Page 119 - Artificial Intelligence for the Internet of Everything
P. 119

Trust and Human-Machine Teaming: A Qualitative Study  105


              6.1.7 Communication Richness
              Related to the above dimension, human-agent teams should be capable of
              rich dialogue to convey task and team-based information between each
              other (Chen & Barnes, 2014). Rich communicative and social cue affor-
              dances may make robots more effective when interacting with humans
              (Mutlu, 2011). The key distinction between this dimension and the above
              dimension is that the above dimension discusses nontask-oriented commu-
              nications, which are geared toward team-building. The current dimension
              focuses on the richness of communication in general, which could include
              both task-oriented and nontask communications. Media richness is believed
              to facilitate team effectiveness due to the added social and task-based infor-
              mation that rich media can convey (Hanumantharao & Grabowski, 2006).
              The greater the richness of communication affordances between the human
              and the technology, the greater the likelihood of the human viewing the
              technology as a teammate versus a tool.


              6.1.8 Synchrony
              Effective teams are comprised of team members who have a shared aware-
              ness of the task, the team, and the context. Indeed, shared awareness and,
              more specifically, having synchronized mental models has been shown to
              enhance team effectiveness (Hinds & Mortensen, 2005). Shared mental
              models have also been hypothesized to be important for human-machine
              teams (Ososky et al., 2013). Having synchrony between team members
              allows the team to share a common perception of the team and its capabil-
              ities/limitations, the context, which facilitates joint adaptation, and the task,
              which enables the team members to anticipate the actions of others.
                 In summary, the current paper examines the antecedents of human-
              machine trust and the components of human-machine teaming using the
              Autonomous Agent Teammate-Likeness model (Wynne & Lyons, 2018)
              as a guiding rubric. It was expected that the trust antecedent themes of reli-
              ability, predictability, helping solve a problem, proactively helping, transpar-
              ency (logic, intent, and state), liking, familiarity, and social interaction
              would be expressed in the open-ended rationale for why individuals report
              trust (or distrust) of the technology. It was expected that perceptions of
              agency, benevolence, interdependence, relationship-building, communica-
              tion richness, and synchrony would be associated with more teammate
              (versus tool) perceptions.
   114   115   116   117   118   119   120   121   122   123   124