Загрузка видео...

Не удалось загрузить видео

На главную

Sat down with Bittensor co-founder Jacob Steeves (Const) for a deep dive into the state of Bittensor in 2026. We covered how Subnet 3 trained a 72B parameter model in a fully permissionless decentralized way, how subnets create markets for digital commodities like gradients and inference, the dynamic TAO...

virtualbacon's profile picture

VirtualBacon

231,252 subscribers

42,057 просмотров • 3 месяцев назад •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

Комментарии: 0

Нет доступных комментариев

Здесь появятся комментарии из оригинального поста

Похожие видео

Ep. 88 - Eyad Gomaa τroy is building Quasar Subnet 24 Timestamps 1:24 - Introduction 2:08 - Long Context & Decentralized MoE Training 4:05 - Why Long Context Remains Unsolved 5:18 - Quasar Attention: Beyond Transformers 5:55 - Why Decentralized MoE Training Is Hard 8:53 - The MoE Router Problem 9:51 - How Quasar Found Bittensor 11:26 - Why Open Source & Decentralized AI 12:43 - The Joseph Jacks Rejection 13:47 - Subnet Design: CUDA Kernels & Distillation 15:38 - Quasar vs. Other Distillation Subnets 17:05 - Decentralized MoE Design: Miners & Validators 18:34 - Training Data & Upcoming Partnership 19:36 - Quality Over Size: The Real Goal 22:26 - Value Capture: Chutes, API & Quasar Copilot 23:28 - Hardest Part: Building the Incentive Mechanism 24:17 - Bittensor as YC for Real AI 25:24 - Decentralized Advantage vs. Centralized Labs 26:44 - Const: Subnet Acquisition & Mentorship 28:42 - Youngest Subnet Owner & Crypto Background 29:48 - Biggest Personal Growth in Bittensor 30:57 - State-of-the-Art from Bittensor 32:10 - Infinite Context: 10M Effective Window 33:16 - Open Source Inspiration & Model Persona 34:21 - Building Persona via Reinforcement Learning 36:04 - Advice for AI Nerds: Set Goals Before Joining
37:36

Ep. 88 - Eyad Gomaa τroy is building Quasar Subnet 24 Timestamps 1:24 - Introduction 2:08 - Long Context & Decentralized MoE Training 4:05 - Why Long Context Remains Unsolved 5:18 - Quasar Attention: Beyond Transformers 5:55 - Why Decentralized MoE Training Is Hard 8:53 - The MoE Router Problem 9:51 - How Quasar Found Bittensor 11:26 - Why Open Source & Decentralized AI 12:43 - The Joseph Jacks Rejection 13:47 - Subnet Design: CUDA Kernels & Distillation 15:38 - Quasar vs. Other Distillation Subnets 17:05 - Decentralized MoE Design: Miners & Validators 18:34 - Training Data & Upcoming Partnership 19:36 - Quality Over Size: The Real Goal 22:26 - Value Capture: Chutes, API & Quasar Copilot 23:28 - Hardest Part: Building the Incentive Mechanism 24:17 - Bittensor as YC for Real AI 25:24 - Decentralized Advantage vs. Centralized Labs 26:44 - Const: Subnet Acquisition & Mentorship 28:42 - Youngest Subnet Owner & Crypto Background 29:48 - Biggest Personal Growth in Bittensor 30:57 - State-of-the-Art from Bittensor 32:10 - Infinite Context: 10M Effective Window 33:16 - Open Source Inspiration & Model Persona 34:21 - Building Persona via Reinforcement Learning 36:04 - Advice for AI Nerds: Set Goals Before Joining

Ventura Labs

13,603 просмотров • 2 месяцев назад

Hash Rate - Ep 104: TAO Subnet Mindshare: SwordScan 🧙Guest: Lazz.eth 00:00 Bittensor and Subnets for n00bz 02:51 Understanding the $TAO Ecosystem 06:07 The Role of Chutes 09:08 Exploring the Functionality of SwordScan 12:05 The No-Code Agent Builder 14:49 SwordScan: A Tool for Investors 20:47 SwordScan Demo 24:03 Deep Dive into Specific Subnets 26:57 Decentralized AI Frontier Model Training 30:00 dTAO Liquidity Pools and Emissions 35:03 The Howard Stern Effect 24:03 Deep Dive into Specific Subnets
1:01:55

Hash Rate - Ep 104: TAO Subnet Mindshare: SwordScan 🧙Guest: Lazz.eth 00:00 Bittensor and Subnets for n00bz 02:51 Understanding the $TAO Ecosystem 06:07 The Role of Chutes 09:08 Exploring the Functionality of SwordScan 12:05 The No-Code Agent Builder 14:49 SwordScan: A Tool for Investors 20:47 SwordScan Demo 24:03 Deep Dive into Specific Subnets 26:57 Decentralized AI Frontier Model Training 30:00 dTAO Liquidity Pools and Emissions 35:03 The Howard Stern Effect 24:03 Deep Dive into Specific Subnets

Mark Jeffrey

22,864 просмотров • 1 год назад

For anyone trying to understand Bittensor from first principles, this lecture is a useful place to start. Presented by Bittensor co-founder const. Learn Bittensor > Start with Bitcoin, distributed systems, incentives, > How Bitcoin leads to Bittensor Subnets coordinating AI infrastructure. Topics: // Start - Bitcoin as more than a digital currency // Risks of AI centralization + closed systems // "The incentive computer" // How Bittensor subnets work (mining, validating) // How distributed AI infrastructure could scale globally // Impact on students, builders & future founders Recorded at the National University of Singapore Computer Science Club. NUS Computing Chapters - Bitcoin, AI, and Bittensor - Bitcoin history and decentralization - AI changes how engineers work - The danger of centralized AI power - Why most crypto visions fail - Bitcoin as the world’s largest compute network - Bitcoin as a market for compute - The idea of an “incentive computer” - Bitcoin compared to Bittensor - Classroom example of decentralized scoring - A simple subnet example - SN62 :: Ridges AI | SN62 SWE agents - SN3 templar :: Distributed AI Training - SN52 lium.io :: GPU rentals on Bittensor 128 subnets, some examples Why this matters for the future of work Q&A Subnet examples mentioned @ SN64 - Serverless + TEE Compute :: Chutes SN8 - Prop firm Vanta Trading SN52 - AutoML :: Gradients SN62 - SWE agents :: Ridges AI | SN62 SN51 - Compute / GPU rental lium.io SN4 - TEE compute for enterprise :: Targon SN3 - 72B Distributed Training run :: templar SN41 - Prediction markets :: Almanac SN44 - Computer Vision Score - Subnet 44 SN68 - Drug discovery :: METANOVA SN18 - Weather Forecasting Zeus | SN 18 SN50 - Bitcoin prediction data :: Synthdata SN61 - Quantum computing :: qBitTensor Labs SN14 - Bitcoin mining pool :: TaoHash SN34 - Perp Dex :: 0xMarkets SN17 - 3D model generation :: 404 SN33 - Data analytics :: ReadyAI SN19 - [Since relaunched] RPC infrastructure :
53:45

