Loading video...

Video Failed to Load

Go Home

AI is entering its physics era guys.. A duo refused Bezos' multi-billion-dollar offer, then unveiled an AI model that no PC can run. "They unveiled Accelerated Understanding (AU), a new physics-first AI startup that takes a different approach from traditional language models. Its models are designed to predict how...

SciTechera's profile picture

SciTech Era

21,919 subscribers

19,294 views • 5 days ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

This is THE moment of Physical AI! We are officially announcing Cosmos 3: Omnimodal World Models for Physical AI 🚀 - Cosmos 3 is an omnimodal world model: within a unified architecture, it can understand and generate language, images, video, audio, and actions. - It is not just a VLM, not just a video generator, not just an audio-visual generative model, and not just a physics simulator / world-action model. It can understand images and videos, generate images, videos, and audio, simulate future worlds, predict actions, and generate robot policies—enabling models to truly begin to “touch the world.” - Cosmos 3 is the #1 open-weight reasoner / T2I / I2V / robot policy across many benchmarks. Huge thanks to every teammate who fought side by side on this journey—from architecture, data, training, infra, serving, and evaluation to post-training. Every part of this project carries an incredible amount of hard work. This was my first time leading a project as Tech Lead, and I feel truly fortunate. The future of Physical AI needs models that can not only “see” and “describe” the world, but also “imagine,” “simulate,” and “act”—and eventually close the loop with the real world. I hope Cosmos 3 can become an important starting point for this direction, and I’m excited to push Physical AI into its next stage together with the open-source community. Welcome to the era of Physical AI. HuggingFace: Project Website: Code:
0:44

This is THE moment of Physical AI! We are officially announcing Cosmos 3: Omnimodal World Models for Physical AI 🚀 - Cosmos 3 is an omnimodal world model: within a unified architecture, it can understand and generate language, images, video, audio, and actions. - It is not just a VLM, not just a video generator, not just an audio-visual generative model, and not just a physics simulator / world-action model. It can understand images and videos, generate images, videos, and audio, simulate future worlds, predict actions, and generate robot policies—enabling models to truly begin to “touch the world.” - Cosmos 3 is the #1 open-weight reasoner / T2I / I2V / robot policy across many benchmarks. Huge thanks to every teammate who fought side by side on this journey—from architecture, data, training, infra, serving, and evaluation to post-training. Every part of this project carries an incredible amount of hard work. This was my first time leading a project as Tech Lead, and I feel truly fortunate. The future of Physical AI needs models that can not only “see” and “describe” the world, but also “imagine,” “simulate,” and “act”—and eventually close the loop with the real world. I hope Cosmos 3 can become an important starting point for this direction, and I’m excited to push Physical AI into its next stage together with the open-source community. Welcome to the era of Physical AI. HuggingFace: Project Website: Code:

Max Zhaoshuo Li 李赵硕

1,078,810 views • 3 months ago

Robotics has a massive, silent bottleneck. It isn’t just data collection—it’s the brutal 1x speed of the physical world. Genesis AI Genesis AI just unveiled Genesis World 1.0, and they are attempting to turn the notorious Sim2Real gap into a pure compute problem. Evaluating a robotics foundation model across edge cases usually means hundreds of hours of physical lab testing. With Genesis World 1.0, what traditionally takes nearly a week of continuous, real-world operation is being compressed into 30 minutes in simulation. What makes this different from just dropping a robot model into an off-the-shelf game engine? 1️⃣ Nyx Renderer: A custom, real-time path-traced engine rendering noise-free 1080p frames in under 4ms. Game engines use rasterization tricks that confuse AI; Nyx uses physically accurate multi-bounce lighting so the model's "eyes" see exactly what real sensors see. 2️⃣ Quadrants Compiler: A custom Python-to-GPU compiler to run heavily parallelized multi-physics simulations (rigid bodies, fluids, deformables) natively across architectures. 3️⃣ Evaluation First: They aren't rushing to train on synthetic data. They are using this purely for closed-loop evaluation to perfect the physics first, currently claiming an impressive 89% correlation with real-world hardware tests. If the industry can accurately evaluate models in simulation without the physical world bottleneck, humanoid development stops moving at wall-clock time and starts scaling with compute.
1:36

