Hassabis Backs Amodei's Call to Slow Down Advanced AI Development

Google DeepMind's CEO Demis Hassabis has expressed support for Anthropic's Dario Amodei's proposal to slow down the development of advanced AI models.

Hassabis stated that while details need working through, 'the direction is correct' for addressing a 'critical moment' in AI evolution. This comes after Amodei called for pacing the frontier and slowing the development of highly capable models due to concerns about losing control over these systems.

As part of this push for greater oversight, Anthropic has proposed a three-part plan including independent third-party evaluations, industry-wide coordination on safety standards, and limits on unchecked AI development. Hassabis also mentioned Google's recent efforts towards establishing common safety standards for frontier AI through an industry-wide standards body.