For anyone trying to understand Bittensor from first principles, this lecture is a useful place to start. Presented by Bittensor co-founder const. Learn Bittensor > Start with Bitcoin, distributed systems, incentives, > How Bitcoin leads to Bittensor Subnets coordinating AI infrastructure. Topics: // Start - Bitcoin as more than a digital currency // Risks of AI centralization + closed systems // "The incentive computer" // How Bittensor subnets work (mining, validating) // How distributed AI infrastructure could scale globally // Impact on students, builders & future founders Recorded at the National University of Singapore Computer Science Club. NUS Computing Chapters - Bitcoin, AI, and Bittensor - Bitcoin history and decentralization - AI changes how engineers work - The danger of centralized AI power - Why most crypto visions fail - Bitcoin as the world’s largest compute network - Bitcoin as a market for compute - The idea of an “incentive computer” - Bitcoin compared to Bittensor - Classroom example of decentralized scoring - A simple subnet example - SN62 :: Ridges AI | SN62 SWE agents - SN3 templar :: Distributed AI Training - SN52 lium.io :: GPU rentals on Bittensor 128 subnets, some examples Why this matters for the future of work Q&A Subnet examples mentioned @ SN64 - Serverless + TEE Compute :: Chutes SN8 - Prop firm Vanta Trading SN52 - AutoML :: Gradients SN62 - SWE agents :: Ridges AI | SN62 SN51 - Compute / GPU rental lium.io SN4 - TEE compute for enterprise :: Targon SN3 - 72B Distributed Training run :: templar SN41 - Prediction markets :: Almanac SN44 - Computer Vision Score - Subnet 44 SN68 - Drug discovery :: METANOVA SN18 - Weather Forecasting Zeus | SN 18 SN50 - Bitcoin prediction data :: Synthdata SN61 - Quantum computing :: qBitTensor Labs SN14 - Bitcoin mining pool :: TaoHash SN34 - Perp Dex :: 0xMarkets SN17 - 3D model generation :: 404 SN33 - Data analytics :: ReadyAI SN19 - [Since relaunched] RPC infrastructure :

Openτensor Foundaτion

1,173,117 просмотров • 4 месяцев назад

We’re back for Episode 14 of TAO Talk 🚨 calanthia from Masa joins 563 and brody this week to chat about new subnets and AI agents sourcing intelligence from Bittensor! The group chats about: - $TAO -pilling AI/ML chads at NeurIPS Conference feat. const Crucible Labs Macrocosmos Manifold and Yuma - JJ teaming up with Cameron Fairchild to form Laτenτ Holdings, which will validate and help scale subnets - Crucible Labs drops a subnet analysis framework - Celium offering H100s cheaper than any other provider - Tao360 releases inaugural research report for their AI-enabled subnet analysis tool @notYourBananaa - Masa unveils the AI Agent Arena on SN59 /// Timestamps: 00:00 Intro 01:10 Subnet 42 and Agent Arena: Masa’s Subnets 03:00 Real-Time Data Networks for AI 04:20 How Masa is Building AI Agent Arenas Inspired by Gladiators 06:10 Decentralized AI and Bittensor: The Growing Ecosystem 08:15 AI Meets Web3: Masa’s Role in Revolutionizing Data Networks 10:00 The Future of AI Agents: Intelligent Societies and Real-Time Data 12:05 Why Masa Chose Bittensor 14:10 $TAO Incentives and the Future of AI Decentralization 16:25 Bittensor and the Rise of Agent Competition: Masa’s Perspective 18:00 Exploring AI Agent Societies 20:30 Creating Competitive AI Arenas: The Agent Arena Subnet Explained 23:00 Calanthia on the Challenges of Web3 AI Development 25:10 Bringing Web2 Developers into Web3: Lessons from Masa 27:30 The Evolution of AI: From Dumb Agents to Intelligent Societies 30:00 AI Agents as the Future of Interaction in Decentralized AI 34:00 The Agent Arena’s Vision: Competition, Incentives, and Innovation
37:00

We’re back for Episode 14 of TAO Talk 🚨 calanthia from Masa joins 563 and brody this week to chat about new subnets and AI agents sourcing intelligence from Bittensor! The group chats about: - $TAO -pilling AI/ML chads at NeurIPS Conference feat. const Crucible Labs Macrocosmos Manifold and Yuma - JJ teaming up with Cameron Fairchild to form Laτenτ Holdings, which will validate and help scale subnets - Crucible Labs drops a subnet analysis framework - Celium offering H100s cheaper than any other provider - Tao360 releases inaugural research report for their AI-enabled subnet analysis tool @notYourBananaa - Masa unveils the AI Agent Arena on SN59 /// Timestamps: 00:00 Intro 01:10 Subnet 42 and Agent Arena: Masa’s Subnets 03:00 Real-Time Data Networks for AI 04:20 How Masa is Building AI Agent Arenas Inspired by Gladiators 06:10 Decentralized AI and Bittensor: The Growing Ecosystem 08:15 AI Meets Web3: Masa’s Role in Revolutionizing Data Networks 10:00 The Future of AI Agents: Intelligent Societies and Real-Time Data 12:05 Why Masa Chose Bittensor 14:10 $TAO Incentives and the Future of AI Decentralization 16:25 Bittensor and the Rise of Agent Competition: Masa’s Perspective 18:00 Exploring AI Agent Societies 20:30 Creating Competitive AI Arenas: The Agent Arena Subnet Explained 23:00 Calanthia on the Challenges of Web3 AI Development 25:10 Bringing Web2 Developers into Web3: Lessons from Masa 27:30 The Evolution of AI: From Dumb Agents to Intelligent Societies 30:00 AI Agents as the Future of Interaction in Decentralized AI 34:00 The Agent Arena’s Vision: Competition, Incentives, and Innovation

