat skills of these same pilots. In our second study, we found that excess personnel in work teams (redundancy) decreased interdependence and performance in the workplace for top global oil-firm teams, replicated in our third study with leading military teams. In our fourth study, we found that intelligence as determined by the average levels of academic schooling in a nation was significantly related to the performance of its teams, as determined by a nation’s patent productivity. Interestingly, the apparently contradictory findings of our last and first study on education and training appear to be orthogonal, a perfect opening to discuss our new work already underway on a social uncertainty principle between orthogonal operators.