Google DeepMind launches institute to widen the AGI debate

Google and DeepMind researchers launched the DeepMind Institute on Wednesday to advance the conversation around artificial general intelligence (AGI). The institute is led by directors Shane Legg, James Manyika, and Demis Hassabis. The institute aims to surface differing views between Google and the broader global research community regarding AGI. The inaugural collection of essays covers topics including economic policies for managing AGI disruption, preserving model reasoning transparency, and frameworks for evaluating frontier AI models. DeepMind safety researchers Rohin Shah and Anca Dragan argue that the shrinking window of transparency in AI models is not inevitable, urging developers to address safety trade-offs directly. Additionally, Demis Hassabis proposed a U.S.-led frontier AI standards body. Under his framework, developers would submit models for voluntary review before release, with the potential for these tests to become mandatory for future deployments.