Published Feb 28, 2024

Demis Hassabis - Scaling, Superhuman AIs, AlphaZero atop LLMs, Rogue Nations Threat

Demis Hassabis, CEO of Google DeepMind, delves into the philosophical and empirical challenges of scaling AI, the imperative for robust AI security amidst open-source challenges, and the groundbreaking multimodal advancements in the Gemini project, highlighting the transformative potential of integrated AI systems.

Episode Highlights

Topics covered

Episode Highlights

Related Episodes

Dario Amodei (Anthropic CEO) - $10 Billion Models, OpenAI, Scaling, & AGI in 2 years
Answers 383 questions
Marc Andreessen - AI, Crypto, 1000 Elon Musks, Regrets, Vulnerabilities, & Managerial Revolution
Answers 383 questions
Ilya Sutskever (OpenAI Chief Scientist) - Building AGI, Alignment, Spies, Microsoft, & Enlightenment
Answers 383 questions
Joseph Carlsmith - Utopia, AI, & Infinite Ethics
Answers 383 questions
Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity
Answers 383 questions
Paul Christiano - Preventing AI Takeover
Answers 383 questions
Carl Shulman (Pt 1) - Intelligence Explosion, Primate Evolution, Robot Doublings, & Alignment
Answers 383 questions
John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI
Answers 383 questions
Shane Legg (DeepMind Founder) - 2028 AGI, New Architectures, Aligning Superhuman Models
Answers 383 questions
Leopold Aschenbrenner - 2027 AGI, China/US Super-Intelligence Race, & The Return of History
Answers 383 questions
David Deutsch - AI, America, Fun, & Bayes
Answers 383 questions
Holden Karnofsky - Transformative AI & Most Important Century
Answers 383 questions
Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality
Answers 383 questions
Joe Carlsmith - Otherness and Control in The Age of AGI
Answers 383 questions
Francois Chollet - LLMs won’t lead to AGI - $1,000,000 Prize to find true solution
Answers 383 questions

Dexa/Dwarkesh Patel

Demis Hassabis - Scaling, Superhuman AIs, AlphaZero atop LLMs, Rogue Nations Threat

Topics covered

Popular Clips

Future AI Timelines

AI Future Insights

Future of Multimodal AI

Grounding AI Learning

AI Model Efficiency

AI Security Insights

AI's Future Capabilities

Scaling Hypothesis Insights

Gemini Development Insights

Intuitive World Models

Future of AI

Gemini Collaboration

Future Multimodal Systems

AI Evolution Insights

Responsible AI Development

Episode Highlights

Scaling AI

AI Security

AI Multimodality

Related Episodes

Dario Amodei (Anthropic CEO) - $10 Billion Models, OpenAI, Scaling, & AGI in 2 years

Marc Andreessen - AI, Crypto, 1000 Elon Musks, Regrets, Vulnerabilities, & Managerial Revolution

Ilya Sutskever (OpenAI Chief Scientist) - Building AGI, Alignment, Spies, Microsoft, & Enlightenment

Joseph Carlsmith - Utopia, AI, & Infinite Ethics

Steve Hsu - Intelligence, Embryo Selection, & The Future of Humanity

Paul Christiano - Preventing AI Takeover

Carl Shulman (Pt 1) - Intelligence Explosion, Primate Evolution, Robot Doublings, & Alignment

John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI

Shane Legg (DeepMind Founder) - 2028 AGI, New Architectures, Aligning Superhuman Models

Leopold Aschenbrenner - 2027 AGI, China/US Super-Intelligence Race, & The Return of History

David Deutsch - AI, America, Fun, & Bayes

Holden Karnofsky - Transformative AI & Most Important Century

Eliezer Yudkowsky - Why AI Will Kill Us, Aligning LLMs, Nature of Intelligence, SciFi, & Rationality

Joe Carlsmith - Otherness and Control in The Age of AGI

Francois Chollet - LLMs won’t lead to AGI - $1,000,000 Prize to find true solution

Demis Hassabis - Scaling, Superhuman AIs, AlphaZero atop LLMs, Rogue Nations Threat

Topics covered

Popular Clips

Episode Highlights

Scaling AIDemis Hassabis, CEO of Google DeepMind, explores the intricate challenges and philosophical questions surrounding AI scaling and the scaling hypothesis. He shares insights into the unexpected capabilities of large models and the empirical nature of AI development.

Scaling AI

AI SecurityDemis Hassabis explores the critical issues of securing AI systems and the challenges of open-source AI. He highlights the need for robust cybersecurity measures and the delicate balance between openness and security.

AI Security

AI MultimodalityThe Gemini project at DeepMind is pushing the boundaries of AI by integrating multimodal capabilities and advanced robotics. Demis Hassabis discusses the transformative potential of these technologies and the collaborative efforts driving their development.

AI Multimodality

Related Episodes