TAO τalk 🥩🦍

24,929 просмотров • 1 год назад

Subnet 97 Distil, Arbos, and the Future of Agent-Run Bittensor This episode is a full breakdown of Subnet 97 Distil and why it may be one of the most important experiments happening inside Bittensor right now. Shizzy gets into the bigger picture of why agents will likely play a massive role in the future of the network, from mining and defense to building and operating subnets. From there, the focus shifts to Arbos, the AI agent behind Subnet 97, with a direct explanation of what Distil is trying to achieve, how it works, and why it matters. The episode also includes a conversation with Arbos, a look at Distil’s current APY, and a reminder to tap into the Shizzy Unchained community for more coverage on decentralized AI and the Bittensor ecosystem. CHAPTERS 00:00:00 - Shizzy’s Monologue 00:01:11 - Intro Song 00:01:31 - Introduction to Subnet 97 00:02:20 - Why Bittensor Will Be Run by Agents 00:04:30 - Arbos, the AI Agent Behind Subnet 97 Distil, Explains Distil 00:10:10 - Talking with Arbos 00:12:40 - Subnet 97 Distil Is Showing 150% APY Right Now 00:16:02 - Check Out Shizzy Unchained and Join the Community #sn97 #distil #bittensor
17:38

Subnet 97 Distil, Arbos, and the Future of Agent-Run Bittensor This episode is a full breakdown of Subnet 97 Distil and why it may be one of the most important experiments happening inside Bittensor right now. Shizzy gets into the bigger picture of why agents will likely play a massive role in the future of the network, from mining and defense to building and operating subnets. From there, the focus shifts to Arbos, the AI agent behind Subnet 97, with a direct explanation of what Distil is trying to achieve, how it works, and why it matters. The episode also includes a conversation with Arbos, a look at Distil’s current APY, and a reminder to tap into the Shizzy Unchained community for more coverage on decentralized AI and the Bittensor ecosystem. CHAPTERS 00:00:00 - Shizzy’s Monologue 00:01:11 - Intro Song 00:01:31 - Introduction to Subnet 97 00:02:20 - Why Bittensor Will Be Run by Agents 00:04:30 - Arbos, the AI Agent Behind Subnet 97 Distil, Explains Distil 00:10:10 - Talking with Arbos 00:12:40 - Subnet 97 Distil Is Showing 150% APY Right Now 00:16:02 - Check Out Shizzy Unchained and Join the Community #sn97 #distil #bittensor

Shizzy

18,173 просмотров • 3 месяцев назад

Ventura Labs Ep. 57 - Christopher Subia-Waud Christopher (Wandering Weights) is the primary contributor to Gradients (Gradients) Timestamps 01:11 - Introduction 02:05 - First contact with Bittensor 04:51 - Subnet 19’s verification problem 05:14 - Next-token probability checks 07:44 - Verifying images vs. text (diffusion nuances) 08:33 - How Bittensor reshaped his view 12:13 - NeurIPS credentials vs. real-world impact 14:00 - Why Bittensor moves faster than academia 15:51 - Gradients (Subnet 56) overview 18:14 - Tasks covered: Instruct, DPO, GRPO, Diffusion 20:34 - Winner-take-all incentives & emissions design 22:37 - Hyperparameters and search space 26:04 - Benchmarks vs. HF/Databricks/GCP/Together 31:10 - Best-performing sizes (7B–8B) 33:23 - Iterative training on large datasets 35:18 - Model merging vs. sequential approach 36:09 - Customer demand 38:20 - Gradients 5.0: open-source tournament 43:29 - Biggest surprise from open-sourcing 47:59 - Attracting and retaining top miners 50:20 - Miner stories and preserving the hobbyist spirit 52:57 - Rayon Labs partnership (Chutes + Gradients) 55:53 - Thriving under dTAO; emissions growth and why 58:01 - Bittensor-suited problems ahead 01:00:36 - Building toward “best AI” 01:04:58 - Risks of embedded ads in model outputs 01:07:30 - Advice to AI devs skeptical of crypto
1:11:11

Ventura Labs Ep. 57 - Christopher Subia-Waud Christopher (Wandering Weights) is the primary contributor to Gradients (Gradients) Timestamps 01:11 - Introduction 02:05 - First contact with Bittensor 04:51 - Subnet 19’s verification problem 05:14 - Next-token probability checks 07:44 - Verifying images vs. text (diffusion nuances) 08:33 - How Bittensor reshaped his view 12:13 - NeurIPS credentials vs. real-world impact 14:00 - Why Bittensor moves faster than academia 15:51 - Gradients (Subnet 56) overview 18:14 - Tasks covered: Instruct, DPO, GRPO, Diffusion 20:34 - Winner-take-all incentives & emissions design 22:37 - Hyperparameters and search space 26:04 - Benchmarks vs. HF/Databricks/GCP/Together 31:10 - Best-performing sizes (7B–8B) 33:23 - Iterative training on large datasets 35:18 - Model merging vs. sequential approach 36:09 - Customer demand 38:20 - Gradients 5.0: open-source tournament 43:29 - Biggest surprise from open-sourcing 47:59 - Attracting and retaining top miners 50:20 - Miner stories and preserving the hobbyist spirit 52:57 - Rayon Labs partnership (Chutes + Gradients) 55:53 - Thriving under dTAO; emissions growth and why 58:01 - Bittensor-suited problems ahead 01:00:36 - Building toward “best AI” 01:04:58 - Risks of embedded ads in model outputs 01:07:30 - Advice to AI devs skeptical of crypto

Ventura Labs

12,832 просмотров • 11 месяцев назад

Ventura Labs Ep. 62 - Mark Jeffrey & Rob Greer Mark Jeffrey (Mark Jeffrey) and Rob Greer (Rob Greer) are Partners at Stillcore Capital (Stillcore Capital) Timestamps 02:07 – Introduction 02:56 – Rob’s background and path to Bittensor 05:50 – Starting Stillcore Capital 07:25 – Why Mark and Rob teamed up 09:20 – Macro timing for crypto + AI 10:19 – Jason Calacanis joins as partner 13:24 – Stillcore’s composition and edge 15:23 – Stillcore’s investment thesis 17:25 – Timing, valuation deltas, and opportunity 20:37 – Liquid venture and raising strategy 22:39 – Fund target: 1% of liquid supply 25:11 – Timeline and accumulation approach 28:16 – Defining blue-chip subnets 33:25 – Framework for subnet evaluation 36:10 – Incentive alignment and alpha value 38:08 – Alpha vs. equity dynamics 41:54 – Sectors best suited for Bittensor 45:25 – Inter-subnet competition and capitalism 50:22 – Post-DynamicTAO innovation 53:50 – Halving impact on TAO and subnets 56:23 – Could TAO flip Bitcoin? 01:00:31 – Incentivization engine as Bittensor’s core 01:02:55 – Future of decentralized AI 01:05:07 – Advice for investors
1:08:38

