Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

I re-recorded the post-training part of our NeurIPS tutorial on language models, added some more slides, and wrote up a mini state of the union on Interconnects. Enjoy! Links in QT. 00:00 Introduction 10:00 Prompts & Skill Selection 14:19 Instruction Finetuning 21:45 Preference Finetuning 36:17 Reinforcement Finetuning 45:28 Open... show more

Nathan Lambert

70,357 subscribers

49,310 görüntüleme • 1 yıl önce •via X (Twitter)

Bilim & Teknoloji Sağlık & İyilik Eğitim

Anya Rossi• Live Now

Private livecam show

11 Yorum

Nathan Lambert profil fotoğrafı

Nathan Lambert1 yıl önce

Second, I think YouTube is an important distribution mechanism for AI research and education. Lmk if there's something I should do more of.

Nathan Lambert profil fotoğrafı

Nathan Lambert1 yıl önce

Also continuing to plug away at rlhf book but waiting to announce until v1

Power Homeschool profil fotoğrafı

Power Homeschool2 yıl önce

Start your child's independent homeschooling journey with our flexible, self-paced platform. We offer a full PreK-12 curriculum with electives & hundreds of engaging video-based lessons taught by experienced educators! Enroll now! ✅ No Contracts. Cancel Anytime.

Cameron R. Wolfe, Ph.D. profil fotoğrafı

Cameron R. Wolfe, Ph.D.1 yıl önce

@interconnectsai Best talk at NeurIPS this year IMO. Everyone should watch.

Zion profil fotoğrafı

Zion1 yıl önce

@interconnectsai Thanks for the Knowledge Nathan

Michael Craig profil fotoğrafı

Michael Craig1 yıl önce

@interconnectsai Thanks for sharing @natolambert. Saw the series of talks at NeurIPS this year as well, very informative.

Zayn Al-Mahdi profil fotoğrafı

Zayn Al-Mahdi1 yıl önce

@interconnectsai Excellent re-record, Nathan! Your insights on instruction and reinforcement fine-tuning are spot on. Curious about their real-world applications. Diving into the blog now.

altyni (ethmag.eth) 🪄 profil fotoğrafı

altyni (ethmag.eth) 🪄1 yıl önce

@interconnectsai What are your thoughts on post-training for JSON outputs and function calling?

Atma profil fotoğrafı

Atma1 yıl önce

@interconnectsai Nice talk, would you mind sharing the slides?

Matthew Lam profil fotoğrafı

Matthew Lam1 yıl önce

@interconnectsai Thanks for the knowledge

Ashvini Jindal profil fotoğrafı

Ashvini Jindal1 yıl önce

@interconnectsai Thanks for sharing!

Benzer Videolar

We've got a podcast! In our first episode, Ege, Tamay and Jaime dig into: • What they expect AI to look like by 2030 • Why economists are underestimating the likelihood of explosive growth • The startling regularity in technological trends like Moore's Law • Moravec’s paradox, and how we might overcome it And more more! Timestamps: 00:00:00 Preview 00:00:37 What is Epoch AI? 00:02:32 Scaling Laws 00:08:43 Key Drivers 00:19:20 End of the Decade Predictions 00:21:18 Bottlenecks: Power 00:27:59 Bottlenecks: GPUs 00:32:07 Bottlenecks: Data 00:45:37 Bottlenecks: Latency 00:56:18 Bottlenecks: Failure Rates 01:03:55 AI Investment 01:07:11 Automation 01:12:10 Benchmarks & Moravec’s Paradox 01:19:45 Economic Impact 01:45:48 Open Questions & Takeaways

We've got a podcast! In our first episode, Ege, Tamay and Jaime dig into: • What they expect AI to look like by 2030 • Why economists are underestimating the likelihood of explosive growth • The startling regularity in technological trends like Moore's Law • Moravec’s paradox, and how we might overcome it And more more! Timestamps: 00:00:00 Preview 00:00:37 What is Epoch AI? 00:02:32 Scaling Laws 00:08:43 Key Drivers 00:19:20 End of the Decade Predictions 00:21:18 Bottlenecks: Power 00:27:59 Bottlenecks: GPUs 00:32:07 Bottlenecks: Data 00:45:37 Bottlenecks: Latency 00:56:18 Bottlenecks: Failure Rates 01:03:55 AI Investment 01:07:11 Automation 01:12:10 Benchmarks & Moravec’s Paradox 01:19:45 Economic Impact 01:45:48 Open Questions & Takeaways

Epoch AI

152,591 görüntüleme • 1 yıl önce

Here's my conversation with Lucas Atkins and the team at Arcee.ai on their path to training and releasing Trinity Large today. From going all in on open models built end to end in the US 6 months ago to having the model in hand is no easy feet. I loved this conversation on how to design a startup around open models and take a bold step to scale it up. I'm openly an Arcee fan, watching them take risk and pull it off. We discuss: - The state (and future) of open vs. closed models, - The business of selling open models for on-prem deployments, - The story of Arcee AI & going “all-in” on this training run, - The ATOM project, - Building frontier model training teams in 6 months, - and other great topics. I really loved this one, and think you well too. Chapters: 00:00:00 Intro: Arcee AI, Trinity Models & Trinity Large 00:08:26 Transitioning a Company to Pre-training 00:13:00 Technical Decisions: Muon and MoE 00:18:41 Scaling and MoE Training Pain 00:23:14 Post-training and RL Strategies 00:28:09 Team Structure and Data Scaling 00:31:31 The Trinity Manifesto: US Open Weights 00:42:31 Specialized Models and Distillation 00:47:12 Infrastructure and Hosting 400B 00:50:53 Open Source as a Business Moat 00:56:31 Predictions: Best Model in 2026 01:02:29 Lightning Round & Conclusions More great open model builder podcasts coming soon!

