Unsolved Problems in ML Safety

TL;DR

This paper outlines a comprehensive roadmap for ML safety, emphasizing four key research problems—robustness, monitoring, alignment, and systemic safety—to address emerging challenges in deploying large-scale ML systems securely.

Contribution

It refines and clarifies core safety challenges in ML, proposing specific research problems and directions to advance the field's understanding and solutions.

Findings

Identifies four critical safety problems in ML.

Provides concrete research directions for each problem.

Highlights importance of safety in high-stakes ML deployment.

Abstract

Machine learning (ML) systems are rapidly increasing in size, are acquiring new capabilities, and are increasingly deployed in high-stakes settings. As with other powerful technologies, safety for ML should be a leading research priority. In response to emerging safety challenges in ML, such as those introduced by recent large-scale models, we provide a new roadmap for ML Safety and refine the technical problems that the field needs to address. We present four problems ready for research, namely withstanding hazards ("Robustness"), identifying hazards ("Monitoring"), reducing inherent model hazards ("Alignment"), and reducing systemic hazards ("Systemic Safety"). Throughout, we clarify each problem's motivation and provide concrete research directions.