While a foundation exists to consider training for the complex, evolving nature of human-agent teaming envisioned in the future, the literature on training has placed very little focus on this precise need. Education and training fields are beginning to focus on the use of individualized, adaptive technologies; however, emphasis is not placed on considering these technologies as teammates and how that might change the content and delivery of training. Team science is advancing training for enhanced teamwork, but the idea of humans and agents as team members does not seem to be a critical element this work. Many research questions must be addressed to begin to prepare for this future. Some examples include:
1. As we move forward, we must understand the human and agent competencies critical for human-agent teamwork. What are the core competencies unique to: (a) human-agent teams, (b) humans working in human-agent teams, and (c) autonomous agents working in human-agent teams? After identifying and defining these competencies, we must explore how to best develop these critical teamwork competencies, through individual and/or team-focused training. How do we synchronize training and development amongst diverse members, including humans and autonomous teammates?
[wpforms id=”1363″ title=”false” description=”false”]2. Envisioning the future, we expect agents capable of learning and adapting at high speeds, and humans that must fluidly adapt with these team members. How do we prepare humans to work with evolving agents? How do we prepare agents to understand and evolve symbiotically with humans? What technologies can be developed to enhance this mutual adaptation in human-agent teams?
[wpforms id=”1364″ title=”false” description=”false”]3. Human-agent teams are expected to be characterized by diversity, and rotating and evolving team membership. What are the critical features of individualized, adaptive training technologies to enhance teamwork in these diverse and evolving teams? What future capabilities are critical for quickly bringing together diverse teams of humans and agents to perform effectively as a group? In what ways can embedding intelligent agent trainers in teams reduce the efficiency costs and increase the productivity benefits associated with diverse and dynamic team membership?
[wpforms id=”1365″ title=”false” description=”false”]Finally, the future we are considering opens potential for novel training capabilities to continually enhance the collective intelligence of the group. Can we use principles of individualized instruction and human-agent teamwork to develop evolving systems of humans and agents with ever-increasing intelligence and capabilities, capable of more complex performance over time?
[wpforms id=”1366″ title=”false” description=”false”]Call for Comments:
- What are other scientific questions critical to preparing for future human-agent teaming?
- Other related comments
Comments
1) Although this section focuses on training, training goes hand-in-hand with other industrial/organizational psychology and human resource issues/factors, such as performance evaluation and selection. In my opinion, individualized training will first require early assessment on a range of characteristics (e.g., different cognitive aptitudes on the part of the human; language proficiency) in order to know how the training should be adapted for that particular human. So, for individualized training to succeed, one would first need an assessment of the entire workforce (human and agent) for those characteristics of interest. This assessment shouldn’t be static: people can learn new languages, and while some people may have some cognitive aptitudes different from others (e.g., associative creativity), training can hone or improve those at least a little as well.
2) What is the modality of the training, either for individual humans, human-agent/tech dyads, or whole teams of humans and agents/tech? Meaning, is it digital? Will it have visual, tactile, auditory elements? Would it require neuropsychological stimulation (e.g., there is research on how vagal nerve stimulation may help with language learning). Training will need to be feasible in the field/in theater/in country as well as before deployment. Would it require complex simulation environments?
3) As with the technology itself, care should be taken that the training can work for a variety of humans: not just those with different cognitive aptitudes, but those with different capabilities and disabilities (e.g., color blindness).
Thank you for a very interesting paper, and thank you for enabling my involvement!
Thanks for commenting!