Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

This video explains how diffusion models are overtaking Large Language Models for generation tasks like: 1. Code Generation 2. Image Generation 3. Video Generation 00:00 Agenda 00:20 How are they different from LLMs? 05:09 Internal Mechanism 10:09 How are vectors generated? 12:08 Conclusion 13:02 Opinion Piece AI Engineering Course:... show more

Gaurav Sen

71,401 subscribers

27,927 views • 8 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Diffusion is the foundational ML framework behind state-of-the-art AI image and video generation, including Sora, Midjourney and Google Veo. In this episode of Decoded, Ankit Gupta sits down with Francois Chaubard to discuss how diffusion works, walk through a code sample, and explain why everyone training models should understand it. 00:00 Intro 00:33 What is diffusion? 02:50 What are applications of diffusion today? 04:06 Key innovations 07:01 Code examples 19:25 The "squint test" 22:27 Other areas diffusion is widely accessible 24:49 Outro

Diffusion is the foundational ML framework behind state-of-the-art AI image and video generation, including Sora, Midjourney and Google Veo. In this episode of Decoded, Ankit Gupta sits down with Francois Chaubard to discuss how diffusion works, walk through a code sample, and explain why everyone training models should understand it. 00:00 Intro 00:33 What is diffusion? 02:50 What are applications of diffusion today? 04:06 Key innovations 07:01 Code examples 19:25 The "squint test" 22:27 Other areas diffusion is widely accessible 24:49 Outro

Y Combinator

74,604 views • 5 months ago

Wow, diffusion models (used in AI image generation) are also game engines - a type of world simulation. By predicting the next frame of the classic shooter DOOM, you get a playable game at 20 fps without any underlying real game engine. This video is from the diffusion model.

Wow, diffusion models (used in AI image generation) are also game engines - a type of world simulation. By predicting the next frame of the classic shooter DOOM, you get a playable game at 20 fps without any underlying real game engine. This video is from the diffusion model.

Ethan Mollick

1,768,873 views • 1 year ago

In this video, we break down the Model Context Protocol — a massively underrated concept in AI that's quietly redefining what large language models can do. While most people are distracted by the hype, MCP is solving a core problem: LLMs that can perform actions. We dive into: 1. The limitations of current AI systems 2. What makes MCP different 3. How MCP can reduce engineering overhead 4. Real-world scenarios where MCP shines 00:00 What is MCP? 01:48 Usecase - SEO 03:50 Usecase - RAG 04:50 Usecase - Apps 06:30 Conclusion 07:00 The future? #MCP #AI #LLM

In this video, we break down the Model Context Protocol — a massively underrated concept in AI that's quietly redefining what large language models can do. While most people are distracted by the hype, MCP is solving a core problem: LLMs that can perform actions. We dive into: 1. The limitations of current AI systems 2. What makes MCP different 3. How MCP can reduce engineering overhead 4. Real-world scenarios where MCP shines 00:00 What is MCP? 01:48 Usecase - SEO 03:50 Usecase - RAG 04:50 Usecase - Apps 06:30 Conclusion 07:00 The future? #MCP #AI #LLM

Gaurav Sen

41,117 views • 1 year ago

Engineers need to communicate effectively when building AI Systems. These terms will help you use a shared vocabulary. This is useful when discussing concepts, reading papers, or collaborating with teammates. Listed in order. 00:00 Agenda 00:28 1. Large Language Model 01:28 2. Tokenization 02:53 3. Vectorization 04:15 4. Attention 07:22 5. Self-Supervised Learning 12:07 6. Transformer 14:32 7. Fine-tuning 17:05 8. Few-shot Prompting 18:11 9. Retrieval Augmented Generation 20:33 10. Vector Database 23:03 11. Model Context Protocol 25:43 12. Context Engineering 28:17 13. Agents 29:19 14. Reinforcement Learning 34:42 15. Chain of Thought 35:55 16. Reasoning Models 36:36 17. Multi-modal Models 38:21 18. Small Language Models 40:24 19. Distillation 41:47 20. Quantization If you are a software engineer looking to transition to AI, click the link below. AI Engineering Course: #AI #SoftwareEngineering #Agents

