Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Computer scientist Judea Pearl: There are mathematical limits to LLMs that cannot be crossed by scaling alone LLMs don't discover world models from raw data; they merely summarize the interpretations humans have already written down "this path is not the way to get AGI"

Haider.

68,900 subscribers

623,037 views • 7 months ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

Yann LeCun argues that large language models (LLMs) cannot reach human-level or superintelligence just by scaling. He says the current LLM paradigm is hitting its limits. Many researchers are now exploring “agentic systems,” but building them on top of LLMs alone is flawed. LLMs can't plan actions well because they don’t truly understand or predict consequences. To get intelligent behavior, we need something fundamentally different.

Yann LeCun argues that large language models (LLMs) cannot reach human-level or superintelligence just by scaling. He says the current LLM paradigm is hitting its limits. Many researchers are now exploring “agentic systems,” but building them on top of LLMs alone is flawed. LLMs can't plan actions well because they don’t truly understand or predict consequences. To get intelligent behavior, we need something fundamentally different.

Wes Roth

71,849 views • 6 months ago

Professor Judea Pearl — the pioneer who invented causal reasoning in AI — says scaling won't save us. "Mathematical limitations that are not crossable by scaling up." The brutal truth: LLMs aren’t learning how the world works. They are learning how we describe the world. This resonates with most biologists: Drug discovery is hitting the same wall. We have mountains of genomic data, but most AI models just find patterns in published papers — not in the raw biology itself. They're learning what scientists think causes disease, not what actually does. Pearl's causal revolution? That's how we move from "this gene correlates with cancer" to "this gene causes cancer" — and finally design drugs that work. Until then, we're building very expensive parrots.

Professor Judea Pearl — the pioneer who invented causal reasoning in AI — says scaling won't save us. "Mathematical limitations that are not crossable by scaling up." The brutal truth: LLMs aren’t learning how the world works. They are learning how we describe the world. This resonates with most biologists: Drug discovery is hitting the same wall. We have mountains of genomic data, but most AI models just find patterns in published papers — not in the raw biology itself. They're learning what scientists think causes disease, not what actually does. Pearl's causal revolution? That's how we move from "this gene correlates with cancer" to "this gene causes cancer" — and finally design drugs that work. Until then, we're building very expensive parrots.

Bo Wang

804,376 views • 6 months ago

Why haven’t LLMs done new science yet? One reason from Periodic Labs cofounder Ekin Dogus Cubuk: science is iterative. LLMs haven’t been able to iterate experimentally yet. “ LLMs can be very smart, but if they're not iterating on science, they won't discover science. To be honest, humans won't either. You put a human in a room without any chance to iterate on something, they won't discover anything important. So we feel like the important thing to teach these LLMs is the method of scientific inquiry.”

Why haven’t LLMs done new science yet? One reason from Periodic Labs cofounder Ekin Dogus Cubuk: science is iterative. LLMs haven’t been able to iterate experimentally yet. “ LLMs can be very smart, but if they're not iterating on science, they won't discover science. To be honest, humans won't either. You put a human in a room without any chance to iterate on something, they won't discover anything important. So we feel like the important thing to teach these LLMs is the method of scientific inquiry.”

a16z

53,361 views • 10 months ago

📁 Yann LeCun says that scaling models will not get us to human intelligence. He explains that the industry remains obsessed with making LLMs bigger, but that this path is fundamentally broken. It does not matter how many parameters we add or how many clusters we build, because the models only imitate language patterns. Human intelligence does not emerge from size, it emerges from understanding the world.

📁 Yann LeCun says that scaling models will not get us to human intelligence. He explains that the industry remains obsessed with making LLMs bigger, but that this path is fundamentally broken. It does not matter how many parameters we add or how many clusters we build, because the models only imitate language patterns. Human intelligence does not emerge from size, it emerges from understanding the world.

