Metacognition as a bridge between System 1 and System 2

Determine whether and how metacognitive theories and mechanisms can be used to connect and coordinate fast, intuitive (System 1) and deliberate (System 2) components in embodied intelligent robots, guiding strategy selection and information integration.

Background

Humans employ metacognitive processes to assess uncertainty, select cognitive strategies, and integrate multisensory information, contributing to robust decision-making in non-stationary environments.

The authors suggest that similar metacognitive mechanisms may help embodied agents know when they do not know, choose appropriate inference strategies, and coordinate System 1 and System 2, but whether metacognition can function as an effective bridge remains an open question.

References

One of the interesting aspects of a Dual Process Theory for robots is the fact that metacognition may find a natural place in such a construct: the open question is whether a theory of metacognition can be used to bridge the two systems.

— From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligence  (2110.15245 - Roy et al., 2021) in Section 3.2 (Robots Thinking Fast and Slow: Opportunities and Future Directions)