Engineers need to communicate effectively when building AI Systems. These terms will help you use a shared vocabulary. This is useful when discussing concepts, reading papers, or collaborating with teammates. Listed in order. 00:00 Agenda 00:28 1. Large Language Model 01:28 2. Tokenization 02:53 3. Vectorization 04:15 4. Attention 07:22 5. Self-Supervised Learning 12:07 6. Transformer 14:32 7. Fine-tuning 17:05 8. Few-shot Prompting 18:11 9. Retrieval Augmented Generation 20:33 10. Vector Database 23:03 11. Model Context Protocol 25:43 12. Context Engineering 28:17 13. Agents 29:19 14. Reinforcement Learning 34:42 15. Chain of Thought 35:55 16. Reasoning Models 36:36 17. Multi-modal Models 38:21 18. Small Language Models 40:24 19. Distillation 41:47 20. Quantization If you are a software engineer looking to transition to AI, click the link below. AI Engineering Course: #AI #SoftwareEngineering #Agents

Gaurav Sen

80,115 views • 10 months ago

How Sui Plans to Beat Hyperliquid with evan.sui Timestamps: 00:00 Intro 01:32 Is Blockchain Just Finance? 02:40 Sui's Vertical Build Strategy 05:00 Who Benefits from Tokenization? 07:21 The Verticalization Shift 09:32 AI Meets Blockchain 11:10 Do We Need Billion TPS? 13:10 Regulatory Clarity Incoming 15:05 Competing with Hyper Liquid 20:02 L1 Fee Models Are Dead 21:32 Leaving Facebook's Libra 24:18 Institutional Appetite for Sui

How Sui Plans to Beat Hyperliquid with evan.sui Timestamps: 00:00 Intro 01:32 Is Blockchain Just Finance? 02:40 Sui's Vertical Build Strategy 05:00 Who Benefits from Tokenization? 07:21 The Verticalization Shift 09:32 AI Meets Blockchain 11:10 Do We Need Billion TPS? 13:10 Regulatory Clarity Incoming 15:05 Competing with Hyper Liquid 20:02 L1 Fee Models Are Dead 21:32 Leaving Facebook's Libra 24:18 Institutional Appetite for Sui

The Rollup

59,517 views • 4 months ago

., Co-founder of fal, on where AI image & video generation stands today: “Image and video models haven’t really generalized yet. There’s still a lot of model chaining and workflow generation." "Models are getting bigger and better, but unlike LLMs where you can write a prompt and expect it to handle various tasks, that doesn’t happen for media yet.” He notes that while chaining multiple models is still the norm, the opportunity lies in building hybrid systems that stitch workflows together reliably until true generalization arrives.

., Co-founder of fal, on where AI image & video generation stands today: “Image and video models haven’t really generalized yet. There’s still a lot of model chaining and workflow generation." "Models are getting bigger and better, but unlike LLMs where you can write a prompt and expect it to handle various tasks, that doesn’t happen for media yet.” He notes that while chaining multiple models is still the norm, the opportunity lies in building hybrid systems that stitch workflows together reliably until true generalization arrives.

TBPN

36,675 views • 11 months ago

Core Memory podcast with Suzanne Gildert on the hunt for AI consciousness in the quantum realm. The theory being that LLMs are too constrained to really get to robots and AIs that are really like us 00:00:00 – Consciousness: The First AI Training Model? 00:02:03 – Exploring the Quantum Mind with AI 00:04:42 – Will Robots Ever Have Rights? 00:07:54 – The AI Paradox: Fear vs. Progress 00:13:20 – Can Language Models Teach Robots? 00:17:29 – Why Today’s AI Feels Like a Zombie 00:23:47 – Quantum Computing Meets Artificial Consciousness 00:29:50 – Free Will, Entanglement, and Decision Making 00:38:21 – Teaching a Robot to Think, Not Just React 00:52:51 – Will Conscious AI Power Space Exploration?

