正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Jensen just said he was reading a text from DeepMind’s Demis confirming it: pre-training + post-training scaling laws are fully intact, and Gemini 3 is the proof. A huge jump in quality exactly what Blackwell and Rubin were built for. This next curve is going to be wild. $NVDA

Ben Pouladian

11,285 subscribers

327,245 次观看 • 8 个月前 •via X (Twitter)

新闻政治科学技术

Anya Rossi• Live Now

Private livecam show

0 条评论

暂无评论

原始帖子的评论将显示在这里

相关视频

🚨Anthropic CEO was just asked: “Is pre-training dead?” > Claude 4 embodies advances in both pre-training and post-training > pre-training scaling laws work the way they’ve worked before > we are going to continue to scale up > the timeline is accelerating > 1-2 years until these models become peers to us Good. Accelerate.

🚨Anthropic CEO was just asked: “Is pre-training dead?” > Claude 4 embodies advances in both pre-training and post-training > pre-training scaling laws work the way they’ve worked before > we are going to continue to scale up > the timeline is accelerating > 1-2 years until these models become peers to us Good. Accelerate.

NIK

61,775 次观看 • 1 年前

Frontier models that use reasoning to "think" during inference are generating 5X more AI tokens per year. ✨ "Inference is now a thinking process. And in order to teach AI how to think, reinforcement learning and very significant computation was introduced into post-training," said NVIDIA CEO Jensen Huang in a recent keynote at #CES2026. Reinforcement learning is increasing computation demands across all AI scaling laws: pre-training, post-training, and test-time scaling. Learn more about reinforcement learning and AI scaling ➡️

Frontier models that use reasoning to "think" during inference are generating 5X more AI tokens per year. ✨ "Inference is now a thinking process. And in order to teach AI how to think, reinforcement learning and very significant computation was introduced into post-training," said NVIDIA CEO Jensen Huang in a recent keynote at #CES2026. Reinforcement learning is increasing computation demands across all AI scaling laws: pre-training, post-training, and test-time scaling. Learn more about reinforcement learning and AI scaling ➡️

NVIDIA AI Infrastructure

13,329 次观看 • 6 个月前

Watch Nvidia CEO Jensen Huang reveal what he described as the state of the art in AI hardware: the new Rubin computing architecture. “Vera Rubin is designed to address this fundamental challenge that we have: The amount of computation necessary for AI is skyrocketing.” Huang said. “Today, I can tell you that Vera Rubin is in full production.” According to Nvidia’s tests, Rubin architecture operates three and a half times faster than the previous Blackwell architecture on model-training tasks and five times faster on inference tasks, and that's just the start.

Watch Nvidia CEO Jensen Huang reveal what he described as the state of the art in AI hardware: the new Rubin computing architecture. “Vera Rubin is designed to address this fundamental challenge that we have: The amount of computation necessary for AI is skyrocketing.” Huang said. “Today, I can tell you that Vera Rubin is in full production.” According to Nvidia’s tests, Rubin architecture operates three and a half times faster than the previous Blackwell architecture on model-training tasks and five times faster on inference tasks, and that's just the start.

TechCrunch

76,412 次观看 • 6 个月前

Google's Jeff Dean says current pre-training is passive: initialize a model, stream the internet past it, let it observe But models need to learn not just from data, but by acting, predicting, and choosing what to learn from next "we have this artificial distinction now between pre and post-training, and it shouldn't exist long term"

Google's Jeff Dean says current pre-training is passive: initialize a model, stream the internet past it, let it observe But models need to learn not just from data, but by acting, predicting, and choosing what to learn from next "we have this artificial distinction now between pre and post-training, and it shouldn't exist long term"

Haider.

55,694 次观看 • 3 个月前

🚨 JUST NOW: JENSEN HINTED ANTHROPIC’S NEXT DEAL AND ITS LIKELY $NBIS On the $NVDA earnings call, Jensen said they’ve secured compute for Anthropic across Azure, AWS and $CRWV then added there’s a “whole list of others” they are bringing online. $NBIS is an NVIDIA Cloud Partner with early access to Vera Rubin, already deploying NVL72 across the US and Europe from H2 2026. The partnerships are in place. The infrastructure is being built. Expect a deal soon. If Jensen is flagging more capacity providers coming online for Anthropic, Nebius has everything in place to be that call.

