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。

Self-learning worked in AlphaGo because the world was a clean rulebook, says Demis Hassabis. Real life has no perfect rulebook, so long-term planning needs a strong “world model” to simulate what happens next. Without that simulation, a robot has to guess in the real world—where mistakes can be costly... show more

Wes Roth

33,923 subscribers

15,757 次观看 • 5 个月前 •via X (Twitter)

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

Demis Hassabis says an “AlphaFold moment” for the physical world could be close—real robots doing useful tasks reliably in the real world. Gemini was built multimodal so a future assistant can understand what you see on a phone or glasses. That same “understand the world” ability is a direct bridge into robotics, says Demis Hassabis. Physical intelligence becomes real when robots stop being demos and start being dependable helpers.

Demis Hassabis says an “AlphaFold moment” for the physical world could be close—real robots doing useful tasks reliably in the real world. Gemini was built multimodal so a future assistant can understand what you see on a phone or glasses. That same “understand the world” ability is a direct bridge into robotics, says Demis Hassabis. Physical intelligence becomes real when robots stop being demos and start being dependable helpers.

Wes Roth

21,658 次观看 • 5 个月前

Demis Hassabis says the next AI leap is continual learning. Not train once and freeze, but learn from real experience, constantly. AlphaGo did it in games. Now comes the messy real world. Would you let an AI learn from you? 👀

Demis Hassabis says the next AI leap is continual learning. Not train once and freeze, but learn from real experience, constantly. AlphaGo did it in games. Now comes the messy real world. Would you let an AI learn from you? 👀

VraserX e/acc

19,202 次观看 • 5 个月前

Demis Hassabis says intuitive physics doesn't need embodiment; it can be learned through passive observation This challenges neuroscience theories that link perception to action and embodiment. "we're starting to get towards a world model" an internal map of how the world works.

Demis Hassabis says intuitive physics doesn't need embodiment; it can be learned through passive observation This challenges neuroscience theories that link perception to action and embodiment. "we're starting to get towards a world model" an internal map of how the world works.

Haider.

36,313 次观看 • 11 个月前

Google DeepMind CEO Demis Hassabis says AI agents are coming that will be able to search the Tree of Knowledge to discover new and original things about the world

Google DeepMind CEO Demis Hassabis says AI agents are coming that will be able to search the Tree of Knowledge to discover new and original things about the world

Tsarathustra

42,916 次观看 • 1 年前

Demis Hassabis emphasizes: there is no wall in sight, but according to Hassabis, AGI still needs 1 to 2 breakthroughs. Here he says what is still missing:

Demis Hassabis emphasizes: there is no wall in sight, but according to Hassabis, AGI still needs 1 to 2 breakthroughs. Here he says what is still missing:

Chubby♨️

20,956 次观看 • 5 个月前

Google DeepMind CEO Demis Hassabis says they will have AI agents that everyday consumers can use available later this year as the enabling capabilities, like a world model that can accurately model physics, are already present in their video model Veo 2

Google DeepMind CEO Demis Hassabis says they will have AI agents that everyday consumers can use available later this year as the enabling capabilities, like a world model that can accurately model physics, are already present in their video model Veo 2

Tsarathustra

35,184 次观看 • 1 年前

A conversation with Demis Hassabis on world models (genie 3), deep think, the need for better evals (game arena), and our progress towards AGI.

A conversation with Demis Hassabis on world models (genie 3), deep think, the need for better evals (game arena), and our progress towards AGI.

Logan Kilpatrick

340,665 次观看 • 11 个月前

Demis Hassabis says AGI was meant to be built like CERN. Carefully, with full understanding at each step. That’s not the world we find ourselves in. We’re rushing to bring AGI into the world under pressure in race conditions.

Demis Hassabis says AGI was meant to be built like CERN. Carefully, with full understanding at each step. That’s not the world we find ourselves in. We’re rushing to bring AGI into the world under pressure in race conditions.

vitrupo

60,086 次观看 • 3 个月前

Robot Learning needs 4D world models! Robot Learning needs 4D world models! Robot Learning needs 4D world models! We introduce TesserAct, a 4D embodied world model that can simulate how agents interact with the 3D world over time! We achieve this by simply extending a pre-trained 2D video generation model to jointly predict RGB, depth, and surface normals. It enables: 1️⃣ Much better policy learning in the wild 2️⃣ Temporal + spatial coherence in 4D dynamic prediction 3️⃣ Novel view synthesis for embodied scenes Code: Paper Link: Project page:

Robot Learning needs 4D world models! Robot Learning needs 4D world models! Robot Learning needs 4D world models! We introduce TesserAct, a 4D embodied world model that can simulate how agents interact with the 3D world over time! We achieve this by simply extending a pre-trained 2D video generation model to jointly predict RGB, depth, and surface normals. It enables: 1️⃣ Much better policy learning in the wild 2️⃣ Temporal + spatial coherence in 4D dynamic prediction 3️⃣ Novel view synthesis for embodied scenes Code: Paper Link: Project page:

Chuang Gan

43,265 次观看 • 1 年前

“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

Demis Hassabis says true AGI is a theoretical benchmark set by the human brain — the only proven architecture for general intelligence today’s models are impressive but inconsistent; anyone can find flaws within minutes. "real AGI should be so strong that it would take experts months to spot a weakness"

Demis Hassabis says true AGI is a theoretical benchmark set by the human brain — the only proven architecture for general intelligence today’s models are impressive but inconsistent; anyone can find flaws within minutes. "real AGI should be so strong that it would take experts months to spot a weakness"

Haider.

461,812 次观看 • 1 年前

Demis Hassabis says Gemini 3 is on track and shows the fastest progress in the industry But general intelligence requires more than just the current trajectory It needs better reasoning, stronger memory, and "world model ideas" to solve physical intelligence AGI is still 5–10 years away

Demis Hassabis says Gemini 3 is on track and shows the fastest progress in the industry But general intelligence requires more than just the current trajectory It needs better reasoning, stronger memory, and "world model ideas" to solve physical intelligence AGI is still 5–10 years away

Haider.

143,920 次观看 • 7 个月前

Demis Hassabis is calling on philosophers to step in. He says technologists shouldn't decide AGI's future alone. “We need a new Kant or Wittgenstein to help map out where society should go next.” Politics, ethics, and theology will be essential to navigating a post-AGI world.

Demis Hassabis is calling on philosophers to step in. He says technologists shouldn't decide AGI's future alone. “We need a new Kant or Wittgenstein to help map out where society should go next.” Politics, ethics, and theology will be essential to navigating a post-AGI world.

vitrupo

406,246 次观看 • 1 年前

Demis Hassabis says the Nano Banana Pro system is moving toward AGI for imaging It doesn't just make images, it understands what's in them The goal is to merge this with world models like Genie and SIMA into one system "that might be a candidate for proto-AGI"

Demis Hassabis says the Nano Banana Pro system is moving toward AGI for imaging It doesn't just make images, it understands what's in them The goal is to merge this with world models like Genie and SIMA into one system "that might be a candidate for proto-AGI"

Haider.

50,740 次观看 • 6 个月前

Demis Hassabis says AlphaGo mastered chess in under 24 hours That same self-improvement loop could work in coding and math, where answers are verifiable Systems could close the loop, but once you remove humans from the loop, capabilities can proliferate in ways you didn't want

Demis Hassabis says AlphaGo mastered chess in under 24 hours That same self-improvement loop could work in coding and math, where answers are verifiable Systems could close the loop, but once you remove humans from the loop, capabilities can proliferate in ways you didn't want

Haider.

29,140 次观看 • 5 个月前

Robot policies must be both reliable and highly capable to be useful; the best way to achieve this level of performance is with reinforcement learning. However, for reinforcement learning you are usually stuck between two difficult options: reinforcement in the real world is often risky and expensive, while reinforcement learning in a traditional simulator takes a lot of engineering work and has a persistent sim-to-real gap. What if instead you could train your robot purely in a world model? RISE by Jiazhi Yang et al. uses a compositional world model to predict the future and evaluate progress. This allows for a self-improving pipeline, which learns a world model from real data and then learns how the robot should perform different tasks. This pipeline results in a data-driven way to improve policy performance from real data but without real-world reinforcement learning. Watch Episode #86 of RoboPapers, with Chris Paxton and Jiafei Duan, to learn more!

Robot policies must be both reliable and highly capable to be useful; the best way to achieve this level of performance is with reinforcement learning. However, for reinforcement learning you are usually stuck between two difficult options: reinforcement in the real world is often risky and expensive, while reinforcement learning in a traditional simulator takes a lot of engineering work and has a persistent sim-to-real gap. What if instead you could train your robot purely in a world model? RISE by Jiazhi Yang et al. uses a compositional world model to predict the future and evaluate progress. This allows for a self-improving pipeline, which learns a world model from real data and then learns how the robot should perform different tasks. This pipeline results in a data-driven way to improve policy performance from real data but without real-world reinforcement learning. Watch Episode #86 of RoboPapers, with Chris Paxton and Jiafei Duan, to learn more!

RoboPapers

38,334 次观看 • 28 天前

Demis Hassabis says our brains are likely to be approximate Turing machines. AlphaFold showed why that matters: protein folding looked like a quantum problem, but a classical neural network could model it well enough to solve it. The world may be quantum, but understanding it may only need the right approximation.

Demis Hassabis says our brains are likely to be approximate Turing machines. AlphaFold showed why that matters: protein folding looked like a quantum problem, but a classical neural network could model it well enough to solve it. The world may be quantum, but understanding it may only need the right approximation.

vitrupo

47,068 次观看 • 2 个月前

Without World Models, There Is No AGI. Google Just Proved It. If AGI ever happens, it will not come from bigger chatbots alone. From the very start of this interview, one thing is crystal clear: without world models, we will never reach AGI. And right now, Google is leading with its world simulator Genie 3. Here is the core of what Demis Hassabis explains in this conversation: • World models are the missing core of AGI Hassabis says his deepest long term focus has always been world models and simulations. Not just language. Not just prediction. Actual internal simulations of reality. • LLMs are impressive, but incomplete Language models understand more about the world than expected because human language encodes a lot of reality. Still, language is only a shadow of the real thing. • What text can never fully teach Reality includes things text struggles to express: •3D space and spatial dynamics •Physical causality and mechanics •Sensorimotor experience like movement, force, smell, or balance • Experience beats description To close the gap, AI must learn from interaction and experience, not just static text. That is how you build an internal world simulator. • Why Genie 3 matters With Google DeepMind pushing systems like Genie 3, AI starts to model reality itself, not just talk about it. • Robots and real world assistants depend on this True robotics, smart glasses, and universal assistants require AI that understands the physical world you live in, not just your screen. Bottom line: AGI will not emerge from better text prediction. It will emerge from systems that can simulate, predict, and understand reality itself. Right now, Google is clearly ahead on that path. Curious what you think. Are world models the real AGI unlock, or just another stepping stone?

Without World Models, There Is No AGI. Google Just Proved It. If AGI ever happens, it will not come from bigger chatbots alone. From the very start of this interview, one thing is crystal clear: without world models, we will never reach AGI. And right now, Google is leading with its world simulator Genie 3. Here is the core of what Demis Hassabis explains in this conversation: • World models are the missing core of AGI Hassabis says his deepest long term focus has always been world models and simulations. Not just language. Not just prediction. Actual internal simulations of reality. • LLMs are impressive, but incomplete Language models understand more about the world than expected because human language encodes a lot of reality. Still, language is only a shadow of the real thing. • What text can never fully teach Reality includes things text struggles to express: •3D space and spatial dynamics •Physical causality and mechanics •Sensorimotor experience like movement, force, smell, or balance • Experience beats description To close the gap, AI must learn from interaction and experience, not just static text. That is how you build an internal world simulator. • Why Genie 3 matters With Google DeepMind pushing systems like Genie 3, AI starts to model reality itself, not just talk about it. • Robots and real world assistants depend on this True robotics, smart glasses, and universal assistants require AI that understands the physical world you live in, not just your screen. Bottom line: AGI will not emerge from better text prediction. It will emerge from systems that can simulate, predict, and understand reality itself. Right now, Google is clearly ahead on that path. Curious what you think. Are world models the real AGI unlock, or just another stepping stone?

VraserX e/acc

23,784 次观看 • 6 个月前

Demis Hassabis: Today’s AI is still frozen after training, unable to continuously learn from real world experience, which blocks true general intelligence.

Demis Hassabis: Today’s AI is still frozen after training, unable to continuously learn from real world experience, which blocks true general intelligence.

AshutoshShrivastava

83,264 次观看 • 4 个月前

The best way to get robust, high-quality robot performance is through reinforcement learning; but RL in either the real world or a traditional simulation has lots of limitations. Instead, Jiazhi Yang in RISE does RL in a compositional world model. Learn more ->

The best way to get robust, high-quality robot performance is through reinforcement learning; but RL in either the real world or a traditional simulation has lots of limitations. Instead, Jiazhi Yang in RISE does RL in a compositional world model. Learn more ->

Chris Paxton

33,766 次观看 • 28 天前