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。

This will redefine “remote” work: Fine-Grained Bimanual Manipulation with Low-Cost Hardware, from Tony Zhao at Stanford. Incredible precision.

AI Breakfast

236,359 subscribers

26,427 次观看 • 3 年前 •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

How can robots acquire fine-grained manipulation skills? Introducing ACT: Action Chunking with Transformers 🤖 Key idea: Imitation, but predict actions in chunks instead of one at a time. Here are results with only ~15min of demonstrations, running on low-cost arms:

How can robots acquire fine-grained manipulation skills? Introducing ACT: Action Chunking with Transformers 🤖 Key idea: Imitation, but predict actions in chunks instead of one at a time. Here are results with only ~15min of demonstrations, running on low-cost arms:

Tony Zhao

247,496 次观看 • 3 年前

Enough with LLMs - exciting things are happening in the world of atoms. This is Stanford ALOHA, a low-cost and agile robot platform. The whole system is open-source (!!): hardware design, CAD models for 3D printing, simulator, and training code. Time to graft a physical arm onto GPTs 🦾 Led by my friend Tony Zhao at Stanford, and advised by Chelsea Finn Sergey Levine Vikash Kumar. Project page: If you don't want to 3D print your own components, you can buy the setup at

Enough with LLMs - exciting things are happening in the world of atoms. This is Stanford ALOHA, a low-cost and agile robot platform. The whole system is open-source (!!): hardware design, CAD models for 3D printing, simulator, and training code. Time to graft a physical arm onto GPTs 🦾 Led by my friend Tony Zhao at Stanford, and advised by Chelsea Finn Sergey Levine Vikash Kumar. Project page: If you don't want to 3D print your own components, you can buy the setup at

Jim Fan

818,445 次观看 • 3 年前

Researchers at Stanford University & University of Michigan, using our technology, developed a BCI enabling finger-based quadcopter control—using thought alone. Key highlights: -4 Degrees of Freedom: Decodes fine finger movements, doubling prior systems’ functionality. -Exceptional Precision: 76 targets/min with 100% accuracy. -Broader Impact: Paves the way for gaming, fine motor skill restoration & remote work. Congrats to the exceptional team! 🔗Read more in Nature Medicine:

Researchers at Stanford University & University of Michigan, using our technology, developed a BCI enabling finger-based quadcopter control—using thought alone. Key highlights: -4 Degrees of Freedom: Decodes fine finger movements, doubling prior systems’ functionality. -Exceptional Precision: 76 targets/min with 100% accuracy. -Broader Impact: Paves the way for gaming, fine motor skill restoration & remote work. Congrats to the exceptional team! 🔗Read more in Nature Medicine:

Blackrock Neurotech

21,921 次观看 • 1 年前

There was incredible momentum at ADAPT's launch! Together with AfCFTA Secretariat Official, @WEF, and Tony Blair Institute for Global Change, IOTA is helping to build the systems that will redefine how trade works in Africa and beyond. This is what digital public infrastructure looks like in action.

There was incredible momentum at ADAPT's launch! Together with AfCFTA Secretariat Official, @WEF, and Tony Blair Institute for Global Change, IOTA is helping to build the systems that will redefine how trade works in Africa and beyond. This is what digital public infrastructure looks like in action.

IOTA

1,034,494 次观看 • 7 个月前

Sunday Robotics unveiled its home robot, Memo, a wheeled robot with two arms and pincerlike hands. Training method: Sunday pays remote workers to perform household tasks wearing gloves that resemble Memo's hands. The Mountain View-based company is focused on a full-stack approach, building both hardware and AI models. Beta testing is scheduled to begin next year. Co-founders: - CEO Tony Zhao, ex-Stanford Ph.D. dropout and ex-DeepMind/Tesla, co-created ALOHA project (imitation learning). - CTO Cheng Chi, a Columbia Ph.D., invented the highly-cited work, Diffusion Policy.

Sunday Robotics unveiled its home robot, Memo, a wheeled robot with two arms and pincerlike hands. Training method: Sunday pays remote workers to perform household tasks wearing gloves that resemble Memo's hands. The Mountain View-based company is focused on a full-stack approach, building both hardware and AI models. Beta testing is scheduled to begin next year. Co-founders: - CEO Tony Zhao, ex-Stanford Ph.D. dropout and ex-DeepMind/Tesla, co-created ALOHA project (imitation learning). - CTO Cheng Chi, a Columbia Ph.D., invented the highly-cited work, Diffusion Policy.

The Humanoid Hub

140,597 次观看 • 8 个月前

Bitcoin is repeating the exact same Accumulation → Manipulation → Distribution pattern from 2021 with precision. According to this chart, $BTC will dump to $25,000 in April. Make sure you’re fully prepared for what’s about to come next.

Bitcoin is repeating the exact same Accumulation → Manipulation → Distribution pattern from 2021 with precision. According to this chart, $BTC will dump to $25,000 in April. Make sure you’re fully prepared for what’s about to come next.

Crypto Lens

16,992 次观看 • 3 个月前

🧐Applying world models to improve real-world policy on challenging manipulation tasks used to be considered out of reach. 😌After sustained effort, we’re now seeing encouraging progress. 🚀Thrilled to introduce RISE: Self-Improving Robot Policy with Compositional World Model RISE is, to our knowledge, the first work to use a world model as an effective learning environment for challenging real-world manipulation, enabling policy improvement on tasks that demand high dynamics, dexterity, and precision. Incredible teamwork with Kunyang Lin Hongyang Li Yue Xiangyu Hao Zhao Chonghao Sima

🧐Applying world models to improve real-world policy on challenging manipulation tasks used to be considered out of reach. 😌After sustained effort, we’re now seeing encouraging progress. 🚀Thrilled to introduce RISE: Self-Improving Robot Policy with Compositional World Model RISE is, to our knowledge, the first work to use a world model as an effective learning environment for challenging real-world manipulation, enabling policy improvement on tasks that demand high dynamics, dexterity, and precision. Incredible teamwork with Kunyang Lin Hongyang Li Yue Xiangyu Hao Zhao Chonghao Sima

Jiazhi Yang ✈️ RSS2026

76,440 次观看 • 5 个月前

Meet Tencent HY 3D Studio 1.2 👋 With this major upgrade to our 3D creation pipeline, you can generate assets with sculpt-level detail and fine-grained interactive control. Starting today, the studio is officially open for Public Beta — no application required. Tencent HY 3D Studio 1.2 introduces a suite of features built for creative precision, including: 🧩 PartGen 1.5: High-Precision Component Partitioning — 1536³ Resolution: Boosted from 1024³ for crystal-clear model splitting and ultra-fine detail retention. — Fine-Grained Interaction: Introducing intuitive, brush-based control for precise, manual component editing. — Shape Adherence: Drastically improved geometry integrity for even the most intricate objects. 🎨 Tencent HY 3D 3.1: Sculpt-Level Detail — Enhanced Geometry: Fine-grained details that adapt perfectly to stylized inputs.. — Texture Fidelity: A massive leap in color accuracy and texture fidelity, reflecting your original input with stunning realism. — 8-View Control: Expanded from 4 to 8 input views for ultimate reconstruction accuracy. ✨We’re so excited to see what you create! Try it now:

Meet Tencent HY 3D Studio 1.2 👋 With this major upgrade to our 3D creation pipeline, you can generate assets with sculpt-level detail and fine-grained interactive control. Starting today, the studio is officially open for Public Beta — no application required. Tencent HY 3D Studio 1.2 introduces a suite of features built for creative precision, including: 🧩 PartGen 1.5: High-Precision Component Partitioning — 1536³ Resolution: Boosted from 1024³ for crystal-clear model splitting and ultra-fine detail retention. — Fine-Grained Interaction: Introducing intuitive, brush-based control for precise, manual component editing. — Shape Adherence: Drastically improved geometry integrity for even the most intricate objects. 🎨 Tencent HY 3D 3.1: Sculpt-Level Detail — Enhanced Geometry: Fine-grained details that adapt perfectly to stylized inputs.. — Texture Fidelity: A massive leap in color accuracy and texture fidelity, reflecting your original input with stunning realism. — 8-View Control: Expanded from 4 to 8 input views for ultimate reconstruction accuracy. ✨We’re so excited to see what you create! Try it now:

Tencent Hy

83,314 次观看 • 6 个月前

It’s been a massive week for embodied AI foundation models: the pace of this field is truly staggering. Throwing it back to the SimToolReal work by Kushal and Tyler Lum (Cornell & Stanford labs). In February, they achieved zero-shot tool manipulation across 24 tasks using a single RL policy and a robotic arm fitted with the SharpaWave hand. Watching the robot nail these high-speed in-hand rotations is incredible. The precision is especially impressive when you consider the policy was never trained on these specific objects or tasks. This is what solving the manipulation bottleneck looks like. ⚡️

It’s been a massive week for embodied AI foundation models: the pace of this field is truly staggering. Throwing it back to the SimToolReal work by Kushal and Tyler Lum (Cornell & Stanford labs). In February, they achieved zero-shot tool manipulation across 24 tasks using a single RL policy and a robotic arm fitted with the SharpaWave hand. Watching the robot nail these high-speed in-hand rotations is incredible. The precision is especially impressive when you consider the policy was never trained on these specific objects or tasks. This is what solving the manipulation bottleneck looks like. ⚡️

Sharpa

868,354 次观看 • 2 个月前

We just released our work on robot soccer. I've been working on this for quite some time with my amazing colleagues at DeepMind. It's exciting how deep RL can produce such beautiful behaviors with low-cost robots. Full paper is available at Enjoy!

We just released our work on robot soccer. I've been working on this for quite some time with my amazing colleagues at DeepMind. It's exciting how deep RL can produce such beautiful behaviors with low-cost robots. Full paper is available at Enjoy!

Tuomas Haarnoja

758,785 次观看 • 3 年前

From trial to operations in under two months. Working in partnership with the Royal Air Force we have brought a game-changing new capability to the frontline at pace. Typhoon pilots are now flying with APKWS, our precision, low-cost weapon helping to counter uncrewed air systems in the Middle East.

From trial to operations in under two months. Working in partnership with the Royal Air Force we have brought a game-changing new capability to the frontline at pace. Typhoon pilots are now flying with APKWS, our precision, low-cost weapon helping to counter uncrewed air systems in the Middle East.

BAE Systems

23,505 次观看 • 2 个月前

Repost: My Bad I put The Videos In The Wrong Order 🤭 Redenomination & Revaluation: You Win From The Onset You know what is so beautiful about this? Whatever Iraq decides to do your low cost entry point to buy IQD at a low price will work either way. We know they will drop the 3 zeros. (Again I told you all before Whatever you hold at this time will determine how much you get at your exchange.) Do you want to see what Grok said? I told you all I got this shit covered! Watch 'Revaluation' Video Below This One Please enjoy.

Repost: My Bad I put The Videos In The Wrong Order 🤭 Redenomination & Revaluation: You Win From The Onset You know what is so beautiful about this? Whatever Iraq decides to do your low cost entry point to buy IQD at a low price will work either way. We know they will drop the 3 zeros. (Again I told you all before Whatever you hold at this time will determine how much you get at your exchange.) Do you want to see what Grok said? I told you all I got this shit covered! Watch 'Revaluation' Video Below This One Please enjoy.

Prolotario

37,911 次观看 • 4 个月前

A low of 2c at my station this morning and another fine morning here in Carlow but cloud moving in from Northeast and an interesting day ahead with some heavy showers this afternoon and possibly thunderstorms.

A low of 2c at my station this morning and another fine morning here in Carlow but cloud moving in from Northeast and an interesting day ahead with some heavy showers this afternoon and possibly thunderstorms.

Carlow Weather

26,531 次观看 • 1 年前

🚨🇮🇷🇺🇸Cheap Iranian drones & missiles bleed US THAAD interceptors dry: Indian Navy Vet to Sputnik India ▪️Iran aimed at the sensors that make THAAD work, striking critical radar sites. ▪️With radar “eyes” knocked out, interceptors lose tracking and precision. ▪️Iranian low-cost UAVs and missiles stress US air-defense stockpiles.

🚨🇮🇷🇺🇸Cheap Iranian drones & missiles bleed US THAAD interceptors dry: Indian Navy Vet to Sputnik India ▪️Iran aimed at the sensors that make THAAD work, striking critical radar sites. ▪️With radar “eyes” knocked out, interceptors lose tracking and precision. ▪️Iranian low-cost UAVs and missiles stress US air-defense stockpiles.

Sputnik India

58,489 次观看 • 4 个月前

We might be solving the wrong problem in robotics. That’s what this makes clear. UMI → Universal Manipulation Interface A simple $400 gripper that lets you teach robots by demonstration. You hold it like a tool. Show the task. The robot learns. No teleoperation. No expensive hardware. No robot-specific data. Stanford open-sourced everything → hardware, code, datasets. What stands out to me is the bottleneck. Not algorithms. Data. Teleoperation → ~35 demos/hour UMI → ~111 demos/hour And the data transfers across robots → UR5, Franka, others. The design is surprisingly practical: → GoPro fisheye lens (155° FOV) + mirrors for depth → SLAM + IMU for precise 6DoF tracking → latency matching for dynamic tasks → diffusion policies for multimodal actions Then it scales. Cheng Chi takes this further with Sunday Robotics (with Tony Zhao). A $200 glove → deployed in 500+ homes → ~10 million real-world interactions. Not lab data. Real human behavior. Their robot learns dishes, laundry, espresso → with zero robot-specific data. This is where the shift becomes obvious. From training robots in controlled environments → to learning directly from humans at scale So here’s the real question: Will robotics be unlocked by better models… or by unlocking data? #ArtificialIntelligence #Robotics #AI #Innovation #FutureOfWork

We might be solving the wrong problem in robotics. That’s what this makes clear. UMI → Universal Manipulation Interface A simple $400 gripper that lets you teach robots by demonstration. You hold it like a tool. Show the task. The robot learns. No teleoperation. No expensive hardware. No robot-specific data. Stanford open-sourced everything → hardware, code, datasets. What stands out to me is the bottleneck. Not algorithms. Data. Teleoperation → ~35 demos/hour UMI → ~111 demos/hour And the data transfers across robots → UR5, Franka, others. The design is surprisingly practical: → GoPro fisheye lens (155° FOV) + mirrors for depth → SLAM + IMU for precise 6DoF tracking → latency matching for dynamic tasks → diffusion policies for multimodal actions Then it scales. Cheng Chi takes this further with Sunday Robotics (with Tony Zhao). A $200 glove → deployed in 500+ homes → ~10 million real-world interactions. Not lab data. Real human behavior. Their robot learns dishes, laundry, espresso → with zero robot-specific data. This is where the shift becomes obvious. From training robots in controlled environments → to learning directly from humans at scale So here’s the real question: Will robotics be unlocked by better models… or by unlocking data? #ArtificialIntelligence #Robotics #AI #Innovation #FutureOfWork

Pascal Bornet

186,223 次观看 • 3 个月前

This 40 minute lecture at Stanford by Apoorv Agarwal on the Economics of AI supercycle is worth a watch Apoorv is currently a Partner at Altimeter and is directly involved in some of their key AI investments, including OpenAI and Glean Still find it incredible that the internet gives us access to this level of information directly. A course I will definitely be following along Sourced from MS&E 435 Stanford University

This 40 minute lecture at Stanford by Apoorv Agarwal on the Economics of AI supercycle is worth a watch Apoorv is currently a Partner at Altimeter and is directly involved in some of their key AI investments, including OpenAI and Glean Still find it incredible that the internet gives us access to this level of information directly. A course I will definitely be following along Sourced from MS&E 435 Stanford University

Boring_Business

243,633 次观看 • 2 个月前

Open-source dexterous hands with fingertip sensors! 🪬 ORCA Dexterity just released three dexterous hand models. Their mission: is to democratize robotic hand dexterity. They're sharing progress on exceptional, low-cost hardware and a software layer from low-level control to robotic hand learning. This is the open-source approach to dexterous manipulation. It has 83 taxels per finger with 0.1 N force detection which pretty is impressive for an open-source design. Tactile sensing is critical for dexterous manipulation, knowing contact forces enables gentle grasping, slip detection, and force-controlled assembly. Also 700g weight for the lite version makes it practical for mounting on robot arms without exceeding payload limits. Lower weight means faster movements and lower torque requirements. Open hardware accelerates robotics by letting researchers and builders modify designs for their specific needs without starting from scratch! ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Open-source dexterous hands with fingertip sensors! 🪬 ORCA Dexterity just released three dexterous hand models. Their mission: is to democratize robotic hand dexterity. They're sharing progress on exceptional, low-cost hardware and a software layer from low-level control to robotic hand learning. This is the open-source approach to dexterous manipulation. It has 83 taxels per finger with 0.1 N force detection which pretty is impressive for an open-source design. Tactile sensing is critical for dexterous manipulation, knowing contact forces enables gentle grasping, slip detection, and force-controlled assembly. Also 700g weight for the lite version makes it practical for mounting on robot arms without exceeding payload limits. Lower weight means faster movements and lower torque requirements. Open hardware accelerates robotics by letting researchers and builders modify designs for their specific needs without starting from scratch! ~~ ♻️ Join the weekly robotics newsletter, and never miss any news →

Lukas Ziegler

38,094 次观看 • 4 个月前

Elon Musk: At Tesla, we basically had two different chip programs: one Dojo and one. Dojo on the training side, and then what we call AI4, it's just our inference chip The AI4 is what's currently shipping in all vehicles, and we're finalizing the design of AI5, which will be an immense jump from AI4. By some metrics, the improvement in AI5 will be 40 times better than AI4. So not 40%, 40 times This is because we work so closely at a very fine-grained level on the AI software and the AI hardware. So we know exactly where the limiting factors are. And so effectively the AI hardware and software teams are co-designing the chip Compared to the worst limitation on AI4, which is running the SoftMax operation, we currently have to run SoftMax in around 40 steps in emulation mode, whereas that'll just be done in a few steps natively in AI5 AI5 will also be able to easily handle mixed precision models, so you don't have it, it'll dynamically handle mixed precision. There's a bunch of sort of technical stuff that AI5 will do a lot better In terms of nominal raw compute, it's eight times more compute, about nine times more memory, and roughly five times more memory bandwidth But because we're addressing some core limitations in AI4, you multiply that 8x compute improvement by another 5x improvement because of optimization at a very fine-grained silicon level of things that are currently suboptimal in AI4, that's where you get the 40x improvement

Elon Musk: At Tesla, we basically had two different chip programs: one Dojo and one. Dojo on the training side, and then what we call AI4, it's just our inference chip The AI4 is what's currently shipping in all vehicles, and we're finalizing the design of AI5, which will be an immense jump from AI4. By some metrics, the improvement in AI5 will be 40 times better than AI4. So not 40%, 40 times This is because we work so closely at a very fine-grained level on the AI software and the AI hardware. So we know exactly where the limiting factors are. And so effectively the AI hardware and software teams are co-designing the chip Compared to the worst limitation on AI4, which is running the SoftMax operation, we currently have to run SoftMax in around 40 steps in emulation mode, whereas that'll just be done in a few steps natively in AI5 AI5 will also be able to easily handle mixed precision models, so you don't have it, it'll dynamically handle mixed precision. There's a bunch of sort of technical stuff that AI5 will do a lot better In terms of nominal raw compute, it's eight times more compute, about nine times more memory, and roughly five times more memory bandwidth But because we're addressing some core limitations in AI4, you multiply that 8x compute improvement by another 5x improvement because of optimization at a very fine-grained silicon level of things that are currently suboptimal in AI4, that's where you get the 40x improvement

X Freeze

21,221,727 次观看 • 8 个月前

This unforgettable fight scene from Millionaires' Express showcases two martial arts legends at the peak of their abilities. Sammo Hung brings his incredible agility, power, and creativity, while Cynthia Rothrock counters with her lightning-fast kicks and martial arts precision.

This unforgettable fight scene from Millionaires' Express showcases two martial arts legends at the peak of their abilities. Sammo Hung brings his incredible agility, power, and creativity, while Cynthia Rothrock counters with her lightning-fast kicks and martial arts precision.

Chief Goddiz🇺🇲

28,248 次观看 • 1 个月前

We caught up with Ben Kennedy from NASCAR about their partnership with Anduril Industries: "We couldn't ask for a better partner than Anduril. You look at the technology they're creating, both on the software and hardware side, it's incredible." "Precision, speed, and innovations in technology are what our sport is all about. Excited to see what they'll bring over the next 11 months!"

We caught up with Ben Kennedy from NASCAR about their partnership with Anduril Industries: "We couldn't ask for a better partner than Anduril. You look at the technology they're creating, both on the software and hardware side, it's incredible." "Precision, speed, and innovations in technology are what our sport is all about. Excited to see what they'll bring over the next 11 months!"

TBPN

42,383 次观看 • 11 个月前