正在加载视频...

视频加载失败

返回首页

capability != learning new benchtalks with Parth Asawa on continual learning, where we discuss teaching models to learn from experience, measuring learning ability, the bet on parametric models, and more 01:06 What is continual learning? 04:10 Why capability and learning are different 06:13 Why build a benchmark? 08:07 Continual...

vincentsunnchen's profile picture

vincent sunn chen

1,693 subscribers

41,444 次观看 • 3 个月前 •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

3 条评论

vincent sunn chen 的头像
vincent sunn chen3 个月前

Kudos to the full Continual Learning Bench team for the work! + @chris_m_glaze @Gorlanski Benji Xu @RamyaRamakri @_asimbiswal @fredsala @matei_zaharia @profjoeyg

vincent sunn chen 的头像
vincent sunn chen3 个月前

Youtube here:

James Alcorn 的头像
James Alcorn3 个月前

@pgasawa @pgasawa a lad destined for greatness

相关视频

Intelligence and Experience are orthogonal vectors Terence Tao is perhaps the world’s smartest person, but drop him into an accounting firm or onto a construction site and on day one he’s not going to be very productive Trajectory calls this The Experience Gap, and they have a way to close it Arjun Karanam explained how at our Sovereign AI event: 00:00 Introduction 00:12 Building the platform for continual learning 01:33 The experience gap: models have IQ but no tenure 02:52 Traceability → model spec → better models and harnesses 05:27 Four wishes for the agent ecosystem 06:34 Wish 1: Trace the whole tree — and capture the corrections 08:03 Wish 2: Evals from real traffic, graded in the real harness 09:26 Wish 3: Let the agents cook, and make tool responses informative 10:34 Wish 4: Get comfortable on open weights, experiment with routers 11:51 Why owning your intelligence shouldn't be consulted away 13:15 Demo: import a benchmark, train a model, deploy it 14:28 Q&A: What's the trainable object — weights, harness, or context? 16:08 Q&A: Continual learning without training on customer data 17:13 Q&A: Episodic memory and the hierarchy of feedback 19:37 Q&A: Where continual learning matters most
21:55

Intelligence and Experience are orthogonal vectors Terence Tao is perhaps the world’s smartest person, but drop him into an accounting firm or onto a construction site and on day one he’s not going to be very productive Trajectory calls this The Experience Gap, and they have a way to close it Arjun Karanam explained how at our Sovereign AI event: 00:00 Introduction 00:12 Building the platform for continual learning 01:33 The experience gap: models have IQ but no tenure 02:52 Traceability → model spec → better models and harnesses 05:27 Four wishes for the agent ecosystem 06:34 Wish 1: Trace the whole tree — and capture the corrections 08:03 Wish 2: Evals from real traffic, graded in the real harness 09:26 Wish 3: Let the agents cook, and make tool responses informative 10:34 Wish 4: Get comfortable on open weights, experiment with routers 11:51 Why owning your intelligence shouldn't be consulted away 13:15 Demo: import a benchmark, train a model, deploy it 14:28 Q&A: What's the trainable object — weights, harness, or context? 16:08 Q&A: Continual learning without training on customer data 17:13 Q&A: Episodic memory and the hierarchy of feedback 19:37 Q&A: Where continual learning matters most

Pat Grady

77,295 次观看 • 1 个月前

Continual Learning in 2026. And AGI with it? In July, Dario said he believes Continual Learning will turn out to be not as difficult as it seems. Sholto Douglas, Anthropic and former Deepmind researcher, predicted Continual Learning gets solved in a satisfying way in 2026. Sholto said that last week. Considering how capable AI models have become in 2025, it is unwise to dismiss this as fundraising or hype. It is much wiser to assume AGI is coming. That it will change the world. And that you can benefit from it... ...than it is to assume the opposite. "Not enough people are ready for AI to not be a bubble." - roon Choose wisely my friends.
0:26

Continual Learning in 2026. And AGI with it? In July, Dario said he believes Continual Learning will turn out to be not as difficult as it seems. Sholto Douglas, Anthropic and former Deepmind researcher, predicted Continual Learning gets solved in a satisfying way in 2026. Sholto said that last week. Considering how capable AI models have become in 2025, it is unwise to dismiss this as fundraising or hype. It is much wiser to assume AGI is coming. That it will change the world. And that you can benefit from it... ...than it is to assume the opposite. "Not enough people are ready for AI to not be a bubble." - roon Choose wisely my friends.

