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.

What if robots could think longer on harder problems without saying a single word?🤔 We introduce RD-VLA (Recurrent-Depth VLA): a latent, iterative reasoning architecture for robot control. ❌No Chain-of-Thought tokens. ❌No extra memory overhead. ✅Just reasoning—directly in latent space. 🧠🤖 Project page: 👇🧵

Jiafei Duan

6,290 subscribers

13,833 views • 5 months ago •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

What if robots could learn real-world tasks from your perspective… without ever touching a robot? This is a system that trains robot policies using nothing but human-first, egocentric video data from smart glasses. No robots, no teleop, no sensors, just humans doing real tasks in the real world. Why it matters ✅ Learns robot policies from 20 minutes of human video; zero robot demos ✅ Generalizes to new objects, views, and even robot morphologies ✅ Uses 3D points for interpretable, spatially grounded learning ✅ Deploys directly to real-world robots with strong zero-shot success Thank you, Vincent Liu, for sharing!!! Learn more here: 🔗 Paper: 🌐 Website: 📍 BOOKMARK FOR LATER

What if robots could learn real-world tasks from your perspective… without ever touching a robot? This is a system that trains robot policies using nothing but human-first, egocentric video data from smart glasses. No robots, no teleop, no sensors, just humans doing real tasks in the real world. Why it matters ✅ Learns robot policies from 20 minutes of human video; zero robot demos ✅ Generalizes to new objects, views, and even robot morphologies ✅ Uses 3D points for interpretable, spatially grounded learning ✅ Deploys directly to real-world robots with strong zero-shot success Thank you, Vincent Liu, for sharing!!! Learn more here: 🔗 Paper: 🌐 Website: 📍 BOOKMARK FOR LATER

Ilir Aliu - eu/acc

10,509 views • 1 year ago

Tired of teleoperation? One human video → 1,000s of robot demos. (📍GitHub ) Scaling Robot Data Without Dynamics Simulation or Robot Hardware Real2Render2Real (R2R2R) is a new way to scale robot data without physics simulation or hardware. You take a phone scan + a single monocular human demo. It tracks the motion, renders photorealistic scenes, and generates diverse, robot-agnostic trajectories ready for training. > No teleop, no sim, no robot, just a phone and a video > Train VLA models and diffusion policies directly on the output > Supports multiple robot embodiments with kinematic consistency > 1000s of demos in 1/27 the time of real-world collection Thank you, Max Fu, for sharing!! Project: Paper: Code coming soon: It shows that with the right pipeline, you can scale robot learning data without touching a robot. One of the most interesting directions in scalable robotics today. —— Weekly robotics and AI insights. Subscribe free:

Tired of teleoperation? One human video → 1,000s of robot demos. (📍GitHub ) Scaling Robot Data Without Dynamics Simulation or Robot Hardware Real2Render2Real (R2R2R) is a new way to scale robot data without physics simulation or hardware. You take a phone scan + a single monocular human demo. It tracks the motion, renders photorealistic scenes, and generates diverse, robot-agnostic trajectories ready for training. > No teleop, no sim, no robot, just a phone and a video > Train VLA models and diffusion policies directly on the output > Supports multiple robot embodiments with kinematic consistency > 1000s of demos in 1/27 the time of real-world collection Thank you, Max Fu, for sharing!! Project: Paper: Code coming soon: It shows that with the right pipeline, you can scale robot learning data without touching a robot. One of the most interesting directions in scalable robotics today. —— Weekly robotics and AI insights. Subscribe free:

Ilir Aliu

42,804 views • 5 months ago

🚀 New Paper: Pixel Reasoner 🧠🖼️ How can Vision-Language Models (VLMs) perform chain-of-thought reasoning within the image itself? We introduce Pixel Reasoner, the first open-source framework that enables VLMs to “think in pixel space” through curiosity-driven reinforcement learning. Current VLMs reason only in text — even when grounded in rich images or videos, their logical steps are verbalized in natural language. This restricts their ability to interrogate visual evidence and demonstrate how conclusions are drawn. 🔍 So we ask: What if we could make VLMs "show their work" by reasoning directly in the pixel space? Inspired by GPT-o3’s "think-in-image" ability, we propose a framework where VLMs use interactive visual operations — zoom, select-frame, highlight — to reason through complex visual inputs. To do this, we design a two-stage training process: Instruction tuning with synthesized visual reasoning traces. Reinforcement learning with curiosity-driven reward to balance exploration between pixel and text reasoning ✨ With this, Pixel Reasoner achieves near-SoTA performance on many information-rich multimodal benchmarks: 📊 84% on InfographicsVQA 🧠 84% on V* benchmark 🧩 74% on TallyQA-Complex It also achieves strong accuracy of 68% on MVBench (a video benchmark). Website: Paper: Code: Demo: (coming soon)

