Advice Taking from Humans and Machines: An fMRI and Effective Connectivity Study.

Area of Science:

• Neuroscience
• Cognitive Science
• Human-Computer Interaction

Background:

• Technological advancements enable advice from both human and machine sources.
• Understanding how individuals utilize advice and its neural underpinnings is crucial.
• Investigating the differences in advice utilization between human and machine agents is needed.

Purpose of the Study:

• To investigate the neural basis of advice utilization from human versus machine agents.
• To explore the effective connectivity within the brain networks involved in processing advice.
• To examine how agent type (human vs. machine) and advice quality affect utilization.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to measure brain activity.
• Multivariate Granger causality analysis was employed to assess effective connectivity.
• Participants performed an X-ray luggage-screening task, accepting/rejecting advice from unreliable agents.

Main Results:

• Unreliable advice generally decreased task performance.
• Participants showed greater depreciation of advice utilization with bad advice from a human agent compared to a machine agent.
• Brain regions associated with trait evaluation (precuneus, posterior cingulate cortex, temporoparietal junction) and interoception (posterior insula) were engaged.
• The right posterior insula and left precuneus acted as key drivers in the advice utilization network.

Conclusions:

• Differences in advice utilization may stem from reevaluating agent credibility.
• Neural networks involved in evaluating personal characteristics and interoception are critical for advice utilization.
• Findings have significant societal implications given increasing human-machine interaction.