Model evaluation for extreme risks Article Swipe

PDF

Related Concepts

Offensive Software deployment Risk analysis (engineering) Harm Computer security Computer science Business Process management Engineering Operations research Political science Operating system Law

Toby Shevlane , Sebastian Farquhar , Ben Garfinkel , Mary Phuong , Jess Whittlestone , Jade Leung , Daniel Kokotajlo , Nahema Marchal , Markus Anderljung , Noam Kolt , L. Lawrence Ho , Divya Siddarth , Shahar Avin , Will Hawkins , Been Kim , Iason Gabriel , Vijay Bolina , Jack A. Clark , Yoshua Bengio , Paul Christiano , Allan Dafoe ·

YOU? · · 2023 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2305.15324 · OA: W4378474292

Current approaches to building general-purpose AI systems tend to produce systems with both beneficial and harmful capabilities. Further progress in AI development could lead to capabilities that pose extreme risks, such as offensive cyber capabilities or strong manipulation skills. We explain why model evaluation is critical for addressing extreme risks. Developers must be able to identify dangerous capabilities (through "dangerous capability evaluations") and the propensity of models to apply their capabilities for harm (through "alignment evaluations"). These evaluations will become critical for keeping policymakers and other stakeholders informed, and for making responsible decisions about model training, deployment, and security.