This study presents a **Human-Virtual Agent Coexistence Framework** aimed at understanding the content and structure of the virtual agent literature. The authors argue that “co-existence” aptly describes the broad concept of how humans and virtual agents work together, sometimes autonomously, with different objectives. This extends the definition of human-robot coexistence to include advanced cognitive capabilities, which take virtual agents beyond automation, suggesting they will live and work alongside humans as independent agents.
In developing the framework, the authors consider humans as the default agents and human society as the default setting, focusing on how virtual agents become integrated and consequently change elements of this social world.
The framework reveals four primary dimensions in the virtual agent literature: context, agent, interactions, and applications.
1. **The context dimension** refers to the conditions in which human-virtual agent interactions occur.
1. **The agent dimension** considers the body/mind dichotomy of the virtual agent.
1. **The interaction dimension** explores how the virtual agent interacts with humans at individual, group, and societal levels.
1. **Applications** represent areas where virtual agents are applied in human life.
The authors categorize all topics within these dimensions, including future topics appearing in italics, and identify potential areas for future research in relation to the dimensions of the framework.
=> https://www.sciencedirect.com/science/article/abs/pii/S0040162523003293