Core Memory podcast with Suzanne Gildert on the hunt for AI consciousness in the quantum realm. The theory being that LLMs are too constrained to really get to robots and AIs that are really like us 00:00:00 – Consciousness: The First AI Training Model? 00:02:03 – Exploring the Quantum Mind with AI 00:04:42 – Will Robots Ever Have Rights? 00:07:54 – The AI Paradox: Fear vs. Progress 00:13:20 – Can Language Models Teach Robots? 00:17:29 – Why Today’s AI Feels Like a Zombie 00:23:47 – Quantum Computing Meets Artificial Consciousness 00:29:50 – Free Will, Entanglement, and Decision Making 00:38:21 – Teaching a Robot to Think, Not Just React 00:52:51 – Will Conscious AI Power Space Exploration?

Ashlee Vance

17,189 views • 1 year ago

SPECIAL: How Erik Voorhees Built the Leading Private, Uncensored AI Venice Timestamps: 00:00 Intro & ShapeShift Origin Story 05:25 Did Crypto Play Out as Expected? 06:25 AI vs. Crypto: Are They Similar? 07:23 What Is Venice AI? 09:38 From ShapeShift to Venice 10:31 Censorship in AI & the Open Source Alternative 11:38 Hibachi, infiniFi Ads 12:00 Are AI Labs Too Censored? 13:24 Relay Ad 13:41 Geopolitics: Chinese Open Source vs. US Closed Models 14:15 SBF, the Anthropic Investment & Crypto Scammers 16:26 NearCon, Illia & the AI/Crypto Convergence

SPECIAL: How Erik Voorhees Built the Leading Private, Uncensored AI Venice Timestamps: 00:00 Intro & ShapeShift Origin Story 05:25 Did Crypto Play Out as Expected? 06:25 AI vs. Crypto: Are They Similar? 07:23 What Is Venice AI? 09:38 From ShapeShift to Venice 10:31 Censorship in AI & the Open Source Alternative 11:38 Hibachi, infiniFi Ads 12:00 Are AI Labs Too Censored? 13:24 Relay Ad 13:41 Geopolitics: Chinese Open Source vs. US Closed Models 14:15 SBF, the Anthropic Investment & Crypto Scammers 16:26 NearCon, Illia & the AI/Crypto Convergence

The Rollup

47,718 views • 4 months ago

New EP for my consumer ai series featuring MoemateAI $mates 00:00 introduction to moemate and its unique features 01:11 exploring moemate agent sandbox and customization 02:21 live streaming and image generation capabilities 03:45 user statistics and platform popularity 06:13 subscription plans and pricing 07:13 creating and customizing your own agents 09:44 advanced features 10:44 conclusion (disclaimer: I hold $MATES token)

New EP for my consumer ai series featuring MoemateAI $mates 00:00 introduction to moemate and its unique features 01:11 exploring moemate agent sandbox and customization 02:21 live streaming and image generation capabilities 03:45 user statistics and platform popularity 06:13 subscription plans and pricing 07:13 creating and customizing your own agents 09:44 advanced features 10:44 conclusion (disclaimer: I hold $MATES token)

Sandra

26,707 views • 1 year ago

Inside Block: How AI Changes Software Development Block's Owen Jennings sat down with a16z GP David Haber to discuss how AI is changing software businesses, including the end of handwritten code, why Block reduced its workforce by 40%, how small teams are doing more with agents, and more. 00:00 Introduction 09:08 The most meaningful difference in how Block is operating 12:57 AI infrastructure build across the org 17:09 The shape of the business: Square, Cash App, Afterpay 20:00 From static UI to generative UI 23:23 Defensibility in the AI era OBJ David Haber Block

