Artwork

Content provided by First Principles and Christian Keil. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by First Principles and Christian Keil or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://ppacc.player.fm/legal.
Player FM - Podcast App
Go offline with the Player FM app!

#12: Deep Prasad - How to Build a Physics-Fluent AI System

1:06:44
 
Share
 

Manage episode 420978276 series 3554927
Content provided by First Principles and Christian Keil. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by First Principles and Christian Keil or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://ppacc.player.fm/legal.

Brought to you by NordVPN. EXCLUSIVE NordVPN Deal ➼ https://nordvpn.com/firstprinciples. Try it risk-free now with a 30-day money-back guarantee!

======

Episode 12: Discovering new materials is how we unlock new branches of the tech tree. But there is no way for the world’s most powerful AI tools to help us do that discovery – until now.

======

(00:00) - Intro

(01:08) - GenMat's Goal: Building an AI System Smarter Than Humans at Physics

(02:17) - Why It's Hard for AI to Understand Physics Like Humans Do

(05:39) - How GenMat Models Atomic Level Activity to Optimize Materials

(09:35) - The Trial and Error Process of Traditional Materials Discovery

(11:28) - Materials Science 101: What Makes a Material and Its Key Properties

(15:00) - Perovskites: A Promising and Complex Material Family

(17:37) - How GenMat Helps Labs Discover and Optimize New Materials

(19:43) - The Challenges of Data, Representations and Feedback in Physics AI

(23:18) - Why GenMat Chose a Broad vs Narrow Focus Across Materials Domains

(25:32) - Expanding Into Geophysics to Find Critical Materials Using Satellites

(28:48) - The Experience of Launching GenMat's First Satellite with SpaceX

(30:58) - Overlap Between GenMat's Materials and Geophysics AI Models

(32:42) - GenMat's Progress: Team Size, Funding, Patents, Customers, Results

(35:46) - The Future 'Yug-Cyberia' Civilization Enabled by Advanced Physics AI

(38:54) - How Materials Breakthroughs Can Accelerate Mars and Space Exploration

(42:10) - The Difficulty of Communicating Complex Technical Work to Others

(44:48) - Breaking Down GenMat's AI Techniques: CDVAEs Explained

(48:55) - Why GenMat Uses Classical Deep Learning vs Quantum Computing Today

(51:37) - The Origins of GenMat: From Quantum Deep Learning to Classical AI

(54:02) - How GenMat Gets People Excited About the Potential of Materials

(56:11) - Rapid Fire Questions: Focus vs Breadth, Challenges, Surprises

(59:40) - Mapping Out a Future with Intelligent Matter and Interstellar Travel

(01:03:50) - The Meaning and Origins of the Term 'Yug-Cyberia'

(01:05:15) - Conclusion: Advanced Physics AI Can Get Us to Mars and Beyond

======

Links:

======

Production and marketing by The Deep View (https://thedeepview.co). For inquiries about sponsoring this podcast, email [email protected]

======

Checkout the video version here → http://tinyurl.com/4fh497n9

🔔 Follow to stay updated with new uploads

  continue reading

20 episodes

Artwork
iconShare
 
Manage episode 420978276 series 3554927
Content provided by First Principles and Christian Keil. All podcast content including episodes, graphics, and podcast descriptions are uploaded and provided directly by First Principles and Christian Keil or their podcast platform partner. If you believe someone is using your copyrighted work without your permission, you can follow the process outlined here https://ppacc.player.fm/legal.

Brought to you by NordVPN. EXCLUSIVE NordVPN Deal ➼ https://nordvpn.com/firstprinciples. Try it risk-free now with a 30-day money-back guarantee!

======

Episode 12: Discovering new materials is how we unlock new branches of the tech tree. But there is no way for the world’s most powerful AI tools to help us do that discovery – until now.

======

(00:00) - Intro

(01:08) - GenMat's Goal: Building an AI System Smarter Than Humans at Physics

(02:17) - Why It's Hard for AI to Understand Physics Like Humans Do

(05:39) - How GenMat Models Atomic Level Activity to Optimize Materials

(09:35) - The Trial and Error Process of Traditional Materials Discovery

(11:28) - Materials Science 101: What Makes a Material and Its Key Properties

(15:00) - Perovskites: A Promising and Complex Material Family

(17:37) - How GenMat Helps Labs Discover and Optimize New Materials

(19:43) - The Challenges of Data, Representations and Feedback in Physics AI

(23:18) - Why GenMat Chose a Broad vs Narrow Focus Across Materials Domains

(25:32) - Expanding Into Geophysics to Find Critical Materials Using Satellites

(28:48) - The Experience of Launching GenMat's First Satellite with SpaceX

(30:58) - Overlap Between GenMat's Materials and Geophysics AI Models

(32:42) - GenMat's Progress: Team Size, Funding, Patents, Customers, Results

(35:46) - The Future 'Yug-Cyberia' Civilization Enabled by Advanced Physics AI

(38:54) - How Materials Breakthroughs Can Accelerate Mars and Space Exploration

(42:10) - The Difficulty of Communicating Complex Technical Work to Others

(44:48) - Breaking Down GenMat's AI Techniques: CDVAEs Explained

(48:55) - Why GenMat Uses Classical Deep Learning vs Quantum Computing Today

(51:37) - The Origins of GenMat: From Quantum Deep Learning to Classical AI

(54:02) - How GenMat Gets People Excited About the Potential of Materials

(56:11) - Rapid Fire Questions: Focus vs Breadth, Challenges, Surprises

(59:40) - Mapping Out a Future with Intelligent Matter and Interstellar Travel

(01:03:50) - The Meaning and Origins of the Term 'Yug-Cyberia'

(01:05:15) - Conclusion: Advanced Physics AI Can Get Us to Mars and Beyond

======

Links:

======

Production and marketing by The Deep View (https://thedeepview.co). For inquiries about sponsoring this podcast, email [email protected]

======

Checkout the video version here → http://tinyurl.com/4fh497n9

🔔 Follow to stay updated with new uploads

  continue reading

20 episodes

All episodes

×
 
Loading …

Welcome to Player FM!

Player FM is scanning the web for high-quality podcasts for you to enjoy right now. It's the best podcast app and works on Android, iPhone, and the web. Signup to sync subscriptions across devices.

 

Quick Reference Guide

Copyright 2025 | Privacy Policy | Terms of Service | | Copyright
Listen to this show while you explore
Play