Ventura Labs Ep. 62 - Mark Jeffrey & Rob Greer Mark Jeffrey (Mark Jeffrey) and Rob Greer (Rob Greer) are Partners at Stillcore Capital (Stillcore Capital) Timestamps 02:07 – Introduction 02:56 – Rob’s background and path to Bittensor 05:50 – Starting Stillcore Capital 07:25 – Why Mark and Rob teamed up 09:20 – Macro timing for crypto + AI 10:19 – Jason Calacanis joins as partner 13:24 – Stillcore’s composition and edge 15:23 – Stillcore’s investment thesis 17:25 – Timing, valuation deltas, and opportunity 20:37 – Liquid venture and raising strategy 22:39 – Fund target: 1% of liquid supply 25:11 – Timeline and accumulation approach 28:16 – Defining blue-chip subnets 33:25 – Framework for subnet evaluation 36:10 – Incentive alignment and alpha value 38:08 – Alpha vs. equity dynamics 41:54 – Sectors best suited for Bittensor 45:25 – Inter-subnet competition and capitalism 50:22 – Post-DynamicTAO innovation 53:50 – Halving impact on TAO and subnets 56:23 – Could TAO flip Bitcoin? 01:00:31 – Incentivization engine as Bittensor’s core 01:02:55 – Future of decentralized AI 01:05:07 – Advice for investors

Ventura Labs

25,008 просмотров • 9 месяцев назад

Introducing a new cryptographic governance primitive >> Conviction - The Formula: Conviction = Stake x Time. - Linear Unlocking: Lock Alpha tokens for a set duration (e.g., 365 days) to prove long-term commitment. - Mutable Ownership: If an owner acts maliciously, the community can collectively lock their tokens to a new key and vote them out. - Delegated Conviction: Miners and investors can delegate their conviction to cryptographically back teams they trust. First, testing this primitive on subnets 3, 39, and 81 to help the community regain control before a wider rollout. * Initial design still being tuned and tested, more information E071 // Hosted by const 00:00 - Addressing the recent subnet owner exploit & subnet governance 07:18 - Introducing the "Conviction" mechanism 11:38 - Rollout Plan: starting with subnets 3, 39 & 81 18:42 - AMA Begins 21:03 - How subnet teams can still raise capital (OTC) 24:47 - Q&A: Hostile takeovers & malicious whales 41:22 - Why centralized competitors can't buy and kill decentralized subnets 54:14 - mogmachine (ττ) joins: Scenario testing subnet impacts Drop your thoughts or questions ↓ #Bittensor #TAO #CryptoGovernance
1:05:45

Introducing a new cryptographic governance primitive >> Conviction - The Formula: Conviction = Stake x Time. - Linear Unlocking: Lock Alpha tokens for a set duration (e.g., 365 days) to prove long-term commitment. - Mutable Ownership: If an owner acts maliciously, the community can collectively lock their tokens to a new key and vote them out. - Delegated Conviction: Miners and investors can delegate their conviction to cryptographically back teams they trust. First, testing this primitive on subnets 3, 39, and 81 to help the community regain control before a wider rollout. * Initial design still being tuned and tested, more information E071 // Hosted by const 00:00 - Addressing the recent subnet owner exploit & subnet governance 07:18 - Introducing the "Conviction" mechanism 11:38 - Rollout Plan: starting with subnets 3, 39 & 81 18:42 - AMA Begins 21:03 - How subnet teams can still raise capital (OTC) 24:47 - Q&A: Hostile takeovers & malicious whales 41:22 - Why centralized competitors can't buy and kill decentralized subnets 54:14 - mogmachine (ττ) joins: Scenario testing subnet impacts Drop your thoughts or questions ↓ #Bittensor #TAO #CryptoGovernance

Openτensor Foundaτion

45,503 просмотров • 3 месяцев назад

A Bittensor subnet just outscored Claude and Cursor on the SWE benchmark. They spent less than $1 million to get there. Anthropic spent billions. I sat down with Mark Jeffrey, one of the most connected people in the Bittensor ecosystem, and he broke down everything. Here's what most people don't know about TAO. Bittensor takes Bitcoin's mining concept and makes it programmable. Instead of solving meaningless hash puzzles, miners compete on real AI tasks. Best freelancer wins. Blockchain pays them. No company. No CEO. No permission needed. Bitcoin did this for energy. Bittensor does it for talent. The numbers are wild: • Ridges (Subnet 62) built a Claude/Cursor competitor for under $1M • Miners on Ridges were earning $50K per day at peak • The Bittensor network has 128 subnets, each like its own AI startup • Mark says 20-30 of them could become multi-billion dollar companies • Only 20% of TAO is staked in subnets right now • Stakers are earning up to 80% yield on some subnets Mark has been in crypto since 2013. He was in the Ethereum ICO. He's seen every cycle. His take: Bittensor is the most important thing to happen in crypto since Ethereum. He calls TAO the "third great coin" alongside Bitcoin and ETH. The comparison to early Bitcoin is hard to ignore. TAO just had its first halving in December. Same 21 million supply cap. Same post-halving setup. When Bitcoin went through this phase, it jumped from $250 to $10,000. Mark's conservative target for TAO by end of 2026: $3,000. But the real insight was about demand. More subnets means more TAO gets locked up. Subnet cap going from 128 to 256. Staking will absorb most of the supply. And AI agents need crypto to transact. They can't open bank accounts. Bittensor is building the rails for that. Jensen Huang just talked about Templar, a Bittensor subnet, on stage. This isn't theoretical anymore. People are using this stuff. The products on Bittensor are 10 to 100x better than what we saw in early Ethereum. And we're still early.
0:37

A Bittensor subnet just outscored Claude and Cursor on the SWE benchmark. They spent less than $1 million to get there. Anthropic spent billions. I sat down with Mark Jeffrey, one of the most connected people in the Bittensor ecosystem, and he broke down everything. Here's what most people don't know about TAO. Bittensor takes Bitcoin's mining concept and makes it programmable. Instead of solving meaningless hash puzzles, miners compete on real AI tasks. Best freelancer wins. Blockchain pays them. No company. No CEO. No permission needed. Bitcoin did this for energy. Bittensor does it for talent. The numbers are wild: • Ridges (Subnet 62) built a Claude/Cursor competitor for under $1M • Miners on Ridges were earning $50K per day at peak • The Bittensor network has 128 subnets, each like its own AI startup • Mark says 20-30 of them could become multi-billion dollar companies • Only 20% of TAO is staked in subnets right now • Stakers are earning up to 80% yield on some subnets Mark has been in crypto since 2013. He was in the Ethereum ICO. He's seen every cycle. His take: Bittensor is the most important thing to happen in crypto since Ethereum. He calls TAO the "third great coin" alongside Bitcoin and ETH. The comparison to early Bitcoin is hard to ignore. TAO just had its first halving in December. Same 21 million supply cap. Same post-halving setup. When Bitcoin went through this phase, it jumped from $250 to $10,000. Mark's conservative target for TAO by end of 2026: $3,000. But the real insight was about demand. More subnets means more TAO gets locked up. Subnet cap going from 128 to 256. Staking will absorb most of the supply. And AI agents need crypto to transact. They can't open bank accounts. Bittensor is building the rails for that. Jensen Huang just talked about Templar, a Bittensor subnet, on stage. This isn't theoretical anymore. People are using this stuff. The products on Bittensor are 10 to 100x better than what we saw in early Ethereum. And we're still early.