Inside Block: How AI Changes Software Development Block's Owen Jennings sat down with a16z GP David Haber to discuss how AI is changing software businesses, including the end of handwritten code, why Block reduced its workforce by 40%, how small teams are doing more with agents, and more. 00:00 Introduction 09:08 The most meaningful difference in how Block is operating 12:57 AI infrastructure build across the org 17:09 The shape of the business: Square, Cash App, Afterpay 20:00 From static UI to generative UI 23:23 Defensibility in the AI era OBJ David Haber Block

a16z

93,210 views • 3 months ago

춤 챌린지 모음 #윈터 #WINTER 00:00 Kick Back 00:17 Weekend 00:38 Second 01:02 Sparkling 01:23 Sneakers 01:35 Candy 01:55 야야야 02:08 Unforgiven 02:25 Movie Star 02:34 이프푸 02:55 Hard 03:20 최애의 아이 03:37 Vujade 03:53 It is what it is 04:06 Get A Guitar 04:16 INVITATION 04:25 Guilty 05:10 Fact Check 05:34 Maniac 05:56 WOP 06:13 Chill Kill 06:34 Trance x I Know 06:45 Sweet Venom 07:03 Siren 07:16 Hot & Cold 07:41 Rockin' Around The Pole Again 08:03 Underneath the Tree 08:21 Be There For Me 08:42 뭣 같아 08:54 UNTOUCHABLE 09:09 Two Of Hearts 09:19 Boyfriend 09:42 Nee, 09:59 움파룸파 10:17 사랑스러워 10:38 Like I Do 10:53 Like I Do (골댕이ver.) 11:08 민정이가 저장해놓은.. 11:29 갈매기.. 11:45 날 바라바라봐 12:01 GGB 12:12 홀씨 12:56 시나모롤 13:10 띵띵땅땅 13:26 EENIE MEENIE 13:47 How Sweet

춤 챌린지 모음 #윈터 #WINTER 00:00 Kick Back 00:17 Weekend 00:38 Second 01:02 Sparkling 01:23 Sneakers 01:35 Candy 01:55 야야야 02:08 Unforgiven 02:25 Movie Star 02:34 이프푸 02:55 Hard 03:20 최애의 아이 03:37 Vujade 03:53 It is what it is 04:06 Get A Guitar 04:16 INVITATION 04:25 Guilty 05:10 Fact Check 05:34 Maniac 05:56 WOP 06:13 Chill Kill 06:34 Trance x I Know 06:45 Sweet Venom 07:03 Siren 07:16 Hot & Cold 07:41 Rockin' Around The Pole Again 08:03 Underneath the Tree 08:21 Be There For Me 08:42 뭣 같아 08:54 UNTOUCHABLE 09:09 Two Of Hearts 09:19 Boyfriend 09:42 Nee, 09:59 움파룸파 10:17 사랑스러워 10:38 Like I Do 10:53 Like I Do (골댕이ver.) 11:08 민정이가 저장해놓은.. 11:29 갈매기.. 11:45 날 바라바라봐 12:01 GGB 12:12 홀씨 12:56 시나모롤 13:10 띵띵땅땅 13:26 EENIE MEENIE 13:47 How Sweet

취향의 윈터

380,597 views • 2 years ago

This week we introduced Google Antigravity, a new agentic development platform where agents operate across the editor, terminal, and browser. Check out this tutorial to get started. Highlights: 00:00 Overview 00:44 Agent Manager, editor, browser 01:48 Agent-assisted development 02:15 Initial prompt and implementation 03:30 Artifacts for agent accountability 05:09 Antigravity browser extension 06:12 Parallel tasks via the Inbox 06:40 Nano Banana image generation integration 08:08 Artifact review 11:18 Testing application 12:47 Walkthrough

This week we introduced Google Antigravity, a new agentic development platform where agents operate across the editor, terminal, and browser. Check out this tutorial to get started. Highlights: 00:00 Overview 00:44 Agent Manager, editor, browser 01:48 Agent-assisted development 02:15 Initial prompt and implementation 03:30 Artifacts for agent accountability 05:09 Antigravity browser extension 06:12 Parallel tasks via the Inbox 06:40 Nano Banana image generation integration 08:08 Artifact review 11:18 Testing application 12:47 Walkthrough

Google AI Developers

42,530 views • 8 months ago

"World models" is one of the buzziest yet ambiguous terms in AI right now. I started this video with many questions: - How are they different from video generation? - Can they do more than AI slop? - Can LeCun be trusted given that he wears knee-high white socks? Many thanks to TJ Galda and NVIDIA AI for helping me answer (most) of these questions!