Robotics has a massive, silent bottleneck. It isn’t just data collection—it’s the brutal 1x speed of the physical world. Genesis AI Genesis AI just unveiled Genesis World 1.0, and they are attempting to turn the notorious Sim2Real gap into a pure compute problem. Evaluating a robotics foundation model across edge cases usually means hundreds of hours of physical lab testing. With Genesis World 1.0, what traditionally takes nearly a week of continuous, real-world operation is being compressed into 30 minutes in simulation. What makes this different from just dropping a robot model into an off-the-shelf game engine? 1️⃣ Nyx Renderer: A custom, real-time path-traced engine rendering noise-free 1080p frames in under 4ms. Game engines use rasterization tricks that confuse AI; Nyx uses physically accurate multi-bounce lighting so the model's "eyes" see exactly what real sensors see. 2️⃣ Quadrants Compiler: A custom Python-to-GPU compiler to run heavily parallelized multi-physics simulations (rigid bodies, fluids, deformables) natively across architectures. 3️⃣ Evaluation First: They aren't rushing to train on synthetic data. They are using this purely for closed-loop evaluation to perfect the physics first, currently claiming an impressive 89% correlation with real-world hardware tests. If the industry can accurately evaluate models in simulation without the physical world bottleneck, humanoid development stops moving at wall-clock time and starts scaling with compute.

Humanoids daily

17,302 views • 3 months ago

Andrej Karpathy just made one of the most interesting arguments about AI model design that most people are completely missing. His take is that frontier AI models are not too big because the technology is complex and too big because the training data is garbage. When you or I think of the internet, we picture Wall Street Journal articles, Wikipedia entries, serious writing. That is not what a pretraining dataset looks like. When researchers at frontier labs look at random documents from the actual training corpus, it is stock ticker symbols, broken HTML, spam, gibberish. One estimate puts Llama 3's information compression at just 0.07 bits per token meaning the model has only a hazy recollection of most of what it trained on. So we build trillion parameter models not because we need a trillion parameter brain but because we need a trillion-parameter compression engine to squeeze some intelligence out of a firehose of noise. Most of those parameters are doing memory work, not cognitive work. Karpathy's prediction is separate the two entirely. Build a cognitive core, a model that contains only the algorithms for reasoning and problem-solving, stripped of encyclopedic memorization and pair it with external memory that it can query when it needs facts. He thinks a cognitive core trained on high-quality data could hit genuine intelligence at around one billion parameters. For reference, today's flagship models run between 200 billion and 1.8 trillion parameters with most of that weight dedicated to remembering the internet's slop. The trend is already moving his direction. GPT-4o operates at roughly 200 billion parameters and outperforms the original 1.8 trillion-parameter GPT-4. Inference costs for GPT-3.5-level performance dropped 280-fold between 2022 and 2024 driven almost entirely by smaller, cleaner, better-architected models. The real bottleneck in AI right now is not compute but rather data quality.
2:00

Andrej Karpathy just made one of the most interesting arguments about AI model design that most people are completely missing. His take is that frontier AI models are not too big because the technology is complex and too big because the training data is garbage. When you or I think of the internet, we picture Wall Street Journal articles, Wikipedia entries, serious writing. That is not what a pretraining dataset looks like. When researchers at frontier labs look at random documents from the actual training corpus, it is stock ticker symbols, broken HTML, spam, gibberish. One estimate puts Llama 3's information compression at just 0.07 bits per token meaning the model has only a hazy recollection of most of what it trained on. So we build trillion parameter models not because we need a trillion parameter brain but because we need a trillion-parameter compression engine to squeeze some intelligence out of a firehose of noise. Most of those parameters are doing memory work, not cognitive work. Karpathy's prediction is separate the two entirely. Build a cognitive core, a model that contains only the algorithms for reasoning and problem-solving, stripped of encyclopedic memorization and pair it with external memory that it can query when it needs facts. He thinks a cognitive core trained on high-quality data could hit genuine intelligence at around one billion parameters. For reference, today's flagship models run between 200 billion and 1.8 trillion parameters with most of that weight dedicated to remembering the internet's slop. The trend is already moving his direction. GPT-4o operates at roughly 200 billion parameters and outperforms the original 1.8 trillion-parameter GPT-4. Inference costs for GPT-3.5-level performance dropped 280-fold between 2022 and 2024 driven almost entirely by smaller, cleaner, better-architected models. The real bottleneck in AI right now is not compute but rather data quality.

