Google DeepMind CEO Demis Hassabis: The Path To AGI, Deceptive AIs, Building a Virtual Cell

This interview podcast features Alex Kantrowitz interviewing Demis Hassabis, CEO of Google DeepMind, about artificial general intelligence (AGI). The conversation explores the current state of AI development, outlining the capabilities of existing models like Gemini and highlighting the gaps in reasoning, memory, and creative invention needed to reach AGI. Hassabis estimates AGI is 3-5 years away, cautioning against overhyped claims. The discussion further delves into the development of AI agents, world models, and their implications for various fields, including robotics, the future of the web, and scientific discovery (e.g., drug development and material science). The interview concludes with a discussion of the ethical considerations surrounding advanced AI, including the potential for deception and the need for robust safety measures.