Dan McAteer

69,496 次观看 • 9 个月前

.Richard Sutton, father of reinforcement learning, doesn’t think LLMs are bitter-lesson-pilled. My steel man of Richard’s position: we need some new architecture to enable continual (on-the-job) learning. And if we have continual learning, we don't need a special training phase - the agent just learns on-the-fly - like all humans, and indeed, like all animals. This new paradigm will render our current approach with LLMs obsolete. I did my best to represent the view that LLMs will function as the foundation on which this experiential learning can happen. Some sparks flew. 0:00:00 – Are LLMs a dead-end? 0:13:51 – Do humans do imitation learning? 0:23:57 – The Era of Experience 0:34:25 – Current architectures generalize poorly out of distribution 0:42:17 – Surprises in the AI field 0:47:28 – Will The Bitter Lesson still apply after AGI? 0:54:35 – Succession to AI
1:07:08

.Richard Sutton, father of reinforcement learning, doesn’t think LLMs are bitter-lesson-pilled. My steel man of Richard’s position: we need some new architecture to enable continual (on-the-job) learning. And if we have continual learning, we don't need a special training phase - the agent just learns on-the-fly - like all humans, and indeed, like all animals. This new paradigm will render our current approach with LLMs obsolete. I did my best to represent the view that LLMs will function as the foundation on which this experiential learning can happen. Some sparks flew. 0:00:00 – Are LLMs a dead-end? 0:13:51 – Do humans do imitation learning? 0:23:57 – The Era of Experience 0:34:25 – Current architectures generalize poorly out of distribution 0:42:17 – Surprises in the AI field 0:47:28 – Will The Bitter Lesson still apply after AGI? 0:54:35 – Succession to AI

Dwarkesh Patel

3,090,100 次观看 • 1 年前

Today we present Morpheus, a persistent enterprise simulation platform designed to make Continual Learning a reality. Morpheus is the world’s first real world Reinforcement Learning environment. Every Reinforcement Learning environment operates in the game world. Benchmarks like Atari, OpenAI Gym, MuJoCo, and Procgen are all small, game-like worlds that reset every few minutes. But the real world never resets. A business keeps running and evolving everyday. We tested how frontier LLMs would perform in realistic and dynamic business environments 🧬on Morpheus. The main conclusion was that LLMs are not continual learners. 🧵Here’s how we did it and what we learned:
1:18

Today we present Morpheus, a persistent enterprise simulation platform designed to make Continual Learning a reality. Morpheus is the world’s first real world Reinforcement Learning environment. Every Reinforcement Learning environment operates in the game world. Benchmarks like Atari, OpenAI Gym, MuJoCo, and Procgen are all small, game-like worlds that reset every few minutes. But the real world never resets. A business keeps running and evolving everyday. We tested how frontier LLMs would perform in realistic and dynamic business environments 🧬on Morpheus. The main conclusion was that LLMs are not continual learners. 🧵Here’s how we did it and what we learned:

Skyfall AI

269,020 次观看 • 2 个月前

Deep learning works extraordinarily well. And we still largely don't know why. A new paper from Jamie Simon, Daniel Kunin, and 12 co-authors argues that a scientific theory of deep learning is emerging, and coins a name for the emerging field: learning mechanics. We sat down with Jamie and Dan on Generally Intelligent to talk about what a physics of deep learning would actually look like, why now, and what's left to figure out. 3:05 Learning mechanics as the physics to mechanistic interpretability's biology 4:13 Why deep learning needs a theory 7:07 Why deep learning is uniquely hard to engineer 12:11 How a week in the woods became a paper 25:59 The barrier to theory isn't opacity, but complexity 36:26 Deep learning's first gas law 47:22 Why more particles makes the problem easier 56:22 The discretization hypothesis 1:01:50 The strongest signal that a compact theory exists 1:05:07 The Platonic Representation Hypothesis 1:15:41 Why learning mechanics and mech interp need each other 1:25:29 Theory as safety infrastructure
1:33:22

Deep learning works extraordinarily well. And we still largely don't know why. A new paper from Jamie Simon, Daniel Kunin, and 12 co-authors argues that a scientific theory of deep learning is emerging, and coins a name for the emerging field: learning mechanics. We sat down with Jamie and Dan on Generally Intelligent to talk about what a physics of deep learning would actually look like, why now, and what's left to figure out. 3:05 Learning mechanics as the physics to mechanistic interpretability's biology 4:13 Why deep learning needs a theory 7:07 Why deep learning is uniquely hard to engineer 12:11 How a week in the woods became a paper 25:59 The barrier to theory isn't opacity, but complexity 36:26 Deep learning's first gas law 47:22 Why more particles makes the problem easier 56:22 The discretization hypothesis 1:01:50 The strongest signal that a compact theory exists 1:05:07 The Platonic Representation Hypothesis 1:15:41 Why learning mechanics and mech interp need each other 1:25:29 Theory as safety infrastructure