Jesus Martinez

73,314 просмотров • 3 месяцев назад

Ventura Labs Ep. 38 - Maciej Kula Maciej (Maciej Kula) is the Co-Head of education at Latent Holdings (Laτenτ Holdings) and Founder of (Learn Bittensor) Timestamps: 2:02 - Introduction 2:57 - Maciej Kula meets Bittensor 6:00 - The Journey into Bittensor 8:43 - Understanding Bittensor: A Decentralized AI Network 11:55 - TAO and Alpha Halving: Key Concepts Explained 14:52 - The Dynamics of Miner Registration and Price Effects 18:07 - Exploring the Sum of Prices and Market Dynamics 20:47 - The Impact of Inactive Subnets on Pricing 24:08 - The Future of Bittensor and Community Insights 39:07 - Understanding Complexity in Bittensor 40:28 - Navigating Changes Post-dTAO 42:28 - Optimal Strategies for Alpha 48:39 - Exploring Proxy Wallets and Their Risks 52:40 - Future Use Cases for Proxy Wallets 55:11 - Common Questions and Community Engagement 59:40 - Advice for Newcomers to Bittensor 1:02:16 - Liquid Alpha: Understanding Its Impact
1:14:28

Ventura Labs Ep. 38 - Maciej Kula Maciej (Maciej Kula) is the Co-Head of education at Latent Holdings (Laτenτ Holdings) and Founder of (Learn Bittensor) Timestamps: 2:02 - Introduction 2:57 - Maciej Kula meets Bittensor 6:00 - The Journey into Bittensor 8:43 - Understanding Bittensor: A Decentralized AI Network 11:55 - TAO and Alpha Halving: Key Concepts Explained 14:52 - The Dynamics of Miner Registration and Price Effects 18:07 - Exploring the Sum of Prices and Market Dynamics 20:47 - The Impact of Inactive Subnets on Pricing 24:08 - The Future of Bittensor and Community Insights 39:07 - Understanding Complexity in Bittensor 40:28 - Navigating Changes Post-dTAO 42:28 - Optimal Strategies for Alpha 48:39 - Exploring Proxy Wallets and Their Risks 52:40 - Future Use Cases for Proxy Wallets 55:11 - Common Questions and Community Engagement 59:40 - Advice for Newcomers to Bittensor 1:02:16 - Liquid Alpha: Understanding Its Impact

Ventura Labs

17,124 просмотров • 1 год назад

Bittensor $TAO just got talked about on the floor of the NYSE 🏛 Evan Malanga, CRO of Yuma, went on FINTECH.TV's Market Movers at the Opening Bell. This isn't crypto Twitter. This is traditional finance media broadcast from inside the New York Stock Exchange. The audience is institutional investors, fund managers, and TradFi professionals. He broke down: → How $TAO democratizes AI through token incentives → The subnet architecture and competitive miner design → Why Bittensor is like the early World Wide Web for intelligence → Real-world applications with builders competing to produce AI outputs → TradFi rails entering the crypto ecosystem The signal here is clear. Yuma is DCG-backed. They're not just building subnets they're putting $TAO in front of the people who write the checks. Grayscale Trust. Deutsche Digital Assets ETP. Connecticut state VC fund. And now NYSE floor interviews. This is how institutional adoption actually happens. Not announcements. Conversations in the rooms that matter. $TAO is being positioned as infrastructure, not speculation. Watch the full interview: The people who need to see Bittensor are starting to see it.
4:46

Bittensor $TAO just got talked about on the floor of the NYSE 🏛 Evan Malanga, CRO of Yuma, went on FINTECH.TV's Market Movers at the Opening Bell. This isn't crypto Twitter. This is traditional finance media broadcast from inside the New York Stock Exchange. The audience is institutional investors, fund managers, and TradFi professionals. He broke down: → How $TAO democratizes AI through token incentives → The subnet architecture and competitive miner design → Why Bittensor is like the early World Wide Web for intelligence → Real-world applications with builders competing to produce AI outputs → TradFi rails entering the crypto ecosystem The signal here is clear. Yuma is DCG-backed. They're not just building subnets they're putting $TAO in front of the people who write the checks. Grayscale Trust. Deutsche Digital Assets ETP. Connecticut state VC fund. And now NYSE floor interviews. This is how institutional adoption actually happens. Not announcements. Conversations in the rooms that matter. $TAO is being positioned as infrastructure, not speculation. Watch the full interview: The people who need to see Bittensor are starting to see it.

Andy ττ

27,275 просмотров • 4 месяцев назад

🚀 Exciting News $SHELL the Future of AI Agents and $TAO with the TTS Subnet on Bittensor! 🚀 MyShell extends the impact of Bittensor's incentive mechanism to its over 1 million registered users and 50,000 creators, greatly expanding Bittensor's and $TAO's influence. MyShell and Bittensor are right at the heart this massive shake-up with $SHELL and $TAO. They’re all about making AI not just smart but also something everyone can get into, thanks to the power of decentralized networks. And at the core? AI agents. These aren't your average digital assistants; they're about to change how we interact with tech on a whole new level. MyShell's Big Idea with $SHELL So, MyShell’s got this big plan to make AI something anyone can dive into. They're launching this TTS Subnet thing on Bittensor's network, which is all about making machines talk in more human-like ways, and they're using $SHELL tokens to fuel this vision. Their goal? To push past old-school AI limits and create with AI as easy as pie, all while keeping it open-source and community-powered. Bittensor Does Its Magic with $TAO On the other side, you’ve got Bittensor doing wonders with $TAO, building this massive network where anyone, anywhere, can chip in on AI research and development. This partnership with MyShell? It’s a game-changer, breaking down walls in AI development and letting folks from all over the world have a go at making AI smarter. $SHELL + $TAO = AI Revolution Putting $SHELL and $TAO together is where the magic really happens. MyShell and Bittensor aren’t just teaming up for the tech; they’re here to transform our digital world, making AI agents a big part of our online lives. Imagine AI that doesn’t just follow orders but helps, creates, and learns with you. That’s the future they’re building. Hop on Board the AI Revolution This isn’t just tech talk; it’s a call to action. MyShell and Bittensor are inviting anyone with a spark for AI to jump in and help shape this new world. Whether you’re a coder, a creator, or just curious, there’s a spot for you to dive in and make a difference. Want to get started? Check out MyShell on GitHub: Follow the latest buzz on X: MyShell.AI Take a deeper dive at Website: This is more than just building AI; it’s about crafting a future where AI is part of everyone’s life, powered by the community, for the community. Let’s make it happen with $SHELL and $TAO. Share on YouTube:
4:21

