LW - AGI Safety and Alignment at Google DeepMind: A Summary of Recent Work by Rohin Shah | The Nonlinear Library | Podwise