Jon Hernandez

157,988 views • 8 months ago

Yann LeCun -- Meta's chief AI scientist and a Turing Award winner -- explains why scaling up LLMs will never reach human-level AI, and names the four things today's systems still can't do: "If you think that we're going to get to human-level AI by just training on more data and scaling up LLMs, you're making a mistake." "If you're an investor and you invest in a company that told you we're going to get to human-level AI and PhD level by just training on more data and with a few tricks... that was probably not a good idea." "There are ideas about how to go forward and have systems that are capable of doing what every intelligent animal and human are capable of doing, and that current AI systems are not capable of doing." "Understanding the physical world, having persistent memory, and being able to reason and plan. Those are the four characteristics that need to be there." "That requires systems that can acquire common sense, that can learn from natural sensors like video, as opposed to just text." The entire market is priced on a straight line from bigger models to AGI. One of the people who invented deep learning is telling you the line doesn't reach -- and that the real bottleneck isn't compute, it's world models. If he's right, the winner isn't whoever stacks the most GPUs. It's whoever teaches a machine to learn from video the way a child does.

Yann LeCun -- Meta's chief AI scientist and a Turing Award winner -- explains why scaling up LLMs will never reach human-level AI, and names the four things today's systems still can't do: "If you think that we're going to get to human-level AI by just training on more data and scaling up LLMs, you're making a mistake." "If you're an investor and you invest in a company that told you we're going to get to human-level AI and PhD level by just training on more data and with a few tricks... that was probably not a good idea." "There are ideas about how to go forward and have systems that are capable of doing what every intelligent animal and human are capable of doing, and that current AI systems are not capable of doing." "Understanding the physical world, having persistent memory, and being able to reason and plan. Those are the four characteristics that need to be there." "That requires systems that can acquire common sense, that can learn from natural sensors like video, as opposed to just text." The entire market is priced on a straight line from bigger models to AGI. One of the people who invented deep learning is telling you the line doesn't reach -- and that the real bottleneck isn't compute, it's world models. If he's right, the winner isn't whoever stacks the most GPUs. It's whoever teaches a machine to learn from video the way a child does.

Karl Mehta

34,797 views • 1 month ago

I taught a course on the mathematical nitty-gritty of GenAI, equation by equation (written). This covers all the SoTA algorithms, from VAEs to DDPMs to LLMs and state-space models (SSMs), in detail. The link for the entire playlist is in the next message.

I taught a course on the mathematical nitty-gritty of GenAI, equation by equation (written). This covers all the SoTA algorithms, from VAEs to DDPMs to LLMs and state-space models (SSMs), in detail. The link for the entire playlist is in the next message.

prathosh A P

98,333 views • 1 month ago

Yann LeCun says you cannot build a reliable agentic system without a world model LLMs don't have world models. They can't predict the consequences of their actions before taking them "they just act, and whatever happens next is someone else's problem" Without that, it's not intelligence

Yann LeCun says you cannot build a reliable agentic system without a world model LLMs don't have world models. They can't predict the consequences of their actions before taking them "they just act, and whatever happens next is someone else's problem" Without that, it's not intelligence

Haider.

332,721 views • 3 months ago

Demis Hassabis says large language models (LLMs) can search and plan but they forget everything once the session ends. This is because today’s models don’t truly “learn” after deployment. He sees continual learning and better memory as key missing ingredients for AGI. Efficient memory, like the brain’s ability to store only what matters, will be crucial. LLMs are powerful, but without memory, they have a "goldfish brain."

Demis Hassabis says large language models (LLMs) can search and plan but they forget everything once the session ends. This is because today’s models don’t truly “learn” after deployment. He sees continual learning and better memory as key missing ingredients for AGI. Efficient memory, like the brain’s ability to store only what matters, will be crucial. LLMs are powerful, but without memory, they have a "goldfish brain."