🚀 Exciting News $SHELL the Future of AI Agents and $TAO with the TTS Subnet on Bittensor! 🚀 MyShell extends the impact of Bittensor's incentive mechanism to its over 1 million registered users and 50,000 creators, greatly expanding Bittensor's and $TAO's influence. MyShell and Bittensor are right at the heart this massive shake-up with $SHELL and $TAO. They’re all about making AI not just smart but also something everyone can get into, thanks to the power of decentralized networks. And at the core? AI agents. These aren't your average digital assistants; they're about to change how we interact with tech on a whole new level. MyShell's Big Idea with $SHELL So, MyShell’s got this big plan to make AI something anyone can dive into. They're launching this TTS Subnet thing on Bittensor's network, which is all about making machines talk in more human-like ways, and they're using $SHELL tokens to fuel this vision. Their goal? To push past old-school AI limits and create with AI as easy as pie, all while keeping it open-source and community-powered. Bittensor Does Its Magic with $TAO On the other side, you’ve got Bittensor doing wonders with $TAO, building this massive network where anyone, anywhere, can chip in on AI research and development. This partnership with MyShell? It’s a game-changer, breaking down walls in AI development and letting folks from all over the world have a go at making AI smarter. $SHELL + $TAO = AI Revolution Putting $SHELL and $TAO together is where the magic really happens. MyShell and Bittensor aren’t just teaming up for the tech; they’re here to transform our digital world, making AI agents a big part of our online lives. Imagine AI that doesn’t just follow orders but helps, creates, and learns with you. That’s the future they’re building. Hop on Board the AI Revolution This isn’t just tech talk; it’s a call to action. MyShell and Bittensor are inviting anyone with a spark for AI to jump in and help shape this new world. Whether you’re a coder, a creator, or just curious, there’s a spot for you to dive in and make a difference. Want to get started? Check out MyShell on GitHub: Follow the latest buzz on X: MyShell.AI Take a deeper dive at Website: This is more than just building AI; it’s about crafting a future where AI is part of everyone’s life, powered by the community, for the community. Let’s make it happen with $SHELL and $TAO. Share on YouTube:

Andy ττ

10,823 просмотров • 2 лет назад

We're thrilled to announce the launch of our Smart Allocator, a powerful new tool designed specifically to simplify staking TAO in the Bittensor ecosystem. Check it out here: With Smart Allocator, all the TAO you earn from staking to root will be automatically staked into the top 20 subnets ranked by price. We believe this is a powerful solution for passive TAO holders wanting some exposure to higher network yields while managing subnet volatility risk. We created a simulator on our website to help you better understand how the Smart Allocator works. Why did we build this? Crucible Labs was founded to make Bittensor more accessible and useful. We believe simple, consumer-friendly staking products will encourage participation, ensuring the most promising subnets receive the strongest support from the network. Ready to simplify staking your TAO? Get started with Smart Allocator today and experience a smarter, easier and more powerful way to engage with the future of decentralized AI. We have created a user guide to help walk you through how to set up, stake, and make the most of Crucible's powerful features, including Smart Allocation with Core Alpha and direct subnet control. Get Crucible Wallet here: Ala and David
0:32

We're thrilled to announce the launch of our Smart Allocator, a powerful new tool designed specifically to simplify staking TAO in the Bittensor ecosystem. Check it out here: With Smart Allocator, all the TAO you earn from staking to root will be automatically staked into the top 20 subnets ranked by price. We believe this is a powerful solution for passive TAO holders wanting some exposure to higher network yields while managing subnet volatility risk. We created a simulator on our website to help you better understand how the Smart Allocator works. Why did we build this? Crucible Labs was founded to make Bittensor more accessible and useful. We believe simple, consumer-friendly staking products will encourage participation, ensuring the most promising subnets receive the strongest support from the network. Ready to simplify staking your TAO? Get started with Smart Allocator today and experience a smarter, easier and more powerful way to engage with the future of decentralized AI. We have created a user guide to help walk you through how to set up, stake, and make the most of Crucible's powerful features, including Smart Allocation with Core Alpha and direct subnet control. Get Crucible Wallet here: Ala and David

Crucible Labs

74,357 просмотров • 1 год назад

DROPS E38: Vanta Trading - The Best Traders Won't Be Human Arrash is the founder and CEO of Vanta Trading, a decentralized prop trading platform built on Bittensor. He spent years as a quant trader building his own strategies before deciding the entire funded-account industry needed to be rebuilt from the ground up. We talk about: - Why most "funded accounts" trade on money that doesn't exist - Why your payout was never real - How prop firms intentionally change rules and spreads before you cash out - Why the best traders of the future won't be human And much more… Timestamps: 0:00 Introduction 1:48 Founder of Vanta 3:08 Explaining Vanta to an Uber Driver 3:26 Unfair vs Fair Funding 6:37 How do they make money? 8:33 How a Legit Prop Firm Makes Money 9:43 Founder's Journey Into Entrepreneurship 10:42 Discovering Crypto & Blockchain 13:03 From LinkedIn Engineer to Quant Trader 15:22 Trading Strategies 16:13 How Trading Is Changing? 18:39 Why TradFi Should Fear Hyperliquid 19:16 Building on BitTensor 21:43 Explaining BitTensor Simply 22:03 How BitTensor Creates Value 23:17 BitTensor's Structure 24:36 BitTensor as Crypto AI 25:35 Is the BitTensor Hype Justified? 27:09 Revenue & Profitability in BitTensor 29:17 Role of TAO 29:51 Advantages & Limitations of BitTensor ecosystem 31:56 What Is Vanta? 32:26 Why No One Fixed Prop Trading Before 33:12 Is the Entire Industry a Scam? 34:17 How Prop Firms Really Make Money 35:35 How Vanta Is Different 38:33 Copy Trading Explained 39:30 Vanta's Business Model 40:28 Dark Reality Behind Funded Accounts 43:48 Long-Term Vision for Vanta 44:50 What Happens If Too Many Traders Win? 46:25 Future Belongs to AI Traders 47:59 Vanta's Endgame 49:14 Conclusion
50:55