Imbue

18,257 次观看 • 5 个月前

.River AI has a new vision for personal AI. We co-led River’s $1.1B funding round because we believe in Igor Babuschkin's ambition to build open-weight models that individuals can train and improve continuously. Watch Hemant Taneja in conversation with Igor. Chapters 02:22 — Co-Founding xAI 04:36 — An AI Experience You Own 05:59 — Betting Early on Open Weights 07:47 — How Do You Make Money on Free Models? 09:25 — Why the US Needs Its Own Open Models 11:27 — Enterprises Want to Control Their AI 13:36 — What Is Continual Learning? 16:56 — An API That Updates the Weights 19:17 — A Personal AI for Everyone 20:46 — The 20-Minute Training Run
22:16

.River AI has a new vision for personal AI. We co-led River’s $1.1B funding round because we believe in Igor Babuschkin's ambition to build open-weight models that individuals can train and improve continuously. Watch Hemant Taneja in conversation with Igor. Chapters 02:22 — Co-Founding xAI 04:36 — An AI Experience You Own 05:59 — Betting Early on Open Weights 07:47 — How Do You Make Money on Free Models? 09:25 — Why the US Needs Its Own Open Models 11:27 — Enterprises Want to Control Their AI 13:36 — What Is Continual Learning? 16:56 — An API That Updates the Weights 19:17 — A Personal AI for Everyone 20:46 — The 20-Minute Training Run

General Catalyst

112,510 次观看 • 1 个月前

On the latest episode of our podcast, research lead Irina and host Hannah Fry discuss the exciting potential of AI tutors – like our LearnLM Learning Coach on YouTube – to personalize learning and support teachers. 🧠 Watch the full episode now ↓ Timestamps: 00:06 Intro 01:37 Breaking into education 04:08 AI tutor 08:12 The future 10:38 Irina's journey 13:53 Reasoning and teaching 15:05 The optimal strategy for teaching 17:54 Good pedagogy 21:15 GenAI for Education 23:50 AI for history 27:55 Arizona State University 29:43 Self disclosure and anthropomorphism 34:45 Cheating and student engagement 37:00 Dependency on AI 39:42 Research results 41:25 Reflections from Hannah
42:53

On the latest episode of our podcast, research lead Irina and host Hannah Fry discuss the exciting potential of AI tutors – like our LearnLM Learning Coach on YouTube – to personalize learning and support teachers. 🧠 Watch the full episode now ↓ Timestamps: 00:06 Intro 01:37 Breaking into education 04:08 AI tutor 08:12 The future 10:38 Irina's journey 13:53 Reasoning and teaching 15:05 The optimal strategy for teaching 17:54 Good pedagogy 21:15 GenAI for Education 23:50 AI for history 27:55 Arizona State University 29:43 Self disclosure and anthropomorphism 34:45 Cheating and student engagement 37:00 Dependency on AI 39:42 Research results 41:25 Reflections from Hannah

Google DeepMind

127,284 次观看 • 1 年前

Our new podcast on evals, with Max Niederman, Ege Erdil, and Stephen Yang. 0:00:00 – What's an eval, and how's it different from an RL environment? 0:19:33 – Why are models bad at building an emulator when the task is fully verifiable? 0:42:00 – How does training on bad data teach models to write terrible code? 1:04:00 – Why is continual learning still so bad? 2:25:24 – Why haven't software engineers been replaced when coding is basically solved? Listen to the Mechanize Podcast on YouTube, Spotify, etc. Enjoy!
2:42:03

Our new podcast on evals, with Max Niederman, Ege Erdil, and Stephen Yang. 0:00:00 – What's an eval, and how's it different from an RL environment? 0:19:33 – Why are models bad at building an emulator when the task is fully verifiable? 0:42:00 – How does training on bad data teach models to write terrible code? 1:04:00 – Why is continual learning still so bad? 2:25:24 – Why haven't software engineers been replaced when coding is basically solved? Listen to the Mechanize Podcast on YouTube, Spotify, etc. Enjoy!

Mechanize

40,871 次观看 • 3 个月前

I filmed a video version of my post 'Why I Don’t Think AGI Is Right Around The Corner' so I could show it to my YouTube audience. You can watch it too I guess. 0:00:00 - Continual Learning 0:08:06 - Computer Use 0:11:27 - Reasoning 0:12:45 - So what are my predictions?
17:13

I filmed a video version of my post 'Why I Don’t Think AGI Is Right Around The Corner' so I could show it to my YouTube audience. You can watch it too I guess. 0:00:00 - Continual Learning 0:08:06 - Computer Use 0:11:27 - Reasoning 0:12:45 - So what are my predictions?

Dwarkesh Patel

113,781 次观看 • 1 年前

🌄 Beyond SWE: The Future of Long Horizon Environments A discussion with our founders about KellyBench, and the need for new environments that require agents to adapt over time and act under uncertainty. 0:00:17 What is KellyBench? 0:02:10 Openendedness, non-stationarity and continual learning 0:03:40 Analytical versus operational capabilities 0:04:13 Why are models bad at KellyBench? 0:05:39 Situational awareness in dynamic environments 0:06:37 Feature stability and real-world non-stationarity 0:07:07 The power of context 0:07:34 "The first principle is that you must not fool yourself" 0:08:20 Machiavelli, fortuna and the ability to adapt to change 0:09:23 How can models improve on evals like KellyBench? 0:10:12 Limitations of KellyBench: data availability and market odds timing 0:11:44 Implications beyond quant finance / sports betting 0:13:26 Civilisations as the ultimate time horizon 0:14:05 Would a mega prompt / better elicitation do much better on the benchmark? 0:14:52 What new types of capability is GR excited about? 0:17:48 Taste and the ability to pursue long-term goals even if they aren't immediately rewarding 0:18:56 Deep learning as an example of a method that took a long time to bear fruit 0:19:25 Optimism about the future of AI
19:46

🌄 Beyond SWE: The Future of Long Horizon Environments A discussion with our founders about KellyBench, and the need for new environments that require agents to adapt over time and act under uncertainty. 0:00:17 What is KellyBench? 0:02:10 Openendedness, non-stationarity and continual learning 0:03:40 Analytical versus operational capabilities 0:04:13 Why are models bad at KellyBench? 0:05:39 Situational awareness in dynamic environments 0:06:37 Feature stability and real-world non-stationarity 0:07:07 The power of context 0:07:34 "The first principle is that you must not fool yourself" 0:08:20 Machiavelli, fortuna and the ability to adapt to change 0:09:23 How can models improve on evals like KellyBench? 0:10:12 Limitations of KellyBench: data availability and market odds timing 0:11:44 Implications beyond quant finance / sports betting 0:13:26 Civilisations as the ultimate time horizon 0:14:05 Would a mega prompt / better elicitation do much better on the benchmark? 0:14:52 What new types of capability is GR excited about? 0:17:48 Taste and the ability to pursue long-term goals even if they aren't immediately rewarding 0:18:56 Deep learning as an example of a method that took a long time to bear fruit 0:19:25 Optimism about the future of AI

General Reasoning

15,585 次观看 • 5 个月前

On agent factories, memory, continual learning & building on the frontier while growing a family In conversation with Sam Whitmore Sam Whitmore, eng at Cursor this is a special one! dive in: 00:25 Sam’s background 01:38 Family building while in frontier tech 11:50 Agent factories at cursor 15:46 Continual learning, agent self-awareness, memory, and user goal alignment 23:51 Building a thoughtful product culture at Cursor 27:00 Thinking about defensibility in 2026 30:01 Selling outside of the tech/X bubble 33:49 Kids & AI use 35:12 On reading fiction 38:32 On pregnancy, hormones, and building 42:58 On agent predictions for the next 6 months 46:35 Hiring at Cursor
47:58

On agent factories, memory, continual learning & building on the frontier while growing a family In conversation with Sam Whitmore Sam Whitmore, eng at Cursor this is a special one! dive in: 00:25 Sam’s background 01:38 Family building while in frontier tech 11:50 Agent factories at cursor 15:46 Continual learning, agent self-awareness, memory, and user goal alignment 23:51 Building a thoughtful product culture at Cursor 27:00 Thinking about defensibility in 2026 30:01 Selling outside of the tech/X bubble 33:49 Kids & AI use 35:12 On reading fiction 38:32 On pregnancy, hormones, and building 42:58 On agent predictions for the next 6 months 46:35 Hiring at Cursor

ashe

60,599 次观看 • 5 个月前

Today we're announcing Base Labs, a dedicated research organization focused on advancing open-source AI. We believe in a healthy, open frontier model ecosystem. To enable this, we are working on: - Blue-sky research on continual learning, the science of RL, and how models learn across their full lifecycle, with every experiment and recipe shared openly. - The BaseHub Data Foundry: the highest-quality open RL environments, training data, and real-world benchmarks, built for anyone to train and benchmark on. - Post-post training: taking open-source models and making them better, safer, and more aligned through continual post-training, built on our research, and deployed with our frontier safety stack so organizations can use open-source models with confidence. - Making models cheaper and more performant through our model performance research. This is a mission-driven research effort, not a commercial product. We believe the health of the open-source AI ecosystem matters and that the best way to advance it is to do science in the open. We’re hiring engineers, researchers, and research fellows to advance this mission.
1:05

Today we're announcing Base Labs, a dedicated research organization focused on advancing open-source AI. We believe in a healthy, open frontier model ecosystem. To enable this, we are working on: - Blue-sky research on continual learning, the science of RL, and how models learn across their full lifecycle, with every experiment and recipe shared openly. - The BaseHub Data Foundry: the highest-quality open RL environments, training data, and real-world benchmarks, built for anyone to train and benchmark on. - Post-post training: taking open-source models and making them better, safer, and more aligned through continual post-training, built on our research, and deployed with our frontier safety stack so organizations can use open-source models with confidence. - Making models cheaper and more performant through our model performance research. This is a mission-driven research effort, not a commercial product. We believe the health of the open-source AI ecosystem matters and that the best way to advance it is to do science in the open. We’re hiring engineers, researchers, and research fellows to advance this mission.

Base Labs

241,481 次观看 • 28 天前

Continual learning sometimes gets discussed as if the goal is to dissolve the context/weights distinction. Let the model just keep accumulating, fine-tuning itself on the fly. Andrej Karpathy points out, though, that this isn't how humans do it. Our working memory gets wiped regularly. What we actually have is a consolidation process (sleep) that distills stuff into the brain, in a weird and lossy way. This is very different from how people sometimes talk about continual learning. It's not obvious it's something you can get for free from doing long enough RL loops.
3:21

Continual learning sometimes gets discussed as if the goal is to dissolve the context/weights distinction. Let the model just keep accumulating, fine-tuning itself on the fly. Andrej Karpathy points out, though, that this isn't how humans do it. Our working memory gets wiped regularly. What we actually have is a consolidation process (sleep) that distills stuff into the brain, in a weird and lossy way. This is very different from how people sometimes talk about continual learning. It's not obvious it's something you can get for free from doing long enough RL loops.

Dwarkesh Patel

60,477 次观看 • 4 个月前

This morning at NeurIPS, Rich Sutton reminded us that we need continual learning to reach AGI. This afternoon, Ali Behrouz presented a Google poster paper, Nested Learning, which provides new ideas on the path to continual learning. I recorded the 40 minute talk as it might be useful for some researchers in the audience. For the rest of us, I subscribe to Andrej Karpathy's suspicion that it will take a 5-10 papers like this to move us to AGI from where we are now, just like it took about 10 papers to move from 2012's AlexNet to ChatGPT. At the very end, I ask Ali how far along to continual learning this represents. Full paper link below, as well as a YouTube link. ps. sorry about the first 2 minutes of bad audio since there were 2 idiots standing beside me have a conversation right in front of this presenter in a rather packed poster presentation. Honestly, tamp down your egos guys and show come common courtesy!
40:27

This morning at NeurIPS, Rich Sutton reminded us that we need continual learning to reach AGI. This afternoon, Ali Behrouz presented a Google poster paper, Nested Learning, which provides new ideas on the path to continual learning. I recorded the 40 minute talk as it might be useful for some researchers in the audience. For the rest of us, I subscribe to Andrej Karpathy's suspicion that it will take a 5-10 papers like this to move us to AGI from where we are now, just like it took about 10 papers to move from 2012's AlexNet to ChatGPT. At the very end, I ask Ali how far along to continual learning this represents. Full paper link below, as well as a YouTube link. ps. sorry about the first 2 minutes of bad audio since there were 2 idiots standing beside me have a conversation right in front of this presenter in a rather packed poster presentation. Honestly, tamp down your egos guys and show come common courtesy!

Phil Trubey

235,997 次观看 • 10 个月前

Why AI Can Now Make Discoveries - my conversation with Dan Roberts, Lead of the Foundations of Reinforcement Learning team at OpenAI 00:00 Intro: AI's wild week in mathematics 01:21 What OpenAI's Foundations of RL team does 03:08 Dan's journey: from black holes and quantum gravity to frontier AI 07:04 Are AI systems becoming useful for real science 08:21 The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic 08:52 Why the OpenAI result was an act of exploration 10:25 OpenAI vs. DeepMind: informal reasoning vs. formal proof 12:13 RL 101: learning by doing, not just watching 15:10 Why reinforcement learning works 15:58 How RL breaks: sparse feedback and long-horizon tasks 17:03 RLHF: how human feedback shaped early language models 18:48 Move 37, self-play, and the search for novel strategies 22:16 Explore vs. exploit in scientific discovery 24:49 Why RL may now be "the cake," not the cherry on top 25:46 Why RL started working with large language models 27:29 Is RL "sucking supervision through a straw"? 28:47 Why language may be the grounding layer for intelligence 31:46 A contrarian take on the Bitter Lesson 32:41 What test-time compute actually is 34:50 How RL gives models the ability to think 35:40 Verifiable rewards, math, coding, and the messy real world 38:00 What physics can teach us about AI 42:08 Is there a thermodynamics of AI? 43:08 From Erdős problems to Einstein-level AI 45:16 Is AI already doing original science? 45:51 How far are we from AI automating AI research 47:41 Why Dan is excited about the future of science
49:06