🚨 JUST NOW: JENSEN HINTED ANTHROPIC’S NEXT DEAL AND ITS LIKELY $NBIS On the $NVDA earnings call, Jensen said they’ve secured compute for Anthropic across Azure, AWS and $CRWV then added there’s a “whole list of others” they are bringing online. $NBIS is an NVIDIA Cloud Partner with early access to Vera Rubin, already deploying NVL72 across the US and Europe from H2 2026. The partnerships are in place. The infrastructure is being built. Expect a deal soon. If Jensen is flagging more capacity providers coming online for Anthropic, Nebius has everything in place to be that call.

Jordan

156,482 次观看 • 2 个月前

“We think we have a competitive advantage.” - Demis Hassibis, Head of Google DeepMind / AI efforts While Anthropic and OpenAI seek to win via training on “text” (written language), Demis Hassabis says Google Gemini is attempting a different path: Multimodal. Gemini aims to be the most effective model in the real world: physics, robotics, voice, visual, sensory, & text. The bet is multimodal unlocks more of what humanity really cares about.

“We think we have a competitive advantage.” - Demis Hassibis, Head of Google DeepMind / AI efforts While Anthropic and OpenAI seek to win via training on “text” (written language), Demis Hassabis says Google Gemini is attempting a different path: Multimodal. Gemini aims to be the most effective model in the real world: physics, robotics, voice, visual, sensory, & text. The bet is multimodal unlocks more of what humanity really cares about.

Compound248 💰

17,731 次观看 • 2 个月前

Gemini 3, scaling laws and the 'finite data' era: my conversation with Sebastian Borgeaud, research engineer at Google DeepMind and a pre-training lead for Gemini 3 00:00 – Cold intro: “We’re ahead of schedule” + AI is now a system 00:58 – Oriol Vinyals's “secret recipe”: better pre- + post-training 02:09 – Why AI progress still isn’t slowing down 03:04 – Are models actually getting smarter? 04:36 – Two–three years out: what changes first? 06:34 – AI doing AI research: faster, not automated 07:45 – Frontier labs: same playbook or different bets? 10:19 – Post-transformers: will a disruption happen? 10:51 – DeepMind’s advantage: research × engineering × infra 12:26 – What a Gemini 3 pre-training lead actually does 13:59 – From Europe to Cambridge to DeepMind 18:06 – Why he left RL for real-world data 20:05 – From Gopher to Chinchilla to RETRO (and why it matters) 20:28 – “Research taste”: integrate or slow everyone down 23:00 – Fixes vs moonshots: how they balance the pipeline 24:37 – Research vs product pressure (and org structure) 26:24 – Gemini 3 under the hood: MoE in plain English 28:30 – Native multimodality: the hidden costs 30:03 – Scaling laws aren’t dead (but scale isn’t everything) 33:07 – Synthetic data: powerful, dangerous? 35:00 – Reasoning traces: what he can’t say (and why) 37:18 – Long context + attention: what’s next 38:40 – Retrieval vs RAG vs long context 41:49 – The real boss fight: evals (and contamination) 42:28 – Alignment: pre-training vs post-training 43:32 – Deep Think + agents + “vibe coding” 46:34 – Continual learning: updating models over time 49:35 – Advice for researchers + founders 53:35 – “No end in sight” for progress + closing

Gemini 3, scaling laws and the 'finite data' era: my conversation with Sebastian Borgeaud, research engineer at Google DeepMind and a pre-training lead for Gemini 3 00:00 – Cold intro: “We’re ahead of schedule” + AI is now a system 00:58 – Oriol Vinyals's “secret recipe”: better pre- + post-training 02:09 – Why AI progress still isn’t slowing down 03:04 – Are models actually getting smarter? 04:36 – Two–three years out: what changes first? 06:34 – AI doing AI research: faster, not automated 07:45 – Frontier labs: same playbook or different bets? 10:19 – Post-transformers: will a disruption happen? 10:51 – DeepMind’s advantage: research × engineering × infra 12:26 – What a Gemini 3 pre-training lead actually does 13:59 – From Europe to Cambridge to DeepMind 18:06 – Why he left RL for real-world data 20:05 – From Gopher to Chinchilla to RETRO (and why it matters) 20:28 – “Research taste”: integrate or slow everyone down 23:00 – Fixes vs moonshots: how they balance the pipeline 24:37 – Research vs product pressure (and org structure) 26:24 – Gemini 3 under the hood: MoE in plain English 28:30 – Native multimodality: the hidden costs 30:03 – Scaling laws aren’t dead (but scale isn’t everything) 33:07 – Synthetic data: powerful, dangerous? 35:00 – Reasoning traces: what he can’t say (and why) 37:18 – Long context + attention: what’s next 38:40 – Retrieval vs RAG vs long context 41:49 – The real boss fight: evals (and contamination) 42:28 – Alignment: pre-training vs post-training 43:32 – Deep Think + agents + “vibe coding” 46:34 – Continual learning: updating models over time 49:35 – Advice for researchers + founders 53:35 – “No end in sight” for progress + closing

Matt Turck

51,317 次观看 • 7 个月前

“We have a crisis of open source models in the Western world. Outside of China, there are no good open source models. We don't even have any in the US now. The talent, the capital and the focus to be best-in-class at pre-training, mid-training, post-training is an extremely scarce skillset. The answer for a lot of countries may just be to take an open weights model from China, post-train or fine-tune their own version, and have that be what they start from." Everett Randle Why does the West have such poor open source models and is it a national security threat to rely on open-source Chinese models Demis Hassabis Yann LeCun Aidan Gomez Lilian Weng Jan Leike

“We have a crisis of open source models in the Western world. Outside of China, there are no good open source models. We don't even have any in the US now. The talent, the capital and the focus to be best-in-class at pre-training, mid-training, post-training is an extremely scarce skillset. The answer for a lot of countries may just be to take an open weights model from China, post-train or fine-tune their own version, and have that be what they start from." Everett Randle Why does the West have such poor open source models and is it a national security threat to rely on open-source Chinese models Demis Hassabis Yann LeCun Aidan Gomez Lilian Weng Jan Leike

Harry Stebbings

63,397 次观看 • 1 个月前

New Course: Post-training of LLMs Learn to post-train and customize an LLM in this short course, taught by Banghua Zhu, Assistant Professor at the University of Washington University of Washington, and co-founder of @NexusflowX. Training an LLM to follow instructions or answer questions has two key stages: pre-training and post-training. In pre-training, it learns to predict the next word or token from large amounts of unlabeled text. In post-training, it learns useful behaviors such as following instructions, tool use, and reasoning. Post-training transforms a general-purpose token predictor—trained on trillions of unlabeled text tokens—into an assistant that follows instructions and performs specific tasks. Because it is much cheaper than pre-training, it is practical for many more teams to incorporate post-training methods into their workflows than pre-training. In this course, you’ll learn three common post-training methods—Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Online Reinforcement Learning (RL)—and how to use each one effectively. With SFT, you train the model on pairs of input and ideal output responses. With DPO, you provide both a preferred (chosen) and a less preferred (rejected) response and train the model to favor the preferred output. With RL, the model generates an output, receives a reward score based on human or automated feedback, and updates the model to improve performance. You’ll learn the basic concepts, common use cases, and principles for curating high-quality data for effective training. Through hands-on labs, you’ll download a pre-trained model from Hugging Face and post-train it using SFT, DPO, and RL to see how each technique shapes model behavior. In detail, you’ll: - Understand what post-training is, when to use it, and how it differs from pre-training. - Build an SFT pipeline to turn a base model into an instruct model. - Explore how DPO reshapes behavior by minimizing contrastive loss—penalizing poor responses and reinforcing preferred ones. - Implement a DPO pipeline to change the identity of a chat assistant. - Learn online RL methods such as Proximal Policy Optimization (PPO) and Group Relative Policy Optimization (GRPO), and how to design reward functions. - Train a model with GRPO to improve its math capabilities using a verifiable reward. Post-training is one of the most rapidly developing areas of LLM training. Whether you’re building a high-accuracy context-specific assistant, fine-tuning a model's tone, or improving task-specific accuracy, this course will give you experience with the most important techniques shaping how LLMs are post-trained today. Please sign up here:

New Course: Post-training of LLMs Learn to post-train and customize an LLM in this short course, taught by Banghua Zhu, Assistant Professor at the University of Washington University of Washington, and co-founder of @NexusflowX. Training an LLM to follow instructions or answer questions has two key stages: pre-training and post-training. In pre-training, it learns to predict the next word or token from large amounts of unlabeled text. In post-training, it learns useful behaviors such as following instructions, tool use, and reasoning. Post-training transforms a general-purpose token predictor—trained on trillions of unlabeled text tokens—into an assistant that follows instructions and performs specific tasks. Because it is much cheaper than pre-training, it is practical for many more teams to incorporate post-training methods into their workflows than pre-training. In this course, you’ll learn three common post-training methods—Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Online Reinforcement Learning (RL)—and how to use each one effectively. With SFT, you train the model on pairs of input and ideal output responses. With DPO, you provide both a preferred (chosen) and a less preferred (rejected) response and train the model to favor the preferred output. With RL, the model generates an output, receives a reward score based on human or automated feedback, and updates the model to improve performance. You’ll learn the basic concepts, common use cases, and principles for curating high-quality data for effective training. Through hands-on labs, you’ll download a pre-trained model from Hugging Face and post-train it using SFT, DPO, and RL to see how each technique shapes model behavior. In detail, you’ll: - Understand what post-training is, when to use it, and how it differs from pre-training. - Build an SFT pipeline to turn a base model into an instruct model. - Explore how DPO reshapes behavior by minimizing contrastive loss—penalizing poor responses and reinforcing preferred ones. - Implement a DPO pipeline to change the identity of a chat assistant. - Learn online RL methods such as Proximal Policy Optimization (PPO) and Group Relative Policy Optimization (GRPO), and how to design reward functions. - Train a model with GRPO to improve its math capabilities using a verifiable reward. Post-training is one of the most rapidly developing areas of LLM training. Whether you’re building a high-accuracy context-specific assistant, fine-tuning a model's tone, or improving task-specific accuracy, this course will give you experience with the most important techniques shaping how LLMs are post-trained today. Please sign up here:

Andrew Ng

125,146 次观看 • 1 年前

No wonder Jensen got so defensive over $NVDA GPUs against custom chips. $GOOG will start using different custom chips for inference and training. $AMZN has already adopted this approach as they have Trainium for training and Inferentia for inference. Hyperscalers obviously see better price/performance in this approach, and this is where the industry is headed. $AVGO and $MRVL are poised to be big winners, while $NVDA dominance will be under increasing pressure.

No wonder Jensen got so defensive over $NVDA GPUs against custom chips. $GOOG will start using different custom chips for inference and training. $AMZN has already adopted this approach as they have Trainium for training and Inferentia for inference. Hyperscalers obviously see better price/performance in this approach, and this is where the industry is headed. $AVGO and $MRVL are poised to be big winners, while $NVDA dominance will be under increasing pressure.

Oguz Erkan

223,244 次观看 • 3 个月前

Jensen Huang said Grok 5 will be 7 Trillion parameter model. On time to train, with the training window fixed at 1 month, the new Vera-Rubin GPU system of Nvidia needs 1/4 the number of systems compared with Blackwell to train the same frontier model. “factory throughput” improves by about 10x over Blackwell, and Blackwell itself was about 10x over Hopper. i.e. overall Rubin is roughly 100x Hopper in factory throughput per watt, which matters hugely, because a 1 GW, $50B data center is power-limited and revenue scales with throughput per watt. --- From Nvidia YT channel

Jensen Huang said Grok 5 will be 7 Trillion parameter model. On time to train, with the training window fixed at 1 month, the new Vera-Rubin GPU system of Nvidia needs 1/4 the number of systems compared with Blackwell to train the same frontier model. “factory throughput” improves by about 10x over Blackwell, and Blackwell itself was about 10x over Hopper. i.e. overall Rubin is roughly 100x Hopper in factory throughput per watt, which matters hugely, because a 1 GW, $50B data center is power-limited and revenue scales with throughput per watt. --- From Nvidia YT channel

Rohan Paul

165,770 次观看 • 6 个月前

HUGE! Exxon Mobile just said it would take just WEEKS to hit the ground RUNNING in Venezuela! Trump: "We want SPEED and QUALITY!" This is going to be MASSIVE!

HUGE! Exxon Mobile just said it would take just WEEKS to hit the ground RUNNING in Venezuela! Trump: "We want SPEED and QUALITY!" This is going to be MASSIVE!

Gunther Eagleman™

728,635 次观看 • 6 个月前

Nvidia $NVDA CEO Jensen Huang was asked about AI and what it will do to the job market "This is the largest infrastructure buildout in human history. That's going to create a lot of jobs"

Nvidia $NVDA CEO Jensen Huang was asked about AI and what it will do to the job market "This is the largest infrastructure buildout in human history. That's going to create a lot of jobs"

Evan

45,803 次观看 • 6 个月前

Ilya Sutskever says the AI scaling era is running out of road because pre-training is about to hit a finite-data wall. "At some point though, pre-training will run out of data. The data is very clearly finite." "Then, okay, what do you do next? Either you do some kind of a souped-up pre-training... or you're doing RL, or maybe something else." "In some sense, we are back to the age of research." His timeline is the part most AI investors and founders should hear: "From 2012 to 2020, it was the age of research. Now from 2020 to 2025, it was the age of scaling... But now the scale is so big." "Is the belief really that... if you had 100x more everything would be so different?" "I don't think that's true. So it's back to the age of research again, just with big computers."

Ilya Sutskever says the AI scaling era is running out of road because pre-training is about to hit a finite-data wall. "At some point though, pre-training will run out of data. The data is very clearly finite." "Then, okay, what do you do next? Either you do some kind of a souped-up pre-training... or you're doing RL, or maybe something else." "In some sense, we are back to the age of research." His timeline is the part most AI investors and founders should hear: "From 2012 to 2020, it was the age of research. Now from 2020 to 2025, it was the age of scaling... But now the scale is so big." "Is the belief really that... if you had 100x more everything would be so different?" "I don't think that's true. So it's back to the age of research again, just with big computers."

Karl Mehta

199,932 次观看 • 1 个月前

#TicketToHeaven #GeminiFourth #เจมีไนน์โฟร์ท Phi Aof: as for Gemini, when we read the script together, the first thing Gemini came and said was that, “since i started working, this is the character i relate to the most in my life. and i immediately understood why he is like that and i already knew how i would portray it.” (🥹👏🏻) its exactly what i said before, that it requires a lot of exchange and commitment to this role, to the point that he had to fully commit and jump very high into it. that is the maturity we feel the nongs actors have encountered through this experience

#TicketToHeaven #GeminiFourth #เจมีไนน์โฟร์ท Phi Aof: as for Gemini, when we read the script together, the first thing Gemini came and said was that, “since i started working, this is the character i relate to the most in my life. and i immediately understood why he is like that and i already knew how i would portray it.” (🥹👏🏻) its exactly what i said before, that it requires a lot of exchange and commitment to this role, to the point that he had to fully commit and jump very high into it. that is the maturity we feel the nongs actors have encountered through this experience

𝗚𝗲𝗺 𝗙𝗼𝗿 𝗠𝘆 𝗙𝗼𝘂𝗿𝘁𝗵🌻

51,954 次观看 • 28 天前

Jensen Huang says building xAI's computing supercluster in just 19 days, from concept to training, was a "superhuman" effort and the only person in the world who could have done it is Elon Musk

Jensen Huang says building xAI's computing supercluster in just 19 days, from concept to training, was a "superhuman" effort and the only person in the world who could have done it is Elon Musk

Tsarathustra

297,452 次观看 • 1 年前

Jensen Huang: “Demand is so strong that rental prices of 2 generation old GPUs are going up.” Jensen said this in January and Blackwell rental prices are up by 50% since then. The demand is so excessive now that Anthropic had to reduce default thinking effort of Claude to be able to serve everybody. There is simply not enough compute. $AMD $NVDA $NBIS $AMZN $ORCL

Jensen Huang: “Demand is so strong that rental prices of 2 generation old GPUs are going up.” Jensen said this in January and Blackwell rental prices are up by 50% since then. The demand is so excessive now that Anthropic had to reduce default thinking effort of Claude to be able to serve everybody. There is simply not enough compute. $AMD $NVDA $NBIS $AMZN $ORCL

Oguz Erkan

73,553 次观看 • 3 个月前

The news from yesterday is that I will be training for my first powerlifting competition. It’s January 20th, 2024 and I’ll be going for 2000 pounds raw total. This has been a huge goal for me since I was a kid and I’m super excited to share this next journey with you all!

The news from yesterday is that I will be training for my first powerlifting competition. It’s January 20th, 2024 and I’ll be going for 2000 pounds raw total. This has been a huge goal for me since I was a kid and I’m super excited to share this next journey with you all!

🐃🥩

16,424 次观看 • 2 年前