Wes Roth

38,833 views • 6 months ago

The CEO of Google DeepMind just went on record saying he disagrees with one of the most respected AI researchers in the world. Demis Hassabis, the man behind AlphaFold, AlphaGo, and Google's entire AI operation publicly pushed back against Yann LeCun's claim that large language models are a dead end for artificial intelligence. LeCun, who left Meta earlier this year to start his own AI lab, has been saying for years that LLMs cannot reason, cannot plan, and will never get us to human-level intelligence. Hassabis disagrees, and he said so directly. His position is that scaling laws are still working, foundation models are still getting more capable, and whatever AGI ends up looking like, LLMs will be a central part of it, not something that gets replaced. He does say there is roughly a 50/50 chance that one or two additional breakthroughs will be needed beyond scaling alone, things like better memory, long-term planning, and world models. But the core disagreement with LeCun is clear, Hassabis believes the current architecture is sound and the current path leads somewhere real. Two Nobel-recognized researchers, two founding figures of modern AI, now publicly on opposite sides of the most important technical question in the industry.

The CEO of Google DeepMind just went on record saying he disagrees with one of the most respected AI researchers in the world. Demis Hassabis, the man behind AlphaFold, AlphaGo, and Google's entire AI operation publicly pushed back against Yann LeCun's claim that large language models are a dead end for artificial intelligence. LeCun, who left Meta earlier this year to start his own AI lab, has been saying for years that LLMs cannot reason, cannot plan, and will never get us to human-level intelligence. Hassabis disagrees, and he said so directly. His position is that scaling laws are still working, foundation models are still getting more capable, and whatever AGI ends up looking like, LLMs will be a central part of it, not something that gets replaced. He does say there is roughly a 50/50 chance that one or two additional breakthroughs will be needed beyond scaling alone, things like better memory, long-term planning, and world models. But the core disagreement with LeCun is clear, Hassabis believes the current architecture is sound and the current path leads somewhere real. Two Nobel-recognized researchers, two founding figures of modern AI, now publicly on opposite sides of the most important technical question in the industry.

Milk Road AI

130,875 views • 4 months ago

this was the most important bit of the ilya interview imo. humans are not AGI, and LLMs won’t become like humans with more scaling. humans lack a huge amount of knowledge and have insanely good “continual learning” powered by extremely good reward functions (except for addiction) that guide them. and so it’s more about the learning rather than knowledge. the term AGI was coined as a response to the narrow AI which was only capable of intelligence in a specific domain (chess, game, code, etc) but General intelligence is able to learn different topics because that’s what “learning” is about. pre-training was thought to be the solution to reaching AGI because you train AI on so many different topics. but they have “overshot” that target. amazing perspective.

this was the most important bit of the ilya interview imo. humans are not AGI, and LLMs won’t become like humans with more scaling. humans lack a huge amount of knowledge and have insanely good “continual learning” powered by extremely good reward functions (except for addiction) that guide them. and so it’s more about the learning rather than knowledge. the term AGI was coined as a response to the narrow AI which was only capable of intelligence in a specific domain (chess, game, code, etc) but General intelligence is able to learn different topics because that’s what “learning” is about. pre-training was thought to be the solution to reaching AGI because you train AI on so many different topics. but they have “overshot” that target. amazing perspective.

ℏεsam

53,444 views • 8 months ago

Google Deepmind CEO gives his prediction on the final AGI architecture: Demis Hassabis just sat down on 20VC and broke down exactly how DeepMind is building toward true AGI, directly pushing back on the idea that LLMs are a dead end. He defends the current trajectory of the models, noting that “we’ve seen how successful these foundation models have been. They can do incredibly impressive things. I don’t think that’s going to go away. We’re still seeing returns from the scaling laws.” But he’s honest about the missing pieces for true superintelligence, saying “I think there might be… there’s a 50-50 chance there’s some things maybe missing that we still need to make breakthroughs in, perhaps their world models, these kinds of approaches.” And when he breaks down what the actual AGI system will look like, he frames it as one question is the LLM foundation model the key component only, or is it the total system. His answer - “I don’t think it’s going to get replaced. I think it’s going to get built on top of these foundation models, just like the way we do with our world models.”

Google Deepmind CEO gives his prediction on the final AGI architecture: Demis Hassabis just sat down on 20VC and broke down exactly how DeepMind is building toward true AGI, directly pushing back on the idea that LLMs are a dead end. He defends the current trajectory of the models, noting that “we’ve seen how successful these foundation models have been. They can do incredibly impressive things. I don’t think that’s going to go away. We’re still seeing returns from the scaling laws.” But he’s honest about the missing pieces for true superintelligence, saying “I think there might be… there’s a 50-50 chance there’s some things maybe missing that we still need to make breakthroughs in, perhaps their world models, these kinds of approaches.” And when he breaks down what the actual AGI system will look like, he frames it as one question is the LLM foundation model the key component only, or is it the total system. His answer - “I don’t think it’s going to get replaced. I think it’s going to get built on top of these foundation models, just like the way we do with our world models.”

Chris

71,716 views • 3 months ago

A 4-year-old child has seen 50x more information than the biggest LLMs. Yann LeCun is the Chief AI Scientist at Meta. He recently spoke on “The Expanding Universe of Generative Models” panel at the World Economic Forum in Davos. Yann highlighted the idea that a 4-year-old child is way smarter than current cutting-edge large language models (LLMs). “Think about what a child sees through vision. Put a number on how much information a 4-year-old child has seen during their life. It’s 20 Mbps going through the optical nerve for 16,000 wake hours in the first 4 years of life. 3,600 seconds per hour is 10^15 bytes. This is 50x more information than the biggest LLMs we have. A 4-year-old child is way smarter than these models having acquired an enormous amount of knowledge about how the world works.” The real constraint right now is the ability of LLMs to think. Today, LLMs are only capable of System 1 thinking. System 1 vs System 2 thinking was popularised in the book 'Thinking, Fast and Slow' by Daniel Kahneman. System 1 tasks involve quick, instinctive, automatic responses. LLMs struggle with discontinuous tasks that require a creative leap in progress as they imitate human responses. It's hard to go above human response accuracy if LLMs are only trained on humans. Models are building the track in front of them with each word being generated. What could it mean to give language models System 2 thinking? This remains a future development I'm excited about.

A 4-year-old child has seen 50x more information than the biggest LLMs. Yann LeCun is the Chief AI Scientist at Meta. He recently spoke on “The Expanding Universe of Generative Models” panel at the World Economic Forum in Davos. Yann highlighted the idea that a 4-year-old child is way smarter than current cutting-edge large language models (LLMs). “Think about what a child sees through vision. Put a number on how much information a 4-year-old child has seen during their life. It’s 20 Mbps going through the optical nerve for 16,000 wake hours in the first 4 years of life. 3,600 seconds per hour is 10^15 bytes. This is 50x more information than the biggest LLMs we have. A 4-year-old child is way smarter than these models having acquired an enormous amount of knowledge about how the world works.” The real constraint right now is the ability of LLMs to think. Today, LLMs are only capable of System 1 thinking. System 1 vs System 2 thinking was popularised in the book 'Thinking, Fast and Slow' by Daniel Kahneman. System 1 tasks involve quick, instinctive, automatic responses. LLMs struggle with discontinuous tasks that require a creative leap in progress as they imitate human responses. It's hard to go above human response accuracy if LLMs are only trained on humans. Models are building the track in front of them with each word being generated. What could it mean to give language models System 2 thinking? This remains a future development I'm excited about.

Alex Banks

22,985 views • 2 years ago

Andrej Karpathy beautifully explains the fundamental difference of learning between a human and an LLM. > “The book I’m reading is a set of prompts for me to do synthetic data generation. It's by manipulating that information that you actually gain that knowledge. We have no equivalent of that with LLMs; they don't really do that.” many of the best minds in this field think LLMs are not really learning anything, and therefore are incapable of surpassing human-level intelligence.

Andrej Karpathy beautifully explains the fundamental difference of learning between a human and an LLM. > “The book I’m reading is a set of prompts for me to do synthetic data generation. It's by manipulating that information that you actually gain that knowledge. We have no equivalent of that with LLMs; they don't really do that.” many of the best minds in this field think LLMs are not really learning anything, and therefore are incapable of surpassing human-level intelligence.

ℏεsam

895,224 views • 10 months ago

Yann LeCun (Yann LeCun ) beautifully explains how the architecture and principles used to train LLMs can not be extended to teach AI the real-world intelligence. In 1 line: LLMs excel where intelligence equals sequence prediction over symbols. Real-world intelligence requires learned world models, abstraction, causality, and action planning under uncertainty, which current next-token training does not provide. He says current LLMs learn by predicting the next token. That objective works very well when the task itself can be reduced to manipulating discrete symbols and sequences. Math, physics problem solving on paper, and coding fit this pattern because success largely comes from searching and composing the right sequences of symbols, equations, or program tokens. With enough data and scale, these models get very good at that kind of structured sequence prediction. Real-world intelligence is different. The physical world is continuous, noisy, uncertain, and high dimensional. To act in it, a system needs internal models that capture objects, dynamics, causality, constraints from the body, and the outcomes of actions over time. Humans and animals build abstract representations from rich sensory streams, then make predictions in that abstract space, not at the raw pixel level. That is why a child can learn intuitive physics, plan multi-step actions, and adapt quickly in new situations with little data. His claim about saturation follows from this gap. Scaling token prediction keeps improving symbol manipulation tasks like math and code, but it hits limits on embodied reasoning and common sense because text alone does not provide the right learning signals for world models. Predicting the next word cannot efficiently teach contact forces, affordances, occlusion, friction, or how actions change the state of the environment. For that, he argues we need architectures that learn abstractions from sensory data and predict futures in abstract latent spaces, then use those predictions to plan actions toward goals with built-in guardrails. --- From 'Pioneer Works' YT Channel (link in comment)

Yann LeCun (Yann LeCun ) beautifully explains how the architecture and principles used to train LLMs can not be extended to teach AI the real-world intelligence. In 1 line: LLMs excel where intelligence equals sequence prediction over symbols. Real-world intelligence requires learned world models, abstraction, causality, and action planning under uncertainty, which current next-token training does not provide. He says current LLMs learn by predicting the next token. That objective works very well when the task itself can be reduced to manipulating discrete symbols and sequences. Math, physics problem solving on paper, and coding fit this pattern because success largely comes from searching and composing the right sequences of symbols, equations, or program tokens. With enough data and scale, these models get very good at that kind of structured sequence prediction. Real-world intelligence is different. The physical world is continuous, noisy, uncertain, and high dimensional. To act in it, a system needs internal models that capture objects, dynamics, causality, constraints from the body, and the outcomes of actions over time. Humans and animals build abstract representations from rich sensory streams, then make predictions in that abstract space, not at the raw pixel level. That is why a child can learn intuitive physics, plan multi-step actions, and adapt quickly in new situations with little data. His claim about saturation follows from this gap. Scaling token prediction keeps improving symbol manipulation tasks like math and code, but it hits limits on embodied reasoning and common sense because text alone does not provide the right learning signals for world models. Predicting the next word cannot efficiently teach contact forces, affordances, occlusion, friction, or how actions change the state of the environment. For that, he argues we need architectures that learn abstractions from sensory data and predict futures in abstract latent spaces, then use those predictions to plan actions toward goals with built-in guardrails. --- From 'Pioneer Works' YT Channel (link in comment)

Rohan Paul

104,460 views • 8 months ago

SITUATION EXPLAINED: A new generation of AI labs are betting that world models are the path to general intelligence that can interact with the real world. We had General Intuition CEO Pim de Witte on MTS to explain what world models are, and why intuitively world models offer a clearer path towards general intelligence than LLMs. For humans, text is a technology for compressing information about the unfolding, three-dimensional world around us and communicating it to other people. When LLMs reason about the world, they do so by reconstructing the world through text. World models work by modeling the three-dimensional world at the architecture level and then allowing agents to interact with that world. Unlike an LLM, which understands space and time through text, world models perform spatial and temporal reasoning natively. General Intuition hopes to use world models to create embodied AI—machines that can reason and act in the real world. But other companies are also pushing the frontier in this space. Odyssey ML today announced a multi-agent world model, Agora-1. Their launch video shows four participants interacting with a real-time world model simulation of GoldenEye, a first-person shooter for the N64. A playable preview of the GoldenEye simulation is available on their site.

SITUATION EXPLAINED: A new generation of AI labs are betting that world models are the path to general intelligence that can interact with the real world. We had General Intuition CEO Pim de Witte on MTS to explain what world models are, and why intuitively world models offer a clearer path towards general intelligence than LLMs. For humans, text is a technology for compressing information about the unfolding, three-dimensional world around us and communicating it to other people. When LLMs reason about the world, they do so by reconstructing the world through text. World models work by modeling the three-dimensional world at the architecture level and then allowing agents to interact with that world. Unlike an LLM, which understands space and time through text, world models perform spatial and temporal reasoning natively. General Intuition hopes to use world models to create embodied AI—machines that can reason and act in the real world. But other companies are also pushing the frontier in this space. Odyssey ML today announced a multi-agent world model, Agora-1. Their launch video shows four participants interacting with a real-time world model simulation of GoldenEye, a first-person shooter for the N64. A playable preview of the GoldenEye simulation is available on their site.

MTS

11,641 views • 3 months ago

LLMs had the internet. Robotics doesn’t, which is why we’re bringing the Scale AI data engine into the physical world. Robotics won’t be solved by blindly collecting 100x more demos. Raw trajectories need structure (intent, subgoals, failures, recoveries, edge cases, quality signals) or you just average mistakes. At Scale Labs, we’re building that data flywheel, mapping the right data to the right capabilities and scaling what works globally !

LLMs had the internet. Robotics doesn’t, which is why we’re bringing the Scale AI data engine into the physical world. Robotics won’t be solved by blindly collecting 100x more demos. Raw trajectories need structure (intent, subgoals, failures, recoveries, edge cases, quality signals) or you just average mistakes. At Scale Labs, we’re building that data flywheel, mapping the right data to the right capabilities and scaling what works globally !

jade

15,262 views • 3 months ago

“Today’s LLMs are the ‘Mainframe Computers’ of our generation” I was on Bloomberg TV today discussing Sakana AI, and to share my view that today’s LLMs are our generation’s “Mainframe Computers”. We are still in the very early stages of AI, and it is inevitable, due to market competition and global innovation (especially from those innovating with resource constraints), that this technology will become a million times more efficient. There is a narrative established in Silicon Valley that AI is a “Winner Takes All” technology, and that scaling up existing models and consuming ever greater resources will require (and even justify) the largest investments of our generation, in order to “win” the AI race. In contrast, I believe that AI is not a “winner takes all” technology. LLMs will be commoditized, become vastly more efficient, and made widely available in all countries. Ultimately, there will be thousands, if not millions, of AI models used by everyone. Just like with the evolution of early clunky mainframe computers to modern computing, how we use AI today will look very different in a few years (or even a year from now), compared to today’s ‘clunky’ LLMs.

“Today’s LLMs are the ‘Mainframe Computers’ of our generation” I was on Bloomberg TV today discussing Sakana AI, and to share my view that today’s LLMs are our generation’s “Mainframe Computers”. We are still in the very early stages of AI, and it is inevitable, due to market competition and global innovation (especially from those innovating with resource constraints), that this technology will become a million times more efficient. There is a narrative established in Silicon Valley that AI is a “Winner Takes All” technology, and that scaling up existing models and consuming ever greater resources will require (and even justify) the largest investments of our generation, in order to “win” the AI race. In contrast, I believe that AI is not a “winner takes all” technology. LLMs will be commoditized, become vastly more efficient, and made widely available in all countries. Ultimately, there will be thousands, if not millions, of AI models used by everyone. Just like with the evolution of early clunky mainframe computers to modern computing, how we use AI today will look very different in a few years (or even a year from now), compared to today’s ‘clunky’ LLMs.

hardmaru

165,382 views • 1 year ago