Why AI Can Now Make Discoveries - my conversation with Dan Roberts, Lead of the Foundations of Reinforcement Learning team at OpenAI 00:00 Intro: AI's wild week in mathematics 01:21 What OpenAI's Foundations of RL team does 03:08 Dan's journey: from black holes and quantum gravity to frontier AI 07:04 Are AI systems becoming useful for real science 08:21 The AI math moment: Erdős, OpenAI, DeepMind, and Anthropic 08:52 Why the OpenAI result was an act of exploration 10:25 OpenAI vs. DeepMind: informal reasoning vs. formal proof 12:13 RL 101: learning by doing, not just watching 15:10 Why reinforcement learning works 15:58 How RL breaks: sparse feedback and long-horizon tasks 17:03 RLHF: how human feedback shaped early language models 18:48 Move 37, self-play, and the search for novel strategies 22:16 Explore vs. exploit in scientific discovery 24:49 Why RL may now be "the cake," not the cherry on top 25:46 Why RL started working with large language models 27:29 Is RL "sucking supervision through a straw"? 28:47 Why language may be the grounding layer for intelligence 31:46 A contrarian take on the Bitter Lesson 32:41 What test-time compute actually is 34:50 How RL gives models the ability to think 35:40 Verifiable rewards, math, coding, and the messy real world 38:00 What physics can teach us about AI 42:08 Is there a thermodynamics of AI? 43:08 From Erdős problems to Einstein-level AI 45:16 Is AI already doing original science? 45:51 How far are we from AI automating AI research 47:41 Why Dan is excited about the future of science

Matt Turck

69,801 次观看 • 3 个月前

Jensen Huang on "distillation" On his new interview with axios, he was asked this question "Should open source model companies be allowed to distill closed models" "Distillation—learning from AI, learning from other people, and learning from other sources of knowledge, is fundamental to intelligence. We are constantly learning from other people. I am learning from you through the questions you are asking, and you are learning from me. All day long, we are learning from one another. AI also has to learn from something. The original AI models, whether they were open or closed, were trained on previously created knowledge from the internet. Now, AI is generating more content than humans. In a few more years, the internet could be 99% AI-generated content, and that content will have been created by some form of AI. As a result, AI systems will constantly be distilling knowledge and intelligence from other AI systems. The fact that AI can learn is a good thing. We want AI systems to be intelligent because a smarter AI can also be a safer AI." ---- From "Axios" YouTube channel, (full video link in comment)
1:07

Jensen Huang on "distillation" On his new interview with axios, he was asked this question "Should open source model companies be allowed to distill closed models" "Distillation—learning from AI, learning from other people, and learning from other sources of knowledge, is fundamental to intelligence. We are constantly learning from other people. I am learning from you through the questions you are asking, and you are learning from me. All day long, we are learning from one another. AI also has to learn from something. The original AI models, whether they were open or closed, were trained on previously created knowledge from the internet. Now, AI is generating more content than humans. In a few more years, the internet could be 99% AI-generated content, and that content will have been created by some form of AI. As a result, AI systems will constantly be distilling knowledge and intelligence from other AI systems. The fact that AI can learn is a good thing. We want AI systems to be intelligent because a smarter AI can also be a safer AI." ---- From "Axios" YouTube channel, (full video link in comment)

Rohan Paul

443,986 次观看 • 2 个月前

Google DeepMind say true AGI will “reason, adapt, and learn continuously,” and they estimate it’s about 5-10 years away. Meta’s chief scientist Yann LeCun likewise views continual learning as a key pillar of human level AI and is optimistic about achieving it by around 2030 Currently, large language models only improve via offline updates (fine-tuning or reinforcement learning from human feedback) rather than meta learning or modular networks that update themselves gradually. Thankfully All major research teams (OpenAI, DeepMind, Anthropic, Meta, etc.) are actively working on these directions, aiming for models that continuously learn from mistakes as they interact a capability they believe will emerge as we approach true AGI later this decade 2029
1:07

Google DeepMind say true AGI will “reason, adapt, and learn continuously,” and they estimate it’s about 5-10 years away. Meta’s chief scientist Yann LeCun likewise views continual learning as a key pillar of human level AI and is optimistic about achieving it by around 2030 Currently, large language models only improve via offline updates (fine-tuning or reinforcement learning from human feedback) rather than meta learning or modular networks that update themselves gradually. Thankfully All major research teams (OpenAI, DeepMind, Anthropic, Meta, etc.) are actively working on these directions, aiming for models that continuously learn from mistakes as they interact a capability they believe will emerge as we approach true AGI later this decade 2029

Chris

284,222 次观看 • 1 年前

Why AI Progress Suddenly Feels Real - my conversation with Yann Dubois, who co-leads the Post-Training Frontiers team at OpenAI 00:00 - Intro 01:30 - Why recent AI progress feels like a step function 04:13 - Model reliability & the emotional rollercoaster of shipping GPT-5.5 07:33 - How OpenAI structures vertical and horizontal teams 09:49 - Improving model efficiency and test-time compute 12:32 - Yann's journey from Switzerland to OpenAI 15:37 - Reasoning in 2026: Real-world utility vs verifiable rewards 18:34 - GPT-5.5 Thinking vs Pro: Scaling test-time compute 20:09 - How reasoning models become more efficient 23:23 - Pre-training scaling and overcoming the data wall 27:03 - Multimodal data, synthetic data, and embodied AI 31:05 - Demystifying mid-training and post-training 37:21 - Does RL create new capabilities in AI? 38:53 - The challenges and frontier of scaling RL 43:09 - Is building AI models a craft or a strict science 48:21 - How AI models generalize across different domains 54:18 - How reinforcement learning cures AI hallucinations 56:04 - Negative generalization and conflicting instructions 58:05 - Can RL scale to law, medicine, and the broader economy? 1:00:19 - The evaluation bottleneck and Model as a Judge 1:04:21 - Continuous AI progress & continual learning 1:08:49 - Will foundation models eat the agent harness 1:11:23 - Why startups should focus on the last mile of AI
1:13:56

Why AI Progress Suddenly Feels Real - my conversation with Yann Dubois, who co-leads the Post-Training Frontiers team at OpenAI 00:00 - Intro 01:30 - Why recent AI progress feels like a step function 04:13 - Model reliability & the emotional rollercoaster of shipping GPT-5.5 07:33 - How OpenAI structures vertical and horizontal teams 09:49 - Improving model efficiency and test-time compute 12:32 - Yann's journey from Switzerland to OpenAI 15:37 - Reasoning in 2026: Real-world utility vs verifiable rewards 18:34 - GPT-5.5 Thinking vs Pro: Scaling test-time compute 20:09 - How reasoning models become more efficient 23:23 - Pre-training scaling and overcoming the data wall 27:03 - Multimodal data, synthetic data, and embodied AI 31:05 - Demystifying mid-training and post-training 37:21 - Does RL create new capabilities in AI? 38:53 - The challenges and frontier of scaling RL 43:09 - Is building AI models a craft or a strict science 48:21 - How AI models generalize across different domains 54:18 - How reinforcement learning cures AI hallucinations 56:04 - Negative generalization and conflicting instructions 58:05 - Can RL scale to law, medicine, and the broader economy? 1:00:19 - The evaluation bottleneck and Model as a Judge 1:04:21 - Continuous AI progress & continual learning 1:08:49 - Will foundation models eat the agent harness 1:11:23 - Why startups should focus on the last mile of AI

Matt Turck

101,314 次观看 • 4 个月前

Today we release my favorite episode of Training Data yet: the great Rich Sutton. Richard Sutton wrote the textbook, wrote The Bitter Lesson (and many other on-point essays like "Self-Verification, The Key to AI"), and trained a mafia of talented students who went on to change the AI landscape forever including David Silver, inventor of built AlphaGo. Khurram Javed was Rich's PhD student at Alberta and wrote The Big World Hypothesis. They just left academia to start Oak Lab Their core argument: (1) The Bitter Lesson: the world is massively more complex than any model of it, so anything trained on human-curated data has a ceiling (2) Continual Learning: intelligence is continual by definition, and today's models stop learning the moment they ship. The conversation covers: — what The Bitter Lesson actually says, and what people get wrong — why synthetic data is "just a big mistake," and the Big World Hypothesis behind it — how LLMs are both a positive and a negative example of his own essay — why no animal learns by supervised learning, and what squirrels can do that we can't — the cure for catastrophic forgetting: per-weight step sizes and continual backprop — why the biggest labs can't take a path where performance gets worse before it gets better — a trillion parameters on 20 watts, and the Moore's Law math that makes it plausible — why the endpoint isn't one mind but one design, running as many minds It was both a fun generative idea- and debate-filled conversation, and a surprisingly human one too. Rich, thank you for beating cancer and changing the trajectory of AI. 💙 00:00 Introduction 02:10 An AI winter, a cancer diagnosis, and the move to Alberta 07:07 Writing "The Bitter Lesson," and what people get wrong 09:53 Are LLMs a positive or a negative example of it? 11:03 Synthetic data is "just a big mistake," and the Big World Hypothesis 18:01 AlphaGo, human priors, and why prior knowledge and learning should be friends 22:37 "Their weights never change": do LLM assistants actually learn? 26:09 Babies, squirrels, and why no animal learns by supervised learning 32:02 Rockets, imagination, and where paradigm shifts come from 36:42 The Alberta Plan and its 12 steps 38:53 Catastrophic forgetting and the cure 43:43 Oak's biggest ambition: a self-maintaining mind 47:56 Why the big labs are stuck in a local minimum 49:13 If everything goes right: LLMs, many minds, and hiring The man who pioneered reinforcement learning thinks the rest of the field is weird, and lays it all out in today's episode. Together w/ Alfred Lin Sequoia Capital
53:43

