Humans do not "play fair" with social robots: Unequal task sharing in an attentional task

Area of Science:

• Human-Computer Interaction
• Cognitive Science
• Robotics

Background:

• Artificial agents support daily tasks, but offloading attentionally demanding tasks is under-researched.
• Humans equally split attention with other humans but not algorithms.
• Social robots with human-like characteristics raise questions about human offloading decisions.

Purpose of the Study:

• Investigate how humans offload attentionally demanding tasks to social robots.
• Determine if social robots are treated as human or algorithmic partners in offloading.
• Examine the impact of known versus unknown robot accuracy on offloading.

Main Methods:

• Participants performed a multiple object tracking (MOT) task, offloading demands to a social robot (MAKI).
• Two experiments varied whether the robot's task accuracy was known or unknown.
• Offloading behavior and performance gains were measured.

Main Results:

• Participants significantly offloaded attentional demands to the robot, improving their own performance.
• Robot accuracy information did not impact offloading or performance gains.
• Participants did not equally split attention with the robot, unlike with human partners.

Conclusions:

• Social robots are treated similarly to algorithms, not human partners, for attention offloading.
• Human-like characteristics did not shift offloading behavior towards human interaction patterns.
• This suggests cognitive processing of social robots differs from human interaction partners.