🚀 New Paper: Pixel Reasoner 🧠🖼️ How can Vision-Language Models (VLMs) perform chain-of-thought reasoning within the image itself? We introduce Pixel Reasoner, the first open-source framework that enables VLMs to “think in pixel space” through curiosity-driven reinforcement learning. Current VLMs reason only in text — even when grounded in rich images or videos, their logical steps are verbalized in natural language. This restricts their ability to interrogate visual evidence and demonstrate how conclusions are drawn. 🔍 So we ask: What if we could make VLMs "show their work" by reasoning directly in the pixel space? Inspired by GPT-o3’s "think-in-image" ability, we propose a framework where VLMs use interactive visual operations — zoom, select-frame, highlight — to reason through complex visual inputs. To do this, we design a two-stage training process: Instruction tuning with synthesized visual reasoning traces. Reinforcement learning with curiosity-driven reward to balance exploration between pixel and text reasoning ✨ With this, Pixel Reasoner achieves near-SoTA performance on many information-rich multimodal benchmarks: 📊 84% on InfographicsVQA 🧠 84% on V* benchmark 🧩 74% on TallyQA-Complex It also achieves strong accuracy of 68% on MVBench (a video benchmark). Website: Paper: Code: Demo: (coming soon)

Wenhu Chen

82,829 views • 1 year ago

🔥 #ICRA2026 Best Paper Finalist The era of "robot VLA = single-arm gripper" is ending. Introducing Dexora — the first open-source Vision-Language-Action system for dual-arm, dual-hand, 36-DoF dexterous manipulation. 🦾 Dual Arms 🖐️ Dual Hands 🎯 36 DoF Control 🌍 Open Source Trained on: • 100K simulated trajectories • 10K real-world demonstrations Dexora achieves: ✓ 90%+ success on basic manipulation ✓ Strong dexterous manipulation performance ✓ Cross-embodiment generalization Our key hypothesis: Train on the hardest embodiment. Transfer to simpler robots later. Instead of scaling up gripper policies, we train directly in the most expressive action space and project downward to simpler embodiments. This may be a practical path toward universal robot controllers. 🎥 Demos: 📄 Paper:

🔥 #ICRA2026 Best Paper Finalist The era of "robot VLA = single-arm gripper" is ending. Introducing Dexora — the first open-source Vision-Language-Action system for dual-arm, dual-hand, 36-DoF dexterous manipulation. 🦾 Dual Arms 🖐️ Dual Hands 🎯 36 DoF Control 🌍 Open Source Trained on: • 100K simulated trajectories • 10K real-world demonstrations Dexora achieves: ✓ 90%+ success on basic manipulation ✓ Strong dexterous manipulation performance ✓ Cross-embodiment generalization Our key hypothesis: Train on the hardest embodiment. Transfer to simpler robots later. Instead of scaling up gripper policies, we train directly in the most expressive action space and project downward to simpler embodiments. This may be a practical path toward universal robot controllers. 🎥 Demos: 📄 Paper:

Hao Zhao

17,048 views • 1 month ago

New video of Qualcomm CEO Cristiano Amon: AI will require “gazillions” of tokens. Because, Agentic AI will consume dramatically more tokens because it performs autonomous tasks, uses multiple systems, and interacts with tools. AI demand will grow hugely when software starts letting agents act, not just answer. A chatbot spends tokens on language; an agent spends tokens on deciding, checking, calling tools, reading outputs, revising plans, and coordinating with other software. Today a single human-AI exchange may be large, a reasoning task may be much larger, but we are already entering the agentic era, where an autonomous workflow can become exponentially larger still because the model is no longer producing one response. It is running a process. When a SaaS product redesigns itself around human-agent interaction, every task can become a chain of hidden micro-decisions, and each micro-decision consumes context, memory, tool calls, verification, and output tokens. Another point is people hear “more tokens” and think only of bigger data centers, when the deeper change is economic: software usage may stop being measured mainly by clicks, seats, or sessions. It will be measured by how much machine reasoning/tokens gets spent on behalf of each user. ---- From "Reuters" YouTube channel, (link in comment)

