Modeling and calibrating trusted interactions between a human agent and AI system- A combined quantum-cognitive and EEG based approach

Abstract

Trust is vital for effective synchronised action in human-AI teaming just as it is in high performing sports teams, or a finely tuned human canine unit. Whilst humans, and even trained animals, have innate abilities which allow them to tune their behaviour according to a situation and collaborating partner, thereby fostering trust, machines have not yet satisfactorily acquired this capability. In order to progress towards this goal, thinking about the trustor and trustee separately must be set aside for models that aim to integrate trustor-trustee actions and decisions in an inseparable whole. The objective of this proposal is to model the gap between a human agent s expectation of an AI system and their perceived trust while interacting with the system. This will be driven by novel research into interactive trust, stemming from theories within the field of quantum cognition, with the aim of producing innovative theory-backed approaches. The outcome of this project is an empirically tested theoretical model for the calibration of dynamically changing expectations of human and AI system in order to promote trusted interactions.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 29, 2024
Source ID
FA95502310258

Entities

People

  • Peter David Bruza

Organizations

  • Air Force Office of Scientific Research
  • Queensland University of Technology
  • United States Air Force

Tags

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Systems Analysis and Design
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • Quantum Computing