Episode Description

AI Deep Dive is a show from The Information's TITV that gets to the bottom of the hardest technical problems in AI. We explore the models, the research and the people building them. Hosted by AI reporter, Rocket Drew.

Luma AI Co-Founder and CEO Amit Jain joins Rocket to tackle one of AI’s biggest questions: What would it take for a model to actually understand the physical world?

They also explore how intelligence in models for language, vision and robotics connect.

Related articles:

Robots Are in Their GPT-2 Era: https://www.theinformation.com/newsletters/ai-agenda/robots-gpt-2-era

China’s New AI Gold Rush: World Models: https://www.theinformation.com/articles/chinas-new-ai-gold-rush-world-models

Subscribe:

YouTube: https://www.youtube.com/@theinformation

The Information: https://www.theinformation.com/subscribe_h

Sign up for the AI Agenda newsletter: https://www.theinformation.com/features/ai-agenda

Follow us:

X: https://x.com/theinformation

IG: https://www.instagram.com/theinformation/

TikTok: https://www.tiktok.com/@titv.theinformation

LinkedIn: https://www.linkedin.com/company/theinformation/

Chapters:

00:00 – Intro

02:24 – What Is a World Model?

07:19 – Teaching AI the Physical World

14:05 – Why Robots Still Struggle

19:07 – Are Video Models World Models?

25:17 – Can AI Understand Physics?

31:01 – Reinforcement Learning in the Real World

37:29 – AI’s Physical Data Problem

42:52 – Building Physical Intelligence

48:28 – The Bitter Lesson for Robotics

57:39 – Is There Just One Kind of Intelligence?

1:07:38 – What Actually Counts as a World Model?

1:21:26 – Could LLMs Become World Models?

1:26:07 – Outro