Alignment Methods for Advanced AI Systems

Develop methods and evidence-backed procedures to align advanced AI systems with human values in a manner that could be credibly justified to policymakers and standards bodies, thereby enabling safe progression beyond halted development under the proposed international agreement.

Background

The proposal argues for halting progress toward ASI because current alignment research is too immature and unreliable to justify continued advancement. The authors explicitly state that no one currently knows how to align advanced AI systems to human values.

Solving the alignment problem is framed as a precondition for lifting the halt; thus, methods that robustly ensure alignment—and that can be credibly validated—are required to exit the pause safely.