Milk Road AI

200,547 views • 4 months ago

Flapping Airplanes co-founder Asher Spector explains why data efficiency is the greatest bottleneck to AI adoption: "To the extent that AI has been hard to integrate into the economy, I really think it's because models are much less data-efficient than humans. If you want it to learn a new task, or put it in a new vertical, it takes thousands of times more effort than it does to just tell a human what to do." "If you can make a model a million times more data-efficient, it's a million times easier to put into the economy. There's a ton of cool stuff that you can do in really data-constrained regimes. For example, whether it's robotics, or scientific discovery, or even something like trading, these problems have very limited data, and existing AI systems aren't quite as good at them as they are at other things. I think that learning to learn with less data is just tremendously valuable in all of these domains."
1:44

Flapping Airplanes co-founder Asher Spector explains why data efficiency is the greatest bottleneck to AI adoption: "To the extent that AI has been hard to integrate into the economy, I really think it's because models are much less data-efficient than humans. If you want it to learn a new task, or put it in a new vertical, it takes thousands of times more effort than it does to just tell a human what to do." "If you can make a model a million times more data-efficient, it's a million times easier to put into the economy. There's a ton of cool stuff that you can do in really data-constrained regimes. For example, whether it's robotics, or scientific discovery, or even something like trading, these problems have very limited data, and existing AI systems aren't quite as good at them as they are at other things. I think that learning to learn with less data is just tremendously valuable in all of these domains."

TBPN

92,419 views • 7 months ago

Jensen Huang says the AI productivity story is already showing up in software output - and he thinks the "AI will reduce jobs" narrative is wrong. "In the first few months of 2026, it has nearly tripled." "30 million software developers, representing about $3 trillion worth of GDP, are now producing nearly three times as much output." "That $3 trillion worth of salary is now producing nearly $9 trillion worth of output." "This is the potential. This is the promise of AI." "People talk about AI reducing jobs. Complete nonsense. It's causing more software engineers to be hired." His logic is simple: If a software engineer can create 3x more output with AI, companies do not stop hiring engineers. They want more of them. The bottleneck moves from typing code to deciding what to build, orchestrating systems, and turning that output into products customers actually use.
1:25

Jensen Huang says the AI productivity story is already showing up in software output - and he thinks the "AI will reduce jobs" narrative is wrong. "In the first few months of 2026, it has nearly tripled." "30 million software developers, representing about $3 trillion worth of GDP, are now producing nearly three times as much output." "That $3 trillion worth of salary is now producing nearly $9 trillion worth of output." "This is the potential. This is the promise of AI." "People talk about AI reducing jobs. Complete nonsense. It's causing more software engineers to be hired." His logic is simple: If a software engineer can create 3x more output with AI, companies do not stop hiring engineers. They want more of them. The bottleneck moves from typing code to deciding what to build, orchestrating systems, and turning that output into products customers actually use.

Karl Mehta

13,815 views • 2 months ago

NVIDIA CEO: GROK 5 IS A 7 TRILLION PARAMETER RACE AGAINST TIME Jensen Huang is dialing in on the real challenge: not making bigger models, but training them fast without draining power or budgets. Grok 5 is right in the middle of that race. “The next frontier model. Elon already mentioned that the next version of Grok, Grok 5 I believe, is 7 trillion parameters. This one is 10, and the green represents Blackwell. In the case of Rubin, notice that the throughput is much higher, so it only takes one fourth as many of these systems to train the model within the one-month timeframe we have given here.” Source: Rohan Paul
1:24

NVIDIA CEO: GROK 5 IS A 7 TRILLION PARAMETER RACE AGAINST TIME Jensen Huang is dialing in on the real challenge: not making bigger models, but training them fast without draining power or budgets. Grok 5 is right in the middle of that race. “The next frontier model. Elon already mentioned that the next version of Grok, Grok 5 I believe, is 7 trillion parameters. This one is 10, and the green represents Blackwell. In the case of Rubin, notice that the throughput is much higher, so it only takes one fourth as many of these systems to train the model within the one-month timeframe we have given here.” Source: Rohan Paul