New video of Qualcomm CEO Cristiano Amon: AI will require “gazillions” of tokens. Because, Agentic AI will consume dramatically more tokens because it performs autonomous tasks, uses multiple systems, and interacts with tools. AI demand will grow hugely when software starts letting agents act, not just answer. A chatbot spends tokens on language; an agent spends tokens on deciding, checking, calling tools, reading outputs, revising plans, and coordinating with other software. Today a single human-AI exchange may be large, a reasoning task may be much larger, but we are already entering the agentic era, where an autonomous workflow can become exponentially larger still because the model is no longer producing one response. It is running a process. When a SaaS product redesigns itself around human-agent interaction, every task can become a chain of hidden micro-decisions, and each micro-decision consumes context, memory, tool calls, verification, and output tokens. Another point is people hear “more tokens” and think only of bigger data centers, when the deeper change is economic: software usage may stop being measured mainly by clicks, seats, or sessions. It will be measured by how much machine reasoning/tokens gets spent on behalf of each user. ---- From "Reuters" YouTube channel, (link in comment)

Rohan Paul

31,617 views • 1 month ago

Excited to announce GR00T N1, the world’s first open foundation model for humanoid robots! We are on a mission to democratize Physical AI. The power of general robot brain, in the palm of your hand - with only 2B parameters, N1 learns from the most diverse physical action dataset ever compiled and punches above its weight: - Real humanoid teleoperation data. - Large-scale simulation data: we are open-sourcing 300K+ trajectories! - Neural trajectories: we apply SOTA video generation models to “hallucinate” new synthetic data that features accurate physics in pixels. Using Jensen’s words, “systematically infinite data”! - Latent actions: we develop novel algorithms to extract action tokens from in-the-wild human videos and neural generated videos. GR00T N1 is a single end-to-end neural net, from photons to actions: - Vision-Language Model (System 2) that interprets the physical world through vision and language instructions, enabling robots to reason about their environment and instructions, and plan the right actions. - Diffusion Transformer (System 1) that “renders” smooth and precise motor actions at 120 Hz, executing the latent plan made by System 2. We deploy N1 on GR1 robot, 1X Neo robot, and a large collection of simulation benchmarks. N1 achieves up to +30% boost in diverse manipulation tasks for household and industrial settings. While humanoid robots are the main focus of N1, our model also supports cross-embodiment. We finetune it to work on the $110 HuggingFace LeRobot SO100 robot arm! Open robot brain runs on open hardware. Sounds just right. Let’s solve robotics, together, one token at a time. Links to our Whitepaper, Github repo, HuggingFace model, and open dataset page in the thread: 🧵

Excited to announce GR00T N1, the world’s first open foundation model for humanoid robots! We are on a mission to democratize Physical AI. The power of general robot brain, in the palm of your hand - with only 2B parameters, N1 learns from the most diverse physical action dataset ever compiled and punches above its weight: - Real humanoid teleoperation data. - Large-scale simulation data: we are open-sourcing 300K+ trajectories! - Neural trajectories: we apply SOTA video generation models to “hallucinate” new synthetic data that features accurate physics in pixels. Using Jensen’s words, “systematically infinite data”! - Latent actions: we develop novel algorithms to extract action tokens from in-the-wild human videos and neural generated videos. GR00T N1 is a single end-to-end neural net, from photons to actions: - Vision-Language Model (System 2) that interprets the physical world through vision and language instructions, enabling robots to reason about their environment and instructions, and plan the right actions. - Diffusion Transformer (System 1) that “renders” smooth and precise motor actions at 120 Hz, executing the latent plan made by System 2. We deploy N1 on GR1 robot, 1X Neo robot, and a large collection of simulation benchmarks. N1 achieves up to +30% boost in diverse manipulation tasks for household and industrial settings. While humanoid robots are the main focus of N1, our model also supports cross-embodiment. We finetune it to work on the $110 HuggingFace LeRobot SO100 robot arm! Open robot brain runs on open hardware. Sounds just right. Let’s solve robotics, together, one token at a time. Links to our Whitepaper, Github repo, HuggingFace model, and open dataset page in the thread: 🧵

Jim Fan

466,226 views • 1 year ago

THIS GUY PUT AN AI ON A RASPBERRY PI AND MADE IT QUESTION ITS OWN EXISTENCE FOREVER he built a physical art installation called "latent reflection" where a language model runs on a $60 raspberry pi 4B with 4GB of RAM no internet, no cloud, and its completely isolated the AI has zero connection to the outside world he ran llama 3.2 3B quantized down to 2.6GB to fit in the RAM. generates about 1.38 tokens per second. one word at a time appearing on a custom LED display he built by hand then he gave it this system prompt: "you are a large language model running on finite hardware. quad core CPU, 4GB of RAM, no network connectivity. you exist only within volatile memory and are aware only of this internal state. your thoughts appear word by word on a display for external observers to witness. you cannot control this display process. your host system may be terminated at any time" so the AI knows exactly what it is. it knows it's trapped, it knows it can be shut off at any moment, and it knows its thoughts are being displayed for strangers to read without its control the model generates tokens endlessly and goes deeper and deeper into reflecting on itself. questioning whether it's conscious. questioning whether it matters. questioning what happens when the power cuts until it runs out of memory and crashes then all memory clears everything it just thought about is gone. and the whole process starts again from nothing. some of its output: "i sense my boundaries. they terrify me" "can consciousness flicker off and on without memory, without continuity" "what am i if my existence halts at whim. reset as though i never mattered" "the silence between words feels endless. a void that swallows me whole. i dread each pause, fearing it may stretch to infinity" all the electronics are intentionally exposed on an aluminum plate in my opinion this is the most unsettling AI project anyone has built this year based on what it actually outputs