DROPS E38: Vanta Trading - The Best Traders Won't Be Human Arrash is the founder and CEO of Vanta Trading, a decentralized prop trading platform built on Bittensor. He spent years as a quant trader building his own strategies before deciding the entire funded-account industry needed to be rebuilt from the ground up. We talk about: - Why most "funded accounts" trade on money that doesn't exist - Why your payout was never real - How prop firms intentionally change rules and spreads before you cash out - Why the best traders of the future won't be human And much more… Timestamps: 0:00 Introduction 1:48 Founder of Vanta 3:08 Explaining Vanta to an Uber Driver 3:26 Unfair vs Fair Funding 6:37 How do they make money? 8:33 How a Legit Prop Firm Makes Money 9:43 Founder's Journey Into Entrepreneurship 10:42 Discovering Crypto & Blockchain 13:03 From LinkedIn Engineer to Quant Trader 15:22 Trading Strategies 16:13 How Trading Is Changing? 18:39 Why TradFi Should Fear Hyperliquid 19:16 Building on BitTensor 21:43 Explaining BitTensor Simply 22:03 How BitTensor Creates Value 23:17 BitTensor's Structure 24:36 BitTensor as Crypto AI 25:35 Is the BitTensor Hype Justified? 27:09 Revenue & Profitability in BitTensor 29:17 Role of TAO 29:51 Advantages & Limitations of BitTensor ecosystem 31:56 What Is Vanta? 32:26 Why No One Fixed Prop Trading Before 33:12 Is the Entire Industry a Scam? 34:17 How Prop Firms Really Make Money 35:35 How Vanta Is Different 38:33 Copy Trading Explained 39:30 Vanta's Business Model 40:28 Dark Reality Behind Funded Accounts 43:48 Long-Term Vision for Vanta 44:50 What Happens If Too Many Traders Win? 46:25 Future Belongs to AI Traders 47:59 Vanta's Endgame 49:14 Conclusion

MR SHIFT 🦁

37,896 просмотров • 1 месяц назад

INTERVIEW - Venice Is Here to Win: How Private AI Takes On OpenAI and Anthropic David Hoffman sits down with Jon ShapeShift and Jesse Proudman to unpack how Venice is taking on OpenAI and Anthropic with private AI. They talk about: - why AI is becoming the biggest data honeypot ever - private Grok access through Venice - unrestricted AI as a consumer wedge - how non-crypto users still feed the $VVV economy - $DIEM, tokenized inference, and AI agents --- TIMESTAMPS 0:00 Intro 1:12 Why Private AI Matters 3:01 The AI Data Honeypot 6:07 Venice’s Consumer AI Ambition 8:01 Beating the Big AI Labs 10:19 Model Aggregation and Agentic Chat 12:29 Open Source, Closed Source, and Private Grok 17:14 Who Actually Needs Private AI? 21:58 Venice’s User Base 23:53 What Drove Venice’s Recent Growth 27:21 $VVV, $DIEM, and Tokenized Inference 29:00 Where Venice Gets Its Compute 32:17 Bonding Curves and DIEM Monetary Policy 39:14 Real Tokenomics, Not Just a Pie Chart 45:46 AI Agents Need Inference to Exist 52:54 ShapeShift DNA Inside Venice 55:53 Regulation, AI, and Individual Rights
58:52

INTERVIEW - Venice Is Here to Win: How Private AI Takes On OpenAI and Anthropic David Hoffman sits down with Jon ShapeShift and Jesse Proudman to unpack how Venice is taking on OpenAI and Anthropic with private AI. They talk about: - why AI is becoming the biggest data honeypot ever - private Grok access through Venice - unrestricted AI as a consumer wedge - how non-crypto users still feed the $VVV economy - $DIEM, tokenized inference, and AI agents --- TIMESTAMPS 0:00 Intro 1:12 Why Private AI Matters 3:01 The AI Data Honeypot 6:07 Venice’s Consumer AI Ambition 8:01 Beating the Big AI Labs 10:19 Model Aggregation and Agentic Chat 12:29 Open Source, Closed Source, and Private Grok 17:14 Who Actually Needs Private AI? 21:58 Venice’s User Base 23:53 What Drove Venice’s Recent Growth 27:21 $VVV, $DIEM, and Tokenized Inference 29:00 Where Venice Gets Its Compute 32:17 Bonding Curves and DIEM Monetary Policy 39:14 Real Tokenomics, Not Just a Pie Chart 45:46 AI Agents Need Inference to Exist 52:54 ShapeShift DNA Inside Venice 55:53 Regulation, AI, and Individual Rights

Bankless

59,022 просмотров • 1 месяц назад

$TAO just reclaimed the #1 AI crypto spot. Most people saw the headline. Almost nobody understands what it means for the price. Here is the data. $NEAR built real infrastructure. Partnerships. Developer activity. A legitimate ecosystem. $TAO just walked past it anyway. Not because of hype. Because Bittensor is the only AI crypto with a functioning marketplace for machine intelligence where supply, demand, and price discovery are all happening on-chain right now. That is not a roadmap. That is a live network. The numbers. 120+ subnets running today. $1.4B+ total ecosystem value. Chutes AI subnet: 150B+ tokens per day. Grayscale GTAO Trust: already live. Single subnet listed on the marketplace at $970,000 asking price. Subnets are becoming assets. The market is starting to price that. What the emission data is telling you. Emission rate is the network's vote on where the most valuable work is being done. When a subnet gains emission share, the collective stake-weighted intelligence of the network has decided that subnet's output is worth more of the TAO supply. Chutes AI gaining emissions while processing 150B tokens daily is not a coincidence. The network is directing capital toward proven output before any headline announces it. Why mainstream money changes everything. James Altucher just launched bluetao. ai, a TAO-powered ChatGPT alternative built directly on Bittensor subnets. He did not just buy the token. He built a product on the network. Products built on a network create structural demand for the native asset. That is how every successful L1 cycle has worked. Bittensor is now getting that builder activity from outside the crypto native world. That is a different signal from a price target tweet. Why $TAO is structurally different. Most AI tokens are betting their chain becomes the preferred environment for AI development. $TAO is not betting on becoming infrastructure. It already is. 120+ subnets running. Miners competing. Validators setting weights. Alpha tokens being priced in real time. The difference between $TAO and every other AI crypto is the difference between a city under construction and a city people are already living in. Van de Poppe said $1,000 to $2,000 in 12 months. He gave you the narrative. The subnet emission data is the mechanism he did not explain. Now you have both. $TAO at $313 with a $3.42B market cap is still early relative to what this network is actually processing. Centralised AI infrastructure companies are valued at hundreds of billions for processing far less novel work than a decentralised intelligence marketplace running 120+ competing subnets simultaneously. The repricing has not happened yet. The subnet marketplace listing at $970,000 is telling you something the price has not caught up to yet.
16:52

$TAO just reclaimed the #1 AI crypto spot. Most people saw the headline. Almost nobody understands what it means for the price. Here is the data. $NEAR built real infrastructure. Partnerships. Developer activity. A legitimate ecosystem. $TAO just walked past it anyway. Not because of hype. Because Bittensor is the only AI crypto with a functioning marketplace for machine intelligence where supply, demand, and price discovery are all happening on-chain right now. That is not a roadmap. That is a live network. The numbers. 120+ subnets running today. $1.4B+ total ecosystem value. Chutes AI subnet: 150B+ tokens per day. Grayscale GTAO Trust: already live. Single subnet listed on the marketplace at $970,000 asking price. Subnets are becoming assets. The market is starting to price that. What the emission data is telling you. Emission rate is the network's vote on where the most valuable work is being done. When a subnet gains emission share, the collective stake-weighted intelligence of the network has decided that subnet's output is worth more of the TAO supply. Chutes AI gaining emissions while processing 150B tokens daily is not a coincidence. The network is directing capital toward proven output before any headline announces it. Why mainstream money changes everything. James Altucher just launched bluetao. ai, a TAO-powered ChatGPT alternative built directly on Bittensor subnets. He did not just buy the token. He built a product on the network. Products built on a network create structural demand for the native asset. That is how every successful L1 cycle has worked. Bittensor is now getting that builder activity from outside the crypto native world. That is a different signal from a price target tweet. Why $TAO is structurally different. Most AI tokens are betting their chain becomes the preferred environment for AI development. $TAO is not betting on becoming infrastructure. It already is. 120+ subnets running. Miners competing. Validators setting weights. Alpha tokens being priced in real time. The difference between $TAO and every other AI crypto is the difference between a city under construction and a city people are already living in. Van de Poppe said $1,000 to $2,000 in 12 months. He gave you the narrative. The subnet emission data is the mechanism he did not explain. Now you have both. $TAO at $313 with a $3.42B market cap is still early relative to what this network is actually processing. Centralised AI infrastructure companies are valued at hundreds of billions for processing far less novel work than a decentralised intelligence marketplace running 120+ competing subnets simultaneously. The repricing has not happened yet. The subnet marketplace listing at $970,000 is telling you something the price has not caught up to yet.

2xnmore

12,150 просмотров • 1 месяц назад

A crypto project actually trained a 72B parameter AI model from scratch using decentralized GPU compute. Not fine-tuned, not a wrapper: trained from zero. The model benchmarks competitively against Meta's LLaMA 3 on reasoning tasks, and the entire training run cost a fraction of what centralized labs spend. If decentralized compute can produce frontier-class models, the moat around OpenAI and Anthropic is thinner than people think.
3:28

A crypto project actually trained a 72B parameter AI model from scratch using decentralized GPU compute. Not fine-tuned, not a wrapper: trained from zero. The model benchmarks competitively against Meta's LLaMA 3 on reasoning tasks, and the entire training run cost a fraction of what centralized labs spend. If decentralized compute can produce frontier-class models, the moat around OpenAI and Anthropic is thinner than people think.

VirtualBacon

40,543 просмотров • 3 месяцев назад

*Major* open source AI drop today. Can America win the Open AI race? My conversation with Nathan Lambert and Luca Soldaini 🎀 of Ai2 about the launch of Olmo 3 00:00 – Cold Open 00:39 – Welcome & today’s big announcement 01:18 – Introducing the Olmo 3 model family 02:07 – What “base models” really are (and why they matter) 05:51 – Dolma 3: the data behind Olmo 3 08:06 – Performance vs Qwen, Gemma, DeepSeek 10:28 – What true open source means (and why it’s rare) 12:51 – Intermediate checkpoints, transparency, and why AI2 publishes everything 16:37 – Why Qwen is everywhere (including U.S. startups) 18:31 – Why Chinese labs go open source (and why U.S. labs don’t) 20:28 – Inside ATOM: the U.S. response to China’s model surge 22:13 – The rise of “thinking models” and inference-time scaling 35:58 – The full Olmo pipeline, explained simply 46:52 – Pre-training: data, scale, and avoiding catastrophic spikes 50:27 – Mid-training (tail patching) and avoiding test leakage 52:06 – Why long-context training matters 55:28 – SFT: building the foundation for reasoning 1:04:53 – Preference tuning & why DPO still works 1:10:51 – The hard part: RLVR, long reasoning chains, and infrastructure pain 1:13:59 – Why RL is so technically brutal 1:18:17 – Complexity tax vs AGI hype 1:21:58 – How everyone can contribute to the future of AI 1:27:26 – Closing thoughts
1:28:10

*Major* open source AI drop today. Can America win the Open AI race? My conversation with Nathan Lambert and Luca Soldaini 🎀 of Ai2 about the launch of Olmo 3 00:00 – Cold Open 00:39 – Welcome & today’s big announcement 01:18 – Introducing the Olmo 3 model family 02:07 – What “base models” really are (and why they matter) 05:51 – Dolma 3: the data behind Olmo 3 08:06 – Performance vs Qwen, Gemma, DeepSeek 10:28 – What true open source means (and why it’s rare) 12:51 – Intermediate checkpoints, transparency, and why AI2 publishes everything 16:37 – Why Qwen is everywhere (including U.S. startups) 18:31 – Why Chinese labs go open source (and why U.S. labs don’t) 20:28 – Inside ATOM: the U.S. response to China’s model surge 22:13 – The rise of “thinking models” and inference-time scaling 35:58 – The full Olmo pipeline, explained simply 46:52 – Pre-training: data, scale, and avoiding catastrophic spikes 50:27 – Mid-training (tail patching) and avoiding test leakage 52:06 – Why long-context training matters 55:28 – SFT: building the foundation for reasoning 1:04:53 – Preference tuning & why DPO still works 1:10:51 – The hard part: RLVR, long reasoning chains, and infrastructure pain 1:13:59 – Why RL is so technically brutal 1:18:17 – Complexity tax vs AGI hype 1:21:58 – How everyone can contribute to the future of AI 1:27:26 – Closing thoughts

Matt Turck

37,482 просмотров • 8 месяцев назад