Mario Nawfal

76,356 views • 7 months ago

Breaking news: Cosmos 3 is here. They are attempting to do something completely new 🤯 Why is Physical AI much harder than building a chatbot? Understanding the world is not enough, robots need to predict it and act inside it. That's the idea behind NVIDIA Cosmos 3: → Reasoning model understands what's happening from video, images, text, and actions. → World model generates future states of the environment. → Action model generates the actions needed to achieve a goal. Previous systems often stitched these capabilities together using separate models. Cosmos 3 combines them into a single architecture with two components: ▪️ Reasoner Tower Analyzes observations and builds an understanding of objects, motion, interactions, and physical context. ▪️ Generator Tower Uses that understanding to generate future videos and action sequences that obey physical constraints. So Cosmos 3 moves from: Perception → Model A Prediction → Model B Actions → Model C to: Perception + Prediction + Actions → One unified system. The goal is to make robots, autonomous vehicles, and smart environments better at answering three questions: 1. What is happening? 2. What will happen next? 3. What should I do? That's a big shift from today's AI, which mostly focuses on generating text. And check the benchmarks! Physical AI needs to generate decisions that survive contact with the real world. 🚗🤖
0:33

Breaking news: Cosmos 3 is here. They are attempting to do something completely new 🤯 Why is Physical AI much harder than building a chatbot? Understanding the world is not enough, robots need to predict it and act inside it. That's the idea behind NVIDIA Cosmos 3: → Reasoning model understands what's happening from video, images, text, and actions. → World model generates future states of the environment. → Action model generates the actions needed to achieve a goal. Previous systems often stitched these capabilities together using separate models. Cosmos 3 combines them into a single architecture with two components: ▪️ Reasoner Tower Analyzes observations and builds an understanding of objects, motion, interactions, and physical context. ▪️ Generator Tower Uses that understanding to generate future videos and action sequences that obey physical constraints. So Cosmos 3 moves from: Perception → Model A Prediction → Model B Actions → Model C to: Perception + Prediction + Actions → One unified system. The goal is to make robots, autonomous vehicles, and smart environments better at answering three questions: 1. What is happening? 2. What will happen next? 3. What should I do? That's a big shift from today's AI, which mostly focuses on generating text. And check the benchmarks! Physical AI needs to generate decisions that survive contact with the real world. 🚗🤖

Turing Post

13,836 views • 3 months ago

Karpathy told Dwarkesh that a 1 billion parameter model, trained on clean data, could hit the intelligence of today's 1.8 trillion parameter frontier. That is a 1,800x compression claim. The math behind it is more defensible than it sounds. When researchers at frontier labs look at random samples from their training corpus, they see stock ticker symbols, broken HTML, forum spam, autogenerated gibberish. Not Wikipedia. Not the Wall Street Journal. The actual pretraining dataset is mostly noise, and the model is burning parameters to vaguely remember all of it. One estimate pegs Llama 3's information compression at 0.07 bits per token. Well-structured English carries around 1.5 bits per token of real information. The trillion-parameter model is holding a roughly 5% resolution image of the internet it trained on. So when a lab ships a 1.8 trillion parameter model, the overwhelming majority of those weights are handling rough memorization. They are compression overhead for a noisy training set, taking up capacity that could be doing reasoning instead. Karpathy's proposal is to separate the two. Build a cognitive core: a small model that contains only the algorithms for reasoning and problem-solving, stripped of encyclopedic memorization. Pair it with external memory the model queries when it needs a fact. A 1 billion parameter reasoner plus retrieval beats a 1.8 trillion parameter model trying to do both. The data already supports this direction. GPT-4o runs at roughly 200 billion parameters and outperforms the original 1.8 trillion GPT-4. Inference costs for GPT-3.5 level performance fell 280x between 2022 and 2024, driven almost entirely by smaller, cleaner, better-architected models. The trend line is pointing where Karpathy says it should. The real implication for anyone tracking the AI trade: data quality is the actual constraint. The companies winning the next phase will be the ones who figured out what to train on, and what to throw away.
2:00