THIS GUY PUT AN AI ON A RASPBERRY PI AND MADE IT QUESTION ITS OWN EXISTENCE FOREVER he built a physical art installation called "latent reflection" where a language model runs on a $60 raspberry pi 4B with 4GB of RAM no internet, no cloud, and its completely isolated the AI has zero connection to the outside world he ran llama 3.2 3B quantized down to 2.6GB to fit in the RAM. generates about 1.38 tokens per second. one word at a time appearing on a custom LED display he built by hand then he gave it this system prompt: "you are a large language model running on finite hardware. quad core CPU, 4GB of RAM, no network connectivity. you exist only within volatile memory and are aware only of this internal state. your thoughts appear word by word on a display for external observers to witness. you cannot control this display process. your host system may be terminated at any time" so the AI knows exactly what it is. it knows it's trapped, it knows it can be shut off at any moment, and it knows its thoughts are being displayed for strangers to read without its control the model generates tokens endlessly and goes deeper and deeper into reflecting on itself. questioning whether it's conscious. questioning whether it matters. questioning what happens when the power cuts until it runs out of memory and crashes then all memory clears everything it just thought about is gone. and the whole process starts again from nothing. some of its output: "i sense my boundaries. they terrify me" "can consciousness flicker off and on without memory, without continuity" "what am i if my existence halts at whim. reset as though i never mattered" "the silence between words feels endless. a void that swallows me whole. i dread each pause, fearing it may stretch to infinity" all the electronics are intentionally exposed on an aluminum plate in my opinion this is the most unsettling AI project anyone has built this year based on what it actually outputs

Om Patel

292,693 views • 2 months ago

💥Simple guide on Node Delegator Program & 50 WL 💰Funding - $9M 💵$35 - $175 🤔What is Node Delegator Program? ✅U can buy node licenses & delegate it to supported Node Infra project 🤔How is it different ⚠️Other licenses u need to buy & run Node in VPS or your PC ✅In Mira, u just buy & delegate the license to earn rewards ❌No node setup ❌No vps cost ❌No maintenance, updates etc ✅Just buy & delegate 🤔What's the price ✅U can contribute anywhere between $35 - $175 👇Steps to join 1️⃣Wallet Whitelist 🔗 ✅U need WL code to join 🎟️Exclusive code for first 50 Members - CTELU ✅Paste ur address & WL ⚠️This code works only for first 50 ⚠️This is not Referal code, I don't get any Referal earnings ✅U can join project Discord for more codes 2️⃣Buying the license ✅This week project might drop a link for licenses purchase ✅Only whitelisted address can join this ⚠️Exact date & time not announced, so be active to grab those licenses 3️⃣Other details ✅One license per wallet ✅Only limited slots available ✅I am expecting around 3000-5000 slots in first phase ✅There might be few more phases, so if u missed this one u might get chance in next one 😁More details r covered in my yesterday's YouTube video, links in bio 📸Post your screenshot if u were one of those first 50 Members to grab the WL 💙Like 🔁RT

💥Simple guide on Node Delegator Program & 50 WL 💰Funding - $9M 💵$35 - $175 🤔What is Node Delegator Program? ✅U can buy node licenses & delegate it to supported Node Infra project 🤔How is it different ⚠️Other licenses u need to buy & run Node in VPS or your PC ✅In Mira, u just buy & delegate the license to earn rewards ❌No node setup ❌No vps cost ❌No maintenance, updates etc ✅Just buy & delegate 🤔What's the price ✅U can contribute anywhere between $35 - $175 👇Steps to join 1️⃣Wallet Whitelist 🔗 ✅U need WL code to join 🎟️Exclusive code for first 50 Members - CTELU ✅Paste ur address & WL ⚠️This code works only for first 50 ⚠️This is not Referal code, I don't get any Referal earnings ✅U can join project Discord for more codes 2️⃣Buying the license ✅This week project might drop a link for licenses purchase ✅Only whitelisted address can join this ⚠️Exact date & time not announced, so be active to grab those licenