Google DeepMind launches the DeepMind Institute to debate AGI

Google DeepMind Launches DeepMind Institute to Focus on AGI Debates

Google DeepMind has established the DeepMind Institute—a platform dedicated to essays on Artificial General Intelligence (AGI) led by Shane Legg, Demis Hassabis, and James Manyika. Legg, in an interview with the Financial Times, acknowledged the need for caution, stating that Dario Amodei's suggestion to slow AI releases is worth considering, and that it's too early to declare AGI's arrival.

About the Institute

The institute aims to serve as a hub for research and discourse on AGI, with its directors including Shane Legg (Chief AGI Scientist at Google DeepMind), Demis Hassabis (co-founder of DeepMind and Alphabet's chief scientist), and James Manyika (Google's president of research, labs, technology, and society).

"This is a critical moment to ensure we build AGI safely and its benefits to society far outweigh any risks," the directors stated in their launch essay.

Key Takeaways from the Launch Essay

  • Definition of AGI: The essay defines AGI as a system that possesses all cognitive capabilities akin to the human brain, emphasizing that current AI systems still struggle with basic tasks.

  • Concerns: It highlights cybersecurity and biorisks as immediate concerns, along with the prospect of loss of control in future self-improving AI systems.

  • Slowing Down AI Releases: Legg expressed support for Anthropic CEO Dario Amodei's call to slow down but not pause the release of frontier models, citing it as an "interesting directionally" and "worth considering."

  • AGI Achievements: While Legg maintains that it's too early to declare AGI's arrival, Nvidia's Jensen Huang and OpenAI have recently made AGI claims.

Initial Essays on the Platform

In addition to the launch essay, the DeepMind Institute features four initial pieces:

  • "The case for reasoning transparency" by Rohin Shah and Anca Dragan
  • "Economic policy for AGI" by Julian Jacobs and Alex Imas
  • "Principles for a new utopianism" by Stephen Cave
  • "A framework for frontier AI and the dawning of a new age" by Demis Hassabis

Leave a Reply