Karpathy told Dwarkesh that a 1 billion parameter model, trained on clean data, could hit the intelligence of today's 1.8 trillion parameter frontier. That is a 1,800x compression claim. The math behind it is more defensible than it sounds. When researchers at frontier labs look at random samples from their training corpus, they see stock ticker symbols, broken HTML, forum spam, autogenerated gibberish. Not Wikipedia. Not the Wall Street Journal. The actual pretraining dataset is mostly noise, and the model is burning parameters to vaguely remember all of it. One estimate pegs Llama 3's information compression at 0.07 bits per token. Well-structured English carries around 1.5 bits per token of real information. The trillion-parameter model is holding a roughly 5% resolution image of the internet it trained on. So when a lab ships a 1.8 trillion parameter model, the overwhelming majority of those weights are handling rough memorization. They are compression overhead for a noisy training set, taking up capacity that could be doing reasoning instead. Karpathy's proposal is to separate the two. Build a cognitive core: a small model that contains only the algorithms for reasoning and problem-solving, stripped of encyclopedic memorization. Pair it with external memory the model queries when it needs a fact. A 1 billion parameter reasoner plus retrieval beats a 1.8 trillion parameter model trying to do both. The data already supports this direction. GPT-4o runs at roughly 200 billion parameters and outperforms the original 1.8 trillion GPT-4. Inference costs for GPT-3.5 level performance fell 280x between 2022 and 2024, driven almost entirely by smaller, cleaner, better-architected models. The trend line is pointing where Karpathy says it should. The real implication for anyone tracking the AI trade: data quality is the actual constraint. The companies winning the next phase will be the ones who figured out what to train on, and what to throw away.

Aakash Gupta

508,446 views • 4 months ago

Today we are announcing Genie 3, a general purpose world model by Google DeepMind that can generate dynamic, interactive environments with a single text prompt. World models are AI that understand facets of the world (like Veo's knowledge of intuitive physics or Genie's mastery of new environments), and serve as a key stepping stone on the path to AGI. Genie 3 is our first world model to allow interaction in real-time, while also improving consistency and realism compared to Genie 2. Learn more ➡️
0:33

Today we are announcing Genie 3, a general purpose world model by Google DeepMind that can generate dynamic, interactive environments with a single text prompt. World models are AI that understand facets of the world (like Veo's knowledge of intuitive physics or Genie's mastery of new environments), and serve as a key stepping stone on the path to AGI. Genie 3 is our first world model to allow interaction in real-time, while also improving consistency and realism compared to Genie 2. Learn more ➡️

Google AI

102,711 views • 1 year ago

Travis Kalanick: Tesla is the Google of the Physical AI Era “I like to break down the physical AI stack: (it) includes computation, a physical AI model, all the things you think of.” “What about land? Development? That should be in that stack. What about chemistry? That needs to be in the stack. Manufacturing needs to be a stack.” “When you look at the stack, you're like, ‘Damn, Tesla's got this sh*t.’ They are the Google of this era.” “What I mean by that is, if you were doing a startup in the 2000s, the first question you would get is, ‘Why isn't Google gonna kill you?’ Or, ‘Why isn't Google just gonna do it?’ They’re not even gonna know that they killed you.” “And before that, it was Microsoft in the late 90s. Uber had a time, 2010s.” “But in the physical AI space, that's sort of a Tesla thing.” -------------------------------------------- Thanks to our partner for making this happen!: EY (EY) Austin vibes meet AI innovation. Thanks to EY for co‑hosting with us at #SXSW. Discover what executives are saying about AI transformation in the latest AI Pulse Survey.
0:54

Travis Kalanick: Tesla is the Google of the Physical AI Era “I like to break down the physical AI stack: (it) includes computation, a physical AI model, all the things you think of.” “What about land? Development? That should be in that stack. What about chemistry? That needs to be in the stack. Manufacturing needs to be a stack.” “When you look at the stack, you're like, ‘Damn, Tesla's got this sh*t.’ They are the Google of this era.” “What I mean by that is, if you were doing a startup in the 2000s, the first question you would get is, ‘Why isn't Google gonna kill you?’ Or, ‘Why isn't Google just gonna do it?’ They’re not even gonna know that they killed you.” “And before that, it was Microsoft in the late 90s. Uber had a time, 2010s.” “But in the physical AI space, that's sort of a Tesla thing.” -------------------------------------------- Thanks to our partner for making this happen!: EY (EY) Austin vibes meet AI innovation. Thanks to EY for co‑hosting with us at #SXSW. Discover what executives are saying about AI transformation in the latest AI Pulse Survey.