"World models" is one of the buzziest yet ambiguous terms in AI right now. I started this video with many questions: - How are they different from video generation? - Can they do more than AI slop? - Can LeCun be trusted given that he wears knee-high white socks? Many thanks to TJ Galda and NVIDIA AI for helping me answer (most) of these questions!

Julia Turc

248,035 views • 1 month ago

Sergej Kunz: The next billion users will be AI agents. 1inch Co-Founder Sergej Kunz sits down with Gareth Jenkinson to discuss how liquidity fragmentation, and user experience challenges are shaping the next generation of protocols, including 1inch’s new approach with Aqua. OUTLINE 00:00 - Introduction 01:37 - Market Recovery 02:35 - Ethereum Direction 03:08 - Institutional Interest 03:45 - Institutional DeFi 05:27 - UX and Trade Risks 07:28 - Security and Hacks 09:48 - AMM Innovation 12:55 - Aave and Competition 14:52 - AI and Agents

Sergej Kunz: The next billion users will be AI agents. 1inch Co-Founder Sergej Kunz sits down with Gareth Jenkinson to discuss how liquidity fragmentation, and user experience challenges are shaping the next generation of protocols, including 1inch’s new approach with Aqua. OUTLINE 00:00 - Introduction 01:37 - Market Recovery 02:35 - Ethereum Direction 03:08 - Institutional Interest 03:45 - Institutional DeFi 05:27 - UX and Trade Risks 07:28 - Security and Hacks 09:48 - AMM Innovation 12:55 - Aave and Competition 14:52 - AI and Agents

The Block

72,304 views • 2 months ago

Everyone is focused on tracking the ways LLMs are getting better. And they are. But we know there are still things that LLMs can’t do well—the tasks where you can feel the architecture fighting the problem. So I was excited to chat with Eve Bodnia (@eve_bodnia), who is developing an alternative AI model to LLMs, on Every 📧's AI & I. Eve's argument: energy-based models (EBMs), which map possible outcomes onto a mathematical landscape, will lead to the next AI phase shift. We get into: - How energy-based models work. Likely outcomes sit in valleys, and unlikely ones sit on peaks. Whereas LLMs process one token at a time, an EBM scans the full terrain to find the lowest point, or the most probable answer. - Language-based versus data-native models. LLMs are language-dependent even when the problem has nothing to do with language. "If your data is numbers, relationships, and functions, and you try to map those rules into words and then search for the next word, you're losing a lot of information," Bodnia says. EBMs work directly with the underlying data structure, including numbers and spatial coordinates. - Sequential versus panoramic reasoning. An LLM is like driving through San Francisco without a map. Each turn constrains the next, and if you go down the wrong street, you can't reverse course. An EBM has the bird's-eye view—it can evaluate multiple routes at once and course-correct before hitting a dead end. - The LLM plateau no one wants to talk about. LLMs are getting incrementally better, step-change improvements aren’t coming, Eve argues. To achieve that, we need new solutions that compensate for what LLMs are inherently bad at, like non-language reasoning, verification, and real-time data analysis. This is a must-watch for anyone who's curious what might come after the LLM. Watch below! Timestamps: Introduction: 00:00:51 Why correctness and verifiability matter in AI: 00:02:09 What an energy-based model is: 00:09:33 How EBMs construct energy landscapes to understand data: 00:14:21 Why modeling intelligence through language alone is a flawed approach: 00:19:00 What it means for a model to "understand" data: 00:26:54 How EBMs solve the vibe coding problem and enable formally verified code: 00:37:21 Why LLM progress is plateauing: 00:43:21 Mission-critical industries haven't adopted LLMs, and why EBMs can fill that gap: 00:49:54

Everyone is focused on tracking the ways LLMs are getting better. And they are. But we know there are still things that LLMs can’t do well—the tasks where you can feel the architecture fighting the problem. So I was excited to chat with Eve Bodnia (@eve_bodnia), who is developing an alternative AI model to LLMs, on Every 📧's AI & I. Eve's argument: energy-based models (EBMs), which map possible outcomes onto a mathematical landscape, will lead to the next AI phase shift. We get into: - How energy-based models work. Likely outcomes sit in valleys, and unlikely ones sit on peaks. Whereas LLMs process one token at a time, an EBM scans the full terrain to find the lowest point, or the most probable answer. - Language-based versus data-native models. LLMs are language-dependent even when the problem has nothing to do with language. "If your data is numbers, relationships, and functions, and you try to map those rules into words and then search for the next word, you're losing a lot of information," Bodnia says. EBMs work directly with the underlying data structure, including numbers and spatial coordinates. - Sequential versus panoramic reasoning. An LLM is like driving through San Francisco without a map. Each turn constrains the next, and if you go down the wrong street, you can't reverse course. An EBM has the bird's-eye view—it can evaluate multiple routes at once and course-correct before hitting a dead end. - The LLM plateau no one wants to talk about. LLMs are getting incrementally better, step-change improvements aren’t coming, Eve argues. To achieve that, we need new solutions that compensate for what LLMs are inherently bad at, like non-language reasoning, verification, and real-time data analysis. This is a must-watch for anyone who's curious what might come after the LLM. Watch below! Timestamps: Introduction: 00:00:51 Why correctness and verifiability matter in AI: 00:02:09 What an energy-based model is: 00:09:33 How EBMs construct energy landscapes to understand data: 00:14:21 Why modeling intelligence through language alone is a flawed approach: 00:19:00 What it means for a model to "understand" data: 00:26:54 How EBMs solve the vibe coding problem and enable formally verified code: 00:37:21 Why LLM progress is plateauing: 00:43:21 Mission-critical industries haven't adopted LLMs, and why EBMs can fill that gap: 00:49:54

Dan Shipper 📧

26,900 views • 3 months ago

How does math research change when the cost of trying your first dumb idea goes to zero? Daniel Litt joins Greg Burnham and Anson Ho to discuss what today’s models can and can’t do in math, and how far they are from doing high-quality research. 0:00:00 What's the hardest math problem AI can solve today? 00:16:08 How helpful are today’s AI models for math research? 00:23:36 Junk papers, LLM-generated proofs, and the refereeing crisis 00:27:21 AI enables searching through problems at scale 00:33:49 When will AI be good enough to publish in top math journals? 00:42:15 What are the returns to intelligence? 00:59:50 Will AI solve Millennium problems? 01:11:54 Is math full of low-hanging fruit? 01:18:47 How Daniel has adapted his professional life to AI progress 01:25:28 What do AI math benchmarks actually measure? 01:33:05 Designing the Open Problems benchmark 01:56:35 Do mathematicians believe heuristic arguments about conjectures? 02:01:24 What if FrontierMath: Open Problems gets solved? 02:06:53 Is AI on the cusp of accelerating math progress?

How does math research change when the cost of trying your first dumb idea goes to zero? Daniel Litt joins Greg Burnham and Anson Ho to discuss what today’s models can and can’t do in math, and how far they are from doing high-quality research. 0:00:00 What's the hardest math problem AI can solve today? 00:16:08 How helpful are today’s AI models for math research? 00:23:36 Junk papers, LLM-generated proofs, and the refereeing crisis 00:27:21 AI enables searching through problems at scale 00:33:49 When will AI be good enough to publish in top math journals? 00:42:15 What are the returns to intelligence? 00:59:50 Will AI solve Millennium problems? 01:11:54 Is math full of low-hanging fruit? 01:18:47 How Daniel has adapted his professional life to AI progress 01:25:28 What do AI math benchmarks actually measure? 01:33:05 Designing the Open Problems benchmark 01:56:35 Do mathematicians believe heuristic arguments about conjectures? 02:01:24 What if FrontierMath: Open Problems gets solved? 02:06:53 Is AI on the cusp of accelerating math progress?