Here's my conversation with Lucas Atkins and the team at Arcee.ai on their path to training and releasing Trinity Large today. From going all in on open models built end to end in the US 6 months ago to having the model in hand is no easy feet. I loved this conversation on how to design a startup around open models and take a bold step to scale it up. I'm openly an Arcee fan, watching them take risk and pull it off. We discuss: - The state (and future) of open vs. closed models, - The business of selling open models for on-prem deployments, - The story of Arcee AI & going “all-in” on this training run, - The ATOM project, - Building frontier model training teams in 6 months, - and other great topics. I really loved this one, and think you well too. Chapters: 00:00:00 Intro: Arcee AI, Trinity Models & Trinity Large 00:08:26 Transitioning a Company to Pre-training 00:13:00 Technical Decisions: Muon and MoE 00:18:41 Scaling and MoE Training Pain 00:23:14 Post-training and RL Strategies 00:28:09 Team Structure and Data Scaling 00:31:31 The Trinity Manifesto: US Open Weights 00:42:31 Specialized Models and Distillation 00:47:12 Infrastructure and Hosting 400B 00:50:53 Open Source as a Business Moat 00:56:31 Predictions: Best Model in 2026 01:02:29 Lightning Round & Conclusions More great open model builder podcasts coming soon!

Nathan Lambert

26,872 görüntüleme • 5 ay önce

Another quick lecture -- I've been asked many times for prereq's to my book and what you should know, so built a little lecture (with GLM 5.2) to cover some more basics. Topics include: 00:00 Introduction & Course Prerequisites 01:37 Language Models Overview 02:47 The LM Head 04:29 Softmax & Log-Probabilities 06:13 Anatomy of an LM Training Example 06:37 Computing LLM Probabilities (+Phoebe the Dog) 09:52 Three Common Masks in Post-Training 11:03 A Small Decoding Review 12:14 Training an LM: Cross-Entropy 13:23 Optimization & Fine-Tuning 13:55 Pretraining to Midtraining to SFT Pipeline 15:25 Probability Essentials: KL Divergence & Entropy 19:36 Sigmoid & Pairwise Likelihood 20:29 Reinforcement Learning Framing (MDP) 22:28 Transitioning Tools into Post-Training 23:12 Recommended Resources & Wrap-Up  Happy learning and I'm still taking questions from during the course for Q&A videos.

Another quick lecture -- I've been asked many times for prereq's to my book and what you should know, so built a little lecture (with GLM 5.2) to cover some more basics. Topics include: 00:00 Introduction & Course Prerequisites 01:37 Language Models Overview 02:47 The LM Head 04:29 Softmax & Log-Probabilities 06:13 Anatomy of an LM Training Example 06:37 Computing LLM Probabilities (+Phoebe the Dog) 09:52 Three Common Masks in Post-Training 11:03 A Small Decoding Review 12:14 Training an LM: Cross-Entropy 13:23 Optimization & Fine-Tuning 13:55 Pretraining to Midtraining to SFT Pipeline 15:25 Probability Essentials: KL Divergence & Entropy 19:36 Sigmoid & Pairwise Likelihood 20:29 Reinforcement Learning Framing (MDP) 22:28 Transitioning Tools into Post-Training 23:12 Recommended Resources & Wrap-Up  Happy learning and I'm still taking questions from during the course for Q&A videos.

Nathan Lambert

50,424 görüntüleme • 21 gün önce

New episode with avery.apt 🇺🇸 out now! We discuss: The Aptos roadmap Crypto's end state Building an L1 The rise of AI, building on Aptos & more! Timestamps: 00:00 Introduction 00:26 The Future Of Aptos 07:39 Building Products In Crypto 11:09 Designing A High Performance L1 15:10 The Move VM 18:52 Crypto's Narrative Cycle 21:43 Building For Institutions vs Crypto Natives 25:14 Ads (Ledger & Mantle) 26:53 The Aptos Roadmap 34:00 The Growth Of Solana 36:17 Crypto's Ultimate End State 39:22 The Rise Of AI 42:03 Misconceptions About Aptos 44:38 Getting Started On Aptos Links below! ↓

New episode with avery.apt 🇺🇸 out now! We discuss: The Aptos roadmap Crypto's end state Building an L1 The rise of AI, building on Aptos & more! Timestamps: 00:00 Introduction 00:26 The Future Of Aptos 07:39 Building Products In Crypto 11:09 Designing A High Performance L1 15:10 The Move VM 18:52 Crypto's Narrative Cycle 21:43 Building For Institutions vs Crypto Natives 25:14 Ads (Ledger & Mantle) 26:53 The Aptos Roadmap 34:00 The Growth Of Solana 36:17 Crypto's Ultimate End State 39:22 The Rise Of AI 42:03 Misconceptions About Aptos 44:38 Getting Started On Aptos Links below! ↓

