Executive summary: Rethink Priorities is developing a hierarchical Bayesian model to estimate consciousness probabilities in near-future AI systems, aiming to incorporate diverse expert views and provide practical guidance for policy decisions.
Key points:
Model combines three key elements: hierarchical structure (stances → features → indicators), Bayesian probability treatment, and thematic interpretation of evidence.
Designed to be flexible and inclusive of various theoretical perspectives while avoiding commitment to highly specific or controversial assumptions.
Takes a “coarse-grained” approach to features of consciousness rather than requiring precise architectural specifications.
Intended for both policymakers and consciousness researchers, with focus on practical application and ability to update with new evidence.
Improves upon existing probabilistic approaches by incorporating both positive and negative indicators, handling degrees of evidence, and tracking correlations between features.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, andcontact us if you have feedback.
Executive summary: Rethink Priorities is developing a hierarchical Bayesian model to estimate consciousness probabilities in near-future AI systems, aiming to incorporate diverse expert views and provide practical guidance for policy decisions.
Key points:
Model combines three key elements: hierarchical structure (stances → features → indicators), Bayesian probability treatment, and thematic interpretation of evidence.
Designed to be flexible and inclusive of various theoretical perspectives while avoiding commitment to highly specific or controversial assumptions.
Takes a “coarse-grained” approach to features of consciousness rather than requiring precise architectural specifications.
Intended for both policymakers and consciousness researchers, with focus on practical application and ability to update with new evidence.
Improves upon existing probabilistic approaches by incorporating both positive and negative indicators, handling degrees of evidence, and tracking correlations between features.
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.