Today we release my favorite episode of Training Data yet: the great Rich Sutton. Richard Sutton wrote the textbook, wrote The Bitter Lesson (and many other on-point essays like "Self-Verification, The Key to AI"), and trained a mafia of talented students who went on to change the AI landscape forever including David Silver, inventor of built AlphaGo. Khurram Javed was Rich's PhD student at Alberta and wrote The Big World Hypothesis. They just left academia to start Oak Lab Their core argument: (1) The Bitter Lesson: the world is massively more complex than any model of it, so anything trained on human-curated data has a ceiling (2) Continual Learning: intelligence is continual by definition, and today's models stop learning the moment they ship. The conversation covers: — what The Bitter Lesson actually says, and what people get wrong — why synthetic data is "just a big mistake," and the Big World Hypothesis behind it — how LLMs are both a positive and a negative example of his own essay — why no animal learns by supervised learning, and what squirrels can do that we can't — the cure for catastrophic forgetting: per-weight step sizes and continual backprop — why the biggest labs can't take a path where performance gets worse before it gets better — a trillion parameters on 20 watts, and the Moore's Law math that makes it plausible — why the endpoint isn't one mind but one design, running as many minds It was both a fun generative idea- and debate-filled conversation, and a surprisingly human one too. Rich, thank you for beating cancer and changing the trajectory of AI. 💙 00:00 Introduction 02:10 An AI winter, a cancer diagnosis, and the move to Alberta 07:07 Writing "The Bitter Lesson," and what people get wrong 09:53 Are LLMs a positive or a negative example of it? 11:03 Synthetic data is "just a big mistake," and the Big World Hypothesis 18:01 AlphaGo, human priors, and why prior knowledge and learning should be friends 22:37 "Their weights never change": do LLM assistants actually learn? 26:09 Babies, squirrels, and why no animal learns by supervised learning 32:02 Rockets, imagination, and where paradigm shifts come from 36:42 The Alberta Plan and its 12 steps 38:53 Catastrophic forgetting and the cure 43:43 Oak's biggest ambition: a self-maintaining mind 47:56 Why the big labs are stuck in a local minimum 49:13 If everything goes right: LLMs, many minds, and hiring The man who pioneered reinforcement learning thinks the rest of the field is weird, and lays it all out in today's episode. Together w/ Alfred Lin Sequoia Capital

Sonya Huang 🐥

119,665 次观看 • 1 个月前

New Benchtalks with John Yang: on ProgramBench (0% frontier models at launch) and the lineage/future of coding benchmarks, from SWE-bench/InterCode to now 01:29 ProgramBench launch and reception 03:41 Why artifact-level evaluation, not code-level 06:03 Why models love Python 08:29 ProgramBench as a research tool 12:45 From SWE-bench & InterCode to ProgramBench 17:47 How to grade a coding model 21:53 The position paper & humans in the loop 25:01 Managing quality with agents-in-the-loop 28:40 Internet access and benchmark integrity 35:26 Where models may surpass human abilities 38:56 When a model hits 80% on ProgramBench 43:55 Benchmarks worth paying attention to 46:24 What benchmark do you wish existed 49:32 Will benchmarks still look like benchmarks in 5 years 52:02 How to contribute to ProgramBench
54:13

New Benchtalks with John Yang: on ProgramBench (0% frontier models at launch) and the lineage/future of coding benchmarks, from SWE-bench/InterCode to now 01:29 ProgramBench launch and reception 03:41 Why artifact-level evaluation, not code-level 06:03 Why models love Python 08:29 ProgramBench as a research tool 12:45 From SWE-bench & InterCode to ProgramBench 17:47 How to grade a coding model 21:53 The position paper & humans in the loop 25:01 Managing quality with agents-in-the-loop 28:40 Internet access and benchmark integrity 35:26 Where models may surpass human abilities 38:56 When a model hits 80% on ProgramBench 43:55 Benchmarks worth paying attention to 46:24 What benchmark do you wish existed 49:32 Will benchmarks still look like benchmarks in 5 years 52:02 How to contribute to ProgramBench

vincent sunn chen

27,008 次观看 • 4 个月前