Epoch AI

178,703 views • 5 months ago

Ep 105 out now on X! Joe Lonsdale sits down with @mbrendan1, founder & chair of Cosmos Institute Harnessing AI for freedom vs control; doomsayers vs e/acc; training "philosopher builders"; why LLMs can't reason; and more. (00:00) Episode intro (02:04) Submarines to building AI companies (06:00) Why Philosopher Builders are needed (13:20) Why AI poses unique challenges (20:50) Doomsayers vs Accelerationists (25:50) Philosophical frameworks for human flourishing (30:10) Regulators & techno-authoritarians (37:34) AI running for political office? (40:07) LLMs don't know how to reason (42:30) Purpose of the Cosmos Institute

Ep 105 out now on X! Joe Lonsdale sits down with @mbrendan1, founder & chair of Cosmos Institute Harnessing AI for freedom vs control; doomsayers vs e/acc; training "philosopher builders"; why LLMs can't reason; and more. (00:00) Episode intro (02:04) Submarines to building AI companies (06:00) Why Philosopher Builders are needed (13:20) Why AI poses unique challenges (20:50) Doomsayers vs Accelerationists (25:50) Philosophical frameworks for human flourishing (30:10) Regulators & techno-authoritarians (37:34) AI running for political office? (40:07) LLMs don't know how to reason (42:30) Purpose of the Cosmos Institute

American Optimist

49,282 views • 1 year ago

The last 5 years have been dominated by LLMs. amit makes a compelling case for visual intelligence dominating the next 5. We discussed the journey from South Park Commons, to Luma, to world models (what does this even mean!?). Enjoy! (00:00) Apple's LiDAR work seeded Luma's founding vision (05:05) - A free app was secretly a data collection operation (06:49) H100s made video generation stop feeling impossible (10:00) Most definitions of "world model" are simply wrong (14:37) Why unified models beat pure video scaling (20:03) Luma Agents and the end-to-end creative production loop (23:00) The knowledge gap no model can close alone (29:34) Blowing James Cameron's mind in five minutes (31:16) Why consumer video generation failed—and who it's actually for (35:23) - Product and research aren't independent things at a foundation lab

The last 5 years have been dominated by LLMs. amit makes a compelling case for visual intelligence dominating the next 5. We discussed the journey from South Park Commons, to Luma, to world models (what does this even mean!?). Enjoy! (00:00) Apple's LiDAR work seeded Luma's founding vision (05:05) - A free app was secretly a data collection operation (06:49) H100s made video generation stop feeling impossible (10:00) Most definitions of "world model" are simply wrong (14:37) Why unified models beat pure video scaling (20:03) Luma Agents and the end-to-end creative production loop (23:00) The knowledge gap no model can close alone (29:34) Blowing James Cameron's mind in five minutes (31:16) Why consumer video generation failed—and who it's actually for (35:23) - Product and research aren't independent things at a foundation lab

Finn Meeks

24,544 views • 1 month ago

Can AI truly understand concepts, or is it just processing data? 🤖 Join Hannah Fry as she discusses this and more with our Principal Scientist Murray Shanahan. They break down the meaning of consciousness and explore how it might – or might not – apply to AI, as well as exploring the anthropomorphism of multimodal models. Watch our podcast now ↓ Timestamps: 00:00 Introduction 02:00 Ex Machina 03:28 John McCarthy & Coining "AI" 05:45 Symbolic AI 09:43 AI Reasoning 14:42 Turing Test 18:10 Alternative Tests 21:58 Anthropomorphism 26:42 Embodied AI 28:17 AI Consciousness 34:00 Octopus Analogy 37:48 Prompt Engineering 39:40 Exotic Mind-Like Entities 40:44 Conclusion

Can AI truly understand concepts, or is it just processing data? 🤖 Join Hannah Fry as she discusses this and more with our Principal Scientist Murray Shanahan. They break down the meaning of consciousness and explore how it might – or might not – apply to AI, as well as exploring the anthropomorphism of multimodal models. Watch our podcast now ↓ Timestamps: 00:00 Introduction 02:00 Ex Machina 03:28 John McCarthy & Coining "AI" 05:45 Symbolic AI 09:43 AI Reasoning 14:42 Turing Test 18:10 Alternative Tests 21:58 Anthropomorphism 26:42 Embodied AI 28:17 AI Consciousness 34:00 Octopus Analogy 37:48 Prompt Engineering 39:40 Exotic Mind-Like Entities 40:44 Conclusion

Google DeepMind

291,210 views • 1 year ago