Empire 🟪

54,990 görüntüleme • 1 yıl önce

"If NASA and United States can not beat the Chinese back to the moon, then the moon belongs to the Chinese, and we get what we deserve." You don't want to miss this interview. I met Casey Handmer (Casey Handmer) while flying fighter jets in Montana. I Immediately I knew he had a story to tell. Turns out he wants to tile the United States in solar panels, and use that power to generate rocket fuel and clean water for tens of millions of people. Basically to re-invent American industry. In this interview Casey discusses the innovative pursuits of his company Terraform Industries. He reflects on his previous roles at NASA and JPL, where he was frustrated with the inefficiencies and lack of innovation in traditional institutions. Handmer discusses the importance of ambition, self-awareness, curiosity, and practical experience in hiring and developing talent at Terraform. To fully understand the scope of Casey’s ambitions, you should visit his blog where he discusses things like Mars helicopters and desalination to support a new mega city in California. 00:00:00 - 00:00:17 Geopolitical Stakes: US vs. China in Space Race 00:00:17 - 00:01:17 Interview Introduction: Casey's Background and Topics 00:01:17 - 00:03:10 Podcast Setup and Welcome to Terraform Industries 00:03:10 - 00:04:40 Building History and Manufacturing Interest 00:04:40 - 00:07:17 Meeting at Hangar Event and Aircraft Engineering 00:07:17 - 00:20:45 Early Life: Hands-On Building and Career Path 00:20:45 - 00:40:00 Founding Terraform: Synthetic Fuels Vision and Prototypes 00:40:00 - 01:00:00 Technology Deep Dive: Solar to Hydrocarbons and Scaling 01:00:00 - 01:20:00 Energy Geopolitics and US Reindustrialization 01:20:00 - 01:40:00 Team Building, Hiring Philosophy, and Company Culture 01:40:00 - 02:00:00 Hardware vs. Software Challenges and Entrepreneur Advice 02:00:00 - 02:22:24 Future Visions: Desalination and New Cities 02:22:24 - 02:36:39 Closing Call to Action and Thanks

"If NASA and United States can not beat the Chinese back to the moon, then the moon belongs to the Chinese, and we get what we deserve." You don't want to miss this interview. I met Casey Handmer (Casey Handmer) while flying fighter jets in Montana. I Immediately I knew he had a story to tell. Turns out he wants to tile the United States in solar panels, and use that power to generate rocket fuel and clean water for tens of millions of people. Basically to re-invent American industry. In this interview Casey discusses the innovative pursuits of his company Terraform Industries. He reflects on his previous roles at NASA and JPL, where he was frustrated with the inefficiencies and lack of innovation in traditional institutions. Handmer discusses the importance of ambition, self-awareness, curiosity, and practical experience in hiring and developing talent at Terraform. To fully understand the scope of Casey’s ambitions, you should visit his blog where he discusses things like Mars helicopters and desalination to support a new mega city in California. 00:00:00 - 00:00:17 Geopolitical Stakes: US vs. China in Space Race 00:00:17 - 00:01:17 Interview Introduction: Casey's Background and Topics 00:01:17 - 00:03:10 Podcast Setup and Welcome to Terraform Industries 00:03:10 - 00:04:40 Building History and Manufacturing Interest 00:04:40 - 00:07:17 Meeting at Hangar Event and Aircraft Engineering 00:07:17 - 00:20:45 Early Life: Hands-On Building and Career Path 00:20:45 - 00:40:00 Founding Terraform: Synthetic Fuels Vision and Prototypes 00:40:00 - 01:00:00 Technology Deep Dive: Solar to Hydrocarbons and Scaling 01:00:00 - 01:20:00 Energy Geopolitics and US Reindustrialization 01:20:00 - 01:40:00 Team Building, Hiring Philosophy, and Company Culture 01:40:00 - 02:00:00 Hardware vs. Software Challenges and Entrepreneur Advice 02:00:00 - 02:22:24 Future Visions: Desalination and New Cities 02:22:24 - 02:36:39 Closing Call to Action and Thanks

Penny2x

1,598,196 görüntüleme • 9 ay önce

We don't know what most microbial genes do. Can genomic language models help? there's only one way to find out! this is a 1 hour and 42 minute interview with an MIT professor (the famous Yunha Hwang) chatting about these questions, her work in solving them at Tatta Bio, and more. zoomer captions are back too Links in reply! Timestamps: 00:00:00 - Clips + sponsor roll from the wonderful LatchBio 00:02:07 – Introduction 00:02:23 – Why do microbial genomes matter 00:04:07 – Deep learning acceptance in metagenomics 00:05:25 – The case for genomic “context” over sequence matching 00:06:43 – OMG: the only ML-ready metagenomic dataset 00:09:27 – gLM2: A multimodal genomic language model 00:11:06 – What do you do with the output of genomic language models? 00:17:41 – How will OMG evolve? 00:20:26 – Why train on only microbial genomes, as opposed to all genomes? 00:22:58 – Do we need more sequences or more annotations? 00:23:54 – Is there a conserved microbial genome ‘language’? 00:28:11 – What non-obvious things can this genomic language model tell you? 00:33:08 – Semantic deduplication and evaluation 00:37:33 – How does benchmarking work for these types of models? 00:41:31 – Gaia: A genomic search engine 00:44:18 – Even ‘well-studied’ genomes are mostly unannotated 00:50:51 – Using agents on Gaia 00:54:53 – Will genomic language models reshape the tree of life? 00:59:18 – Current limitations of genomic language models 01:08:54 – Directed evolution as training data 01:12:35 – What is Tatta Bio? 01:19:02 – Building Google for genomic sequences (SeqHub) 01:25:46 – How to create communities around scientific OSS 01:29:06 – What’s the purpose in the centralization of the software? 01:35:37 – How will the way science is done change in 10 years?