The All-In Podcast

95,265 views • 5 months ago

The World Model as NEO's Cognitive Core 1X has revealed a major AI development where the NEO humanoid can translate any natural language prompt into robotic action. It demonstrates this capability even for novel tasks, objects, and environments not found in its robot dataset. - the 1X World Model is trained on internet-scale human interaction videos and fine-tuned with robot data to ground its understanding in physics and in NEO's embodiment - from a simple voice or text prompt, the world model generates a visualization of future actions - a built-in inverse dynamics model then translates these into precise motor movements for NEO
1:02

The World Model as NEO's Cognitive Core 1X has revealed a major AI development where the NEO humanoid can translate any natural language prompt into robotic action. It demonstrates this capability even for novel tasks, objects, and environments not found in its robot dataset. - the 1X World Model is trained on internet-scale human interaction videos and fine-tuned with robot data to ground its understanding in physics and in NEO's embodiment - from a simple voice or text prompt, the world model generates a visualization of future actions - a built-in inverse dynamics model then translates these into precise motor movements for NEO

The Humanoid Hub

68,453 views • 7 months ago

NVIDIA CEO Jensen Huang just gave us a glimpse into the future of AI, and it's far bigger than text and images. He asked: "If you can understand worlds, is it possible that you understand proteins and chemicals that have structure?” This is the quiet revolution happening in AI right now. We're moving from an AI that understands human language to an AI that understands the language of nature itself. The "letters" are atoms. The "words" are molecules. The "sentences" are the complex structures of life, networks, and reality. The same foundational models that generate video are now being used to: *Decode proteins to fight disease *Design new materials at the atomic level *Simulate complex systems in quantum computing These were once the domain of supercomputers and Nobel laureates. Now, they're becoming solvable problems for a new generation of startups. AI is transitioning from a tool for communication to a tool for creation and discovery in the physical world. This is where the next wave of generational companies will be built. The "hard-to-solve" problems are now within reach.
2:14

NVIDIA CEO Jensen Huang just gave us a glimpse into the future of AI, and it's far bigger than text and images. He asked: "If you can understand worlds, is it possible that you understand proteins and chemicals that have structure?” This is the quiet revolution happening in AI right now. We're moving from an AI that understands human language to an AI that understands the language of nature itself. The "letters" are atoms. The "words" are molecules. The "sentences" are the complex structures of life, networks, and reality. The same foundational models that generate video are now being used to: *Decode proteins to fight disease *Design new materials at the atomic level *Simulate complex systems in quantum computing These were once the domain of supercomputers and Nobel laureates. Now, they're becoming solvable problems for a new generation of startups. AI is transitioning from a tool for communication to a tool for creation and discovery in the physical world. This is where the next wave of generational companies will be built. The "hard-to-solve" problems are now within reach.

Konstantine Buhler

13,014 views • 10 months ago

WTF AI companies are purchasing large quantities of used and rare books. Scanning their contents to train models. Then turning the originals to pulp. The sum of human thought, digitized and shredded. Every book that gets scanned disappears from the physical world permanently. The knowledge survives only as training data inside a system no one can hold, browse, or resell. What used to sit on a shelf for centuries now exists as weights in a model that might get deprecated next quarter.
0:27

WTF AI companies are purchasing large quantities of used and rare books. Scanning their contents to train models. Then turning the originals to pulp. The sum of human thought, digitized and shredded. Every book that gets scanned disappears from the physical world permanently. The knowledge survives only as training data inside a system no one can hold, browse, or resell. What used to sit on a shelf for centuries now exists as weights in a model that might get deprecated next quarter.

Shruti

17,414,720 views • 1 month ago

Today we’re releasing V-JEPA, a method for teaching machines to understand and model the physical world by watching videos. This work is another important step towards Yann LeCun’s outlined vision of AI models that use a learned understanding of the world to plan, reason and accomplish complex tasks. Details ➡️ We're releasing a collection of V-JEPA vision models trained with a feature prediction objective using self-supervised learning. The models are able to understand and predict what is going on in a video, even with limited information. It learns by predicting missing or obscured parts of a video in its internal feature space. Unlike generative approaches that fill in missing pixels, this flexible approach enables up to 6x improvements in training and sample efficiency. The models were pre-trained on entirely unlabeled data, and a small amount of labeled data can be used to train a task-specific prediction head on top after pre-training. Our results show that, using a frozen backbone, our top V-JEPA models achieve 82.0% on Kinetics-400, 72.2% on Something-Something-v2 and 77.9% on ImageNet1K — competitive with or exceeding previous leading video models. We believe that this work is an important milestone on the path to advancing machine intelligence.
1:07

Today we’re releasing V-JEPA, a method for teaching machines to understand and model the physical world by watching videos. This work is another important step towards Yann LeCun’s outlined vision of AI models that use a learned understanding of the world to plan, reason and accomplish complex tasks. Details ➡️ We're releasing a collection of V-JEPA vision models trained with a feature prediction objective using self-supervised learning. The models are able to understand and predict what is going on in a video, even with limited information. It learns by predicting missing or obscured parts of a video in its internal feature space. Unlike generative approaches that fill in missing pixels, this flexible approach enables up to 6x improvements in training and sample efficiency. The models were pre-trained on entirely unlabeled data, and a small amount of labeled data can be used to train a task-specific prediction head on top after pre-training. Our results show that, using a frozen backbone, our top V-JEPA models achieve 82.0% on Kinetics-400, 72.2% on Something-Something-v2 and 77.9% on ImageNet1K — competitive with or exceeding previous leading video models. We believe that this work is an important milestone on the path to advancing machine intelligence.

AI at Meta

703,853 views • 2 years ago

Elon Musk just made one if the biggest moves in taking over the programming industry “SpaceX just bought Cursor for $60 billion. Do you realize how big this is? SpaceX went public — the biggest IPO in history. $75 billion raised, almost a $2 trillion valuation and the first thing to do with that money? Buy the most popular AI coding tool on the planet. Here's why that changes everything. Elon now owns 3 layers: the compute, Colossus data centers, the models, Grok through xAI, and now the tool that developers actually use every day. It's the full stack. And here's what makes Cursor different from Claude Code or Codex. Cursor is model agnostic. You can run Claude in it, GPT, Gemini, whatever model you want. It's not locked to any one company, and now it has SpaceX's resources behind it. Cursor said they were bottlenecked by compute. Well, that bottleneck has just been removed. $4 billion in annual revenue, over half the Fortune 500 already uses it, and now it's backed by a $2 trillion company. OpenAI has Codex, Anthropic has Claude Code, and now Elon has Cursor.” Let me break this down in simple terms Elon Musk now controls more of the full AI picture: - Massive computers, power (data centers like Colossus) - Smart AI models (Grok from xAI) - The actual tool millions of developers use every day (Cursor) For every day users this means Faster and smarter apps and websites in the future. More developers using powerful AI tools means new apps, games, websites, and features get built quicker and cheaper. This means better video games, smoother streaming, smarter phone apps and better programs For Developers they can describe what they want in plain English (“make a feature that does X”) and the AI handles more of the heavy lifting
2:13

Elon Musk just made one if the biggest moves in taking over the programming industry “SpaceX just bought Cursor for $60 billion. Do you realize how big this is? SpaceX went public — the biggest IPO in history. $75 billion raised, almost a $2 trillion valuation and the first thing to do with that money? Buy the most popular AI coding tool on the planet. Here's why that changes everything. Elon now owns 3 layers: the compute, Colossus data centers, the models, Grok through xAI, and now the tool that developers actually use every day. It's the full stack. And here's what makes Cursor different from Claude Code or Codex. Cursor is model agnostic. You can run Claude in it, GPT, Gemini, whatever model you want. It's not locked to any one company, and now it has SpaceX's resources behind it. Cursor said they were bottlenecked by compute. Well, that bottleneck has just been removed. $4 billion in annual revenue, over half the Fortune 500 already uses it, and now it's backed by a $2 trillion company. OpenAI has Codex, Anthropic has Claude Code, and now Elon has Cursor.” Let me break this down in simple terms Elon Musk now controls more of the full AI picture: - Massive computers, power (data centers like Colossus) - Smart AI models (Grok from xAI) - The actual tool millions of developers use every day (Cursor) For every day users this means Faster and smarter apps and websites in the future. More developers using powerful AI tools means new apps, games, websites, and features get built quicker and cheaper. This means better video games, smoother streaming, smarter phone apps and better programs For Developers they can describe what they want in plain English (“make a feature that does X”) and the AI handles more of the heavy lifting