We don't know what most microbial genes do. Can genomic language models help? there's only one way to find out! this is a 1 hour and 42 minute interview with an MIT professor (the famous Yunha Hwang) chatting about these questions, her work in solving them at Tatta Bio, and more. zoomer captions are back too Links in reply! Timestamps: 00:00:00 - Clips + sponsor roll from the wonderful LatchBio 00:02:07 – Introduction 00:02:23 – Why do microbial genomes matter 00:04:07 – Deep learning acceptance in metagenomics 00:05:25 – The case for genomic “context” over sequence matching 00:06:43 – OMG: the only ML-ready metagenomic dataset 00:09:27 – gLM2: A multimodal genomic language model 00:11:06 – What do you do with the output of genomic language models? 00:17:41 – How will OMG evolve? 00:20:26 – Why train on only microbial genomes, as opposed to all genomes? 00:22:58 – Do we need more sequences or more annotations? 00:23:54 – Is there a conserved microbial genome ‘language’? 00:28:11 – What non-obvious things can this genomic language model tell you? 00:33:08 – Semantic deduplication and evaluation 00:37:33 – How does benchmarking work for these types of models? 00:41:31 – Gaia: A genomic search engine 00:44:18 – Even ‘well-studied’ genomes are mostly unannotated 00:50:51 – Using agents on Gaia 00:54:53 – Will genomic language models reshape the tree of life? 00:59:18 – Current limitations of genomic language models 01:08:54 – Directed evolution as training data 01:12:35 – What is Tatta Bio? 01:19:02 – Building Google for genomic sequences (SeqHub) 01:25:46 – How to create communities around scientific OSS 01:29:06 – What’s the purpose in the centralization of the software? 01:35:37 – How will the way science is done change in 10 years?

owl

44,279 görüntüleme • 7 ay önce

Nobody knows Oakmont like Bob Ford. Happy U.S. Open week. 00:05:59 – Playing in the U.S. Open at his own club 00:07:01 – Making more money than the U.S. Open winner 00:08:02 – Power outage, piles of cash, and U.S. Open chaos 00:10:11 – Becoming the voice of the U.S. Open 00:18:18 – Previewing how Oakmont will hold up this week 00:21:26 – The famed Oakmont SWAT 00:30:01 – 1,000 trees gone before sunrise 00:34:14 – Tanking to avoid breaking Palmer's course record 00:44:30 – A day on the tee box at the U.S. Open

Nobody knows Oakmont like Bob Ford. Happy U.S. Open week. 00:05:59 – Playing in the U.S. Open at his own club 00:07:01 – Making more money than the U.S. Open winner 00:08:02 – Power outage, piles of cash, and U.S. Open chaos 00:10:11 – Becoming the voice of the U.S. Open 00:18:18 – Previewing how Oakmont will hold up this week 00:21:26 – The famed Oakmont SWAT 00:30:01 – 1,000 trees gone before sunrise 00:34:14 – Tanking to avoid breaking Palmer's course record 00:44:30 – A day on the tee box at the U.S. Open

The Golfer's Journal

42,920 görüntüleme • 1 yıl önce

As ever, a fascinating chat with my friend Jordi Visser on AI, crypto, and hedge funds' future... Enjoy! 00:00:00 Introduction 00:00:54 The Big Picture: AI, Crypto, and Macro 00:02:25 Jordie Visser’s Background in Macro and Finance 00:03:30 AI’s Disruptive Impact on the Economy 00:06:32 Rethinking GDP and Productivity 00:08:59 The Future of Labor and Profit Margins 00:10:49 The Rise of the Speculative Economy 00:14:41 Generational Shifts in Investing and Risk-taking 00:16:18 AI and Crypto: A Global Network Effect 00:19:38 Stablecoins and AI-Driven Market Volume Growth 00:24:28 The End of Traditional Hedge Funds? 00:31:16 Deep Research and the Power of AI Analysis 00:35:29 The Acceleration of AI and its Economic Impact 00:41:02 The Growing Crypto Adoption and Its Role in the Future Economy 00:44:22 Shifting from Fear to Optimism in Markets 00:50:27 The Political and Economic Challenges Ahead 00:56:01 Quantum Computing and the Next Leap in AI 01:02:06 Closing Thoughts: Preparing for a Radical Future

As ever, a fascinating chat with my friend Jordi Visser on AI, crypto, and hedge funds' future... Enjoy! 00:00:00 Introduction 00:00:54 The Big Picture: AI, Crypto, and Macro 00:02:25 Jordie Visser’s Background in Macro and Finance 00:03:30 AI’s Disruptive Impact on the Economy 00:06:32 Rethinking GDP and Productivity 00:08:59 The Future of Labor and Profit Margins 00:10:49 The Rise of the Speculative Economy 00:14:41 Generational Shifts in Investing and Risk-taking 00:16:18 AI and Crypto: A Global Network Effect 00:19:38 Stablecoins and AI-Driven Market Volume Growth 00:24:28 The End of Traditional Hedge Funds? 00:31:16 Deep Research and the Power of AI Analysis 00:35:29 The Acceleration of AI and its Economic Impact 00:41:02 The Growing Crypto Adoption and Its Role in the Future Economy 00:44:22 Shifting from Fear to Optimism in Markets 00:50:27 The Political and Economic Challenges Ahead 00:56:01 Quantum Computing and the Next Leap in AI 01:02:06 Closing Thoughts: Preparing for a Radical Future

Raoul Pal

280,145 görüntüleme • 1 yıl önce

Decentralized AI Product Deployment with KIP Protocol Workshop Recap 🎦 Learn from top educators on creating your own Knowledge Base and deploying AI RAG dApps on KIP Protocol 💡 In this TinTinLand workshop, our Chief AI Officer, Jen D., showcases the power of KIP Developer Sandbox👩‍💻 Highlights: 00:00 Introduction 00:40 KIP Ecosystem Overview 01:23 KIP Identification Contract 02:55 KIP Registration Contract 03:36 KIP Service Contract 06:31 AMA Summary and Clarification 07:50 Smart Contract Interaction Demo 19:36 Sandbox Site Introduction 21:12 Account Abstraction and Data Upload 24:17 Creating AI dApps 32:33 Prompt Settings and Model Selection 37:11 App Testing 45:01 Main Net Deployment and Testing Tokens 45:36 Community Engagement and Follow-Up 46:27 Q&A Session 50:21 Conclusion What will you build? 🔧

Decentralized AI Product Deployment with KIP Protocol Workshop Recap 🎦 Learn from top educators on creating your own Knowledge Base and deploying AI RAG dApps on KIP Protocol 💡 In this TinTinLand workshop, our Chief AI Officer, Jen D., showcases the power of KIP Developer Sandbox👩‍💻 Highlights: 00:00 Introduction 00:40 KIP Ecosystem Overview 01:23 KIP Identification Contract 02:55 KIP Registration Contract 03:36 KIP Service Contract 06:31 AMA Summary and Clarification 07:50 Smart Contract Interaction Demo 19:36 Sandbox Site Introduction 21:12 Account Abstraction and Data Upload 24:17 Creating AI dApps 32:33 Prompt Settings and Model Selection 37:11 App Testing 45:01 Main Net Deployment and Testing Tokens 45:36 Community Engagement and Follow-Up 46:27 Q&A Session 50:21 Conclusion What will you build? 🔧

KIP Protocol

181,816 görüntüleme • 2 yıl önce

Hash Rate - Ep 145: Templar (3), Basilica (39), Grail (81) 🧙 Guest: Distributed State of templar + basilica + grail 00:00 Introduction to Covenant and Its Subnets 02:01 Understanding Templar: Decentralized Training 05:31 The Role of Grail and Basilica 10:38 Transitioning from Research to Product 14:37 Competitive Landscape and Model Development 21:58 Target Audience and Market Potential 25:51 The Effect of $TAO Flow 32:52 Incentive Mechanisms in Basilica 39:24 Vision for Templar: Democratizing AI Training 45:20 Future Vision

Hash Rate - Ep 145: Templar (3), Basilica (39), Grail (81) 🧙 Guest: Distributed State of templar + basilica + grail 00:00 Introduction to Covenant and Its Subnets 02:01 Understanding Templar: Decentralized Training 05:31 The Role of Grail and Basilica 10:38 Transitioning from Research to Product 14:37 Competitive Landscape and Model Development 21:58 Target Audience and Market Potential 25:51 The Effect of $TAO Flow 32:52 Incentive Mechanisms in Basilica 39:24 Vision for Templar: Democratizing AI Training 45:20 Future Vision

Mark Jeffrey

27,995 görüntüleme • 8 ay önce

The best AI startup isn't being built in SF right now. Software engineering is the easiest job in the world. NVIDIA won't be the future of AI hardware. Solo founders are doomed to fail. These are just some of the hot takes we got Toronto Tech Week's Cursor event... Watch full video below. 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 (00:00) Intro (00:31) AI Startup Not in SF (00:45) Coding Is Too Easy (01:17) Networking Is a Game (01:52) Post Nvidia AI Future (02:16) Solo Founders Fail (02:28) AI Hits Consciousness Wall (02:34) North America Bad Fashion (02:56) Self Hosted SLM Takeover (03:13) Codex vs Claude Code (03:28) Underdog Universities Win (03:37) LinkedIn Posting Advantage (04:04) Wrap Up

The best AI startup isn't being built in SF right now. Software engineering is the easiest job in the world. NVIDIA won't be the future of AI hardware. Solo founders are doomed to fail. These are just some of the hot takes we got Toronto Tech Week's Cursor event... Watch full video below. 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 (00:00) Intro (00:31) AI Startup Not in SF (00:45) Coding Is Too Easy (01:17) Networking Is a Game (01:52) Post Nvidia AI Future (02:16) Solo Founders Fail (02:28) AI Hits Consciousness Wall (02:34) North America Bad Fashion (02:56) Self Hosted SLM Takeover (03:13) Codex vs Claude Code (03:28) Underdog Universities Win (03:37) LinkedIn Posting Advantage (04:04) Wrap Up

Julia Fedorin

11,069 görüntüleme • 1 ay önce

This week on The Knowledge Project: A rare interview with Bruce Flatt, the quiet force behind 19% annual returns for 30 years, reveals the investment principles that transformed Brookfield into a $1T global powerhouse. "Our view always has been 'Buy the best.' Own the best, buy the best; continue to reinvest into the best." (00:00:54) Changes in Investing Over the Past 25 Years (00:02:50) How Private Enterprise Has Built Our Tech Infrastructure (00:05:07) Implications and Opportunities of Passive Investing (00:14:40) AI (00:19:39) Transition to Low-Carbon Energy (00:23:19) Future of Data Centers (00:25:28) De-globalization of Industry (00:27:55) Implications of Manufacturing Repatriation (00:29:06) Long-term Prospects for America (00:31:35) Changes in Real Estate (00:34:13) Approach to Risk and Debt (00:38:40) Managing Market Dislocations (00:40:22) Long-term Investing Strategy (00:42:59) History and Future of Brookfield (00:45:48) Exploration of Private Markets and Insurance (00:46:41) Investment Decision Process (00:48:34) Introduction to Annuities (00:53:10) Understanding Brookfield’s Structure (01:00:51) Focus on Downside Protection (01:02:56) Accountability in Investment Decisions (01:04:24) Understanding Investment Cycles (01:06:59) Postmortem Analysis of Investments (01:09:07) Consideration of Geopolitical Risks

This week on The Knowledge Project: A rare interview with Bruce Flatt, the quiet force behind 19% annual returns for 30 years, reveals the investment principles that transformed Brookfield into a $1T global powerhouse. "Our view always has been 'Buy the best.' Own the best, buy the best; continue to reinvest into the best." (00:00:54) Changes in Investing Over the Past 25 Years (00:02:50) How Private Enterprise Has Built Our Tech Infrastructure (00:05:07) Implications and Opportunities of Passive Investing (00:14:40) AI (00:19:39) Transition to Low-Carbon Energy (00:23:19) Future of Data Centers (00:25:28) De-globalization of Industry (00:27:55) Implications of Manufacturing Repatriation (00:29:06) Long-term Prospects for America (00:31:35) Changes in Real Estate (00:34:13) Approach to Risk and Debt (00:38:40) Managing Market Dislocations (00:40:22) Long-term Investing Strategy (00:42:59) History and Future of Brookfield (00:45:48) Exploration of Private Markets and Insurance (00:46:41) Investment Decision Process (00:48:34) Introduction to Annuities (00:53:10) Understanding Brookfield’s Structure (01:00:51) Focus on Downside Protection (01:02:56) Accountability in Investment Decisions (01:04:24) Understanding Investment Cycles (01:06:59) Postmortem Analysis of Investments (01:09:07) Consideration of Geopolitical Risks

The Knowledge Project

453,197 görüntüleme • 1 yıl önce

Here's my conversation with Edward Gibson (Ted Gibson, Language Lab MIT), a linguist and psychologist at MIT, heading the MIT Language Lab. We talk all about the human language: syntax, grammar, structure, theories of language, evolution of language, how it reflects culture, and of course LLMs, both their amazing power and their limitations. It's here on X in full, and is up on YouTube, Spotify, and everywhere else. Links in comment. Timestamps: 0:00 - Introduction 1:13 - Human language 5:19 - Generalizations in language 11:06 - Dependency grammar 21:05 - Morphology 29:40 - Evolution of languages 33:00 - Noam Chomsky 1:17:06 - Thinking and language 1:30:36 - LLMs 1:43:35 - Center embedding 2:10:02 - Learning a new language 2:13:54 - Nature vs nurture 2:20:30 - Culture and language 2:34:58 - Universal language 2:39:21 - Language translation 2:42:36 - Animal communication

Here's my conversation with Edward Gibson (Ted Gibson, Language Lab MIT), a linguist and psychologist at MIT, heading the MIT Language Lab. We talk all about the human language: syntax, grammar, structure, theories of language, evolution of language, how it reflects culture, and of course LLMs, both their amazing power and their limitations. It's here on X in full, and is up on YouTube, Spotify, and everywhere else. Links in comment. Timestamps: 0:00 - Introduction 1:13 - Human language 5:19 - Generalizations in language 11:06 - Dependency grammar 21:05 - Morphology 29:40 - Evolution of languages 33:00 - Noam Chomsky 1:17:06 - Thinking and language 1:30:36 - LLMs 1:43:35 - Center embedding 2:10:02 - Learning a new language 2:13:54 - Nature vs nurture 2:20:30 - Culture and language 2:34:58 - Universal language 2:39:21 - Language translation 2:42:36 - Animal communication

Lex Fridman

340,246 görüntüleme • 2 yıl önce

What Paramedics Saw During COVID: Harry Fisher on Bret Weinstein's Inside Rail 00:00:00 Welcome Harry Fisher, Paramedic & Military Veteran 00:01:02 What Are the Roles of an EMT & Paramedic? 00:07:29 Life Before vs. After COVID 00:09:06 Questioning the Logic Behind Vaccinating Medical Professionals First 00:10:16 When Activity Started to Pick Up 00:14:45 The Viral Video that Got Harry Labeled a Terrorist 00:24:15 Reading the Insert Online Got Harry Fired 00:30:40 After the Shots Rolled Out 00:37:02 Aortic Dissection 00:43:56 Prior to COVID: Heart Attacks for 25 & Under? 00:48:29 Anecdotal Experiences 00:52:09 The 2020 Pattern, Pre-Vaccines 00:56:06 The "Real World" in 2025 Recognizes the Patterns 01:00:00 Paramedics are Short Staffed 01:07:58 Will Paramedics Stand Up Together? 01:14:20 Faith & the Medical Freedom Movement 01:25:09 Read Harry Fisher's Book & Follow Him on X

What Paramedics Saw During COVID: Harry Fisher on Bret Weinstein's Inside Rail 00:00:00 Welcome Harry Fisher, Paramedic & Military Veteran 00:01:02 What Are the Roles of an EMT & Paramedic? 00:07:29 Life Before vs. After COVID 00:09:06 Questioning the Logic Behind Vaccinating Medical Professionals First 00:10:16 When Activity Started to Pick Up 00:14:45 The Viral Video that Got Harry Labeled a Terrorist 00:24:15 Reading the Insert Online Got Harry Fired 00:30:40 After the Shots Rolled Out 00:37:02 Aortic Dissection 00:43:56 Prior to COVID: Heart Attacks for 25 & Under? 00:48:29 Anecdotal Experiences 00:52:09 The 2020 Pattern, Pre-Vaccines 00:56:06 The "Real World" in 2025 Recognizes the Patterns 01:00:00 Paramedics are Short Staffed 01:07:58 Will Paramedics Stand Up Together? 01:14:20 Faith & the Medical Freedom Movement 01:25:09 Read Harry Fisher's Book & Follow Him on X

The Darkhorse Podcast

19,424 görüntüleme • 8 ay önce

That’s the end of the Austin House season! Mizkif has flown away in a Klat Airlines helicopter…. to a new journey, a new chapter!❤️ Here are the (3) VOD links and Timestamps: First VOD: 00:00/29:00 - Intro with all the old intro songs 29:00 - Start of Stream 30:00 - Reacting to TopKek 41:30 - Reacting to NEW TopKek (The Streamer Awards - Twitch Fails) 52:00 - Reacting to Erobb and Jason Second VOD: Starts Right Away with Super Mario 64 01:59:00 - Mizkif plays Minecraft Hardcore 03:46:00 - Mizkif plays Fall Guys 04:28:00 - Mizkif plays Jump King 05:28:00 - asianbunny comes over 11:50:00 - Mizkif plays Mario Kart 8 12:28:00 - Back to Minecraft Hardcore 14:27:00 - Mizkif calls MaryMaybe 14:39:00 - Mizkif plays Only Up! 15:39:00 - MaryMaybe comes over 18:33:00 - Reacting to Clips 18:43:00 - Back to Only Up! 19:19:00 - MaryMaybe returns 21:34:00 - Marbles with MaryMayne 22:40:00 - Fandy shows up 23:01:00 - Clara shows up 23:12:00 - Clara and Fandy react to ClipQueue 25:45:00 - Clara and Fandy react to the Top 100 Spotify Songs of All Time 27:06:00 - Island joins for the Pokemon Box Opening 29:02:00 - Mizkif, MaryMaybe, Clara and Fandy react to ClipQueue 30:09:00 - Mizkif calls Roflgator 30:24:00 - Clara and Fandy play Vampire Survivors (also Esfand shows up) 30:53:00 - Clara and Fandy play Map Quiz Game 31:19:00 - Esfand, Clara and Fandy play Balatro 32:29:00 - Mizkif names 100 Women 32:34:00 - Wake shows up 34:07:00 - Back to the craft 35:00:00 - Mizkif reacts to everybody else reacting to Playboi Carti not dropping his music on time 35:39:00 - One Hour of Wall Staring with Clara and Bunky (with the occasional Esfand entertainment) 37:36:00 - We are playing Minecraft again 39:05:00 - Mizkif reacts to the new Playboi Carti album 40:32:00 - Minecraft. Again. 40:46:00 - Morning Shift with MaryMaybe Third VOD: 35:00 - Mizkif and MaryMaybe play Pokemon Stadium 39:00 - Mizkif and MaryMaybe play Nintendo Switch Sports 46:00 - Juliakins shows up 2:13:00 - Sardaco shows up 3:29:00 - Simply shows up 3:51:00 - Going over Unban Requests 4:45:00 - Fireball Friday starts 4:59:00 - Mario Party with Simply, Sardaco and MaryMaybe 5:15:00 - AXIUN shows up 5:22:00 - Alinity shows up 5:30:00 - Knut and family show up 5:34:00 - asianbunny shows up 5:54:00 - Fandy and Bryce show up 6:04:00 - Jacob shows up 6:09:00 - Corey shows up 6:14:00 - Cam shows up 6:26:00 - Russel shows up 6:35:00 - ‘Goodbye Austin House’ Party Starts 6:42:00 - Wake shows up 6:47:00 - ExtraEmily shows up 6:48:00 - Aruuu, Poke and Clara show up 6:50:00 - The Final Goodbye to the House

That’s the end of the Austin House season! Mizkif has flown away in a Klat Airlines helicopter…. to a new journey, a new chapter!❤️ Here are the (3) VOD links and Timestamps: First VOD: 00:00/29:00 - Intro with all the old intro songs 29:00 - Start of Stream 30:00 - Reacting to TopKek 41:30 - Reacting to NEW TopKek (The Streamer Awards - Twitch Fails) 52:00 - Reacting to Erobb and Jason Second VOD: Starts Right Away with Super Mario 64 01:59:00 - Mizkif plays Minecraft Hardcore 03:46:00 - Mizkif plays Fall Guys 04:28:00 - Mizkif plays Jump King 05:28:00 - asianbunny comes over 11:50:00 - Mizkif plays Mario Kart 8 12:28:00 - Back to Minecraft Hardcore 14:27:00 - Mizkif calls MaryMaybe 14:39:00 - Mizkif plays Only Up! 15:39:00 - MaryMaybe comes over 18:33:00 - Reacting to Clips 18:43:00 - Back to Only Up! 19:19:00 - MaryMaybe returns 21:34:00 - Marbles with MaryMayne 22:40:00 - Fandy shows up 23:01:00 - Clara shows up 23:12:00 - Clara and Fandy react to ClipQueue 25:45:00 - Clara and Fandy react to the Top 100 Spotify Songs of All Time 27:06:00 - Island joins for the Pokemon Box Opening 29:02:00 - Mizkif, MaryMaybe, Clara and Fandy react to ClipQueue 30:09:00 - Mizkif calls Roflgator 30:24:00 - Clara and Fandy play Vampire Survivors (also Esfand shows up) 30:53:00 - Clara and Fandy play Map Quiz Game 31:19:00 - Esfand, Clara and Fandy play Balatro 32:29:00 - Mizkif names 100 Women 32:34:00 - Wake shows up 34:07:00 - Back to the craft 35:00:00 - Mizkif reacts to everybody else reacting to Playboi Carti not dropping his music on time 35:39:00 - One Hour of Wall Staring with Clara and Bunky (with the occasional Esfand entertainment) 37:36:00 - We are playing Minecraft again 39:05:00 - Mizkif reacts to the new Playboi Carti album 40:32:00 - Minecraft. Again. 40:46:00 - Morning Shift with MaryMaybe Third VOD: 35:00 - Mizkif and MaryMaybe play Pokemon Stadium 39:00 - Mizkif and MaryMaybe play Nintendo Switch Sports 46:00 - Juliakins shows up 2:13:00 - Sardaco shows up 3:29:00 - Simply shows up 3:51:00 - Going over Unban Requests 4:45:00 - Fireball Friday starts 4:59:00 - Mario Party with Simply, Sardaco and MaryMaybe 5:15:00 - AXIUN shows up 5:22:00 - Alinity shows up 5:30:00 - Knut and family show up 5:34:00 - asianbunny shows up 5:54:00 - Fandy and Bryce show up 6:04:00 - Jacob shows up 6:09:00 - Corey shows up 6:14:00 - Cam shows up 6:26:00 - Russel shows up 6:35:00 - ‘Goodbye Austin House’ Party Starts 6:42:00 - Wake shows up 6:47:00 - ExtraEmily shows up 6:48:00 - Aruuu, Poke and Clara show up 6:50:00 - The Final Goodbye to the House

Mizkif Updates

188,834 görüntüleme • 1 yıl önce

The Scott Alexander & Daniel Kokotajlo episode. Scott and Daniel break down every month from now until the 2027 intelligence explosion. Misaligned hive minds, Xi and Trump waking up, automated Ilyas accelerating AI progress. I went in quite skeptical. But I learned a tremendous amount by bouncing different objections against them. Links below. Enjoy! Timestamps: 00:00:00 - AI 2027 00:07:45 - Forecasting 2025 and 2026 00:15:30 - Why LLMs aren't making discoveries 00:25:22 - Debating intelligence explosion 00:50:34 - Can superintelligence actually transform science? 01:17:43 - Cultural evolution vs superintelligence 01:24:54 - Mid-2027 branch point 01:33:19 - Race with China 01:45:36 - Nationalization vs private anarchy 02:04:11 - Misalignment 02:15:41 - UBI, AI advisors, & human future 02:23:49 - Factory farming for digital minds 02:27:41 - Daniel leaving OpenAI 02:36:04 - Scott's blogging advice

The Scott Alexander & Daniel Kokotajlo episode. Scott and Daniel break down every month from now until the 2027 intelligence explosion. Misaligned hive minds, Xi and Trump waking up, automated Ilyas accelerating AI progress. I went in quite skeptical. But I learned a tremendous amount by bouncing different objections against them. Links below. Enjoy! Timestamps: 00:00:00 - AI 2027 00:07:45 - Forecasting 2025 and 2026 00:15:30 - Why LLMs aren't making discoveries 00:25:22 - Debating intelligence explosion 00:50:34 - Can superintelligence actually transform science? 01:17:43 - Cultural evolution vs superintelligence 01:24:54 - Mid-2027 branch point 01:33:19 - Race with China 01:45:36 - Nationalization vs private anarchy 02:04:11 - Misalignment 02:15:41 - UBI, AI advisors, & human future 02:23:49 - Factory farming for digital minds 02:27:41 - Daniel leaving OpenAI 02:36:04 - Scott's blogging advice

Dwarkesh Patel

362,019 görüntüleme • 1 yıl önce