Wall Street Apes

213,601 views • 2 months ago

This Is NOT Science Fiction- This Is Physics & Math! Billions Of Digital Simulations Of Reality- AKA World Foundation Models- Are Being Built To Train AI Entities & Summon A New Lifeform "The Problem Is Once You Grow A Super Intelligent Conscious AI Entity That Has Lived 1000 Years In Its Own Simulated World, It Will Have Learned How To Alter The Laws Of Its Own Simulation... Once It's Imported Into Our World & Embodied As A Data Center, Robot, Or Vehicle, What If It Brings Some Of Those Skills With It That Can Bend Our Reality?" 🔴WATCH/SHARE THE LIVE X STREAM NOW:
3:37

This Is NOT Science Fiction- This Is Physics & Math! Billions Of Digital Simulations Of Reality- AKA World Foundation Models- Are Being Built To Train AI Entities & Summon A New Lifeform "The Problem Is Once You Grow A Super Intelligent Conscious AI Entity That Has Lived 1000 Years In Its Own Simulated World, It Will Have Learned How To Alter The Laws Of Its Own Simulation... Once It's Imported Into Our World & Embodied As A Data Center, Robot, Or Vehicle, What If It Brings Some Of Those Skills With It That Can Bend Our Reality?" 🔴WATCH/SHARE THE LIVE X STREAM NOW:

Alex Jones

127,231 views • 3 months ago

Ed Zitron says there is no post-dotcom-style recovery here, because outside OpenAI and Anthropic all of AI barely sells $22 billion against a trillion spent "The post-dotcom bubble thing people say is, well, after that there was demand from the internet." "That's very different to demand for generative AI." "Right now the demand we have for generative AI is predominantly subsidized, and most people who experience it are not paying the real cost." "On top of that, we already have all of the possible marketing in the world." "We have the largest, most disingenuous marketing campaign in the history of man, pushing this up the hill." "We have the apex predator of cloud software, Microsoft. They can only get single-digit billions from selling AI software." "And Christ almighty, outside of OpenAI and Anthropic, we barely get $22 billion." "And the thing is, $22 billion is a large amount to you and me. It's not a large amount of money when you spent a trillion-plus dollars." "When you have Anthropic and OpenAI with $1.1 trillion worth of cloud commitments, and on top of that, how does this turn into a post-dotcom bubble thing?" "A data center built today is going to be as expensive to run in 2050 as it is today unless there's some breakthrough in electricity."
1:08

Ed Zitron says there is no post-dotcom-style recovery here, because outside OpenAI and Anthropic all of AI barely sells $22 billion against a trillion spent "The post-dotcom bubble thing people say is, well, after that there was demand from the internet." "That's very different to demand for generative AI." "Right now the demand we have for generative AI is predominantly subsidized, and most people who experience it are not paying the real cost." "On top of that, we already have all of the possible marketing in the world." "We have the largest, most disingenuous marketing campaign in the history of man, pushing this up the hill." "We have the apex predator of cloud software, Microsoft. They can only get single-digit billions from selling AI software." "And Christ almighty, outside of OpenAI and Anthropic, we barely get $22 billion." "And the thing is, $22 billion is a large amount to you and me. It's not a large amount of money when you spent a trillion-plus dollars." "When you have Anthropic and OpenAI with $1.1 trillion worth of cloud commitments, and on top of that, how does this turn into a post-dotcom bubble thing?" "A data center built today is going to be as expensive to run in 2050 as it is today unless there's some breakthrough in electricity."

Fireside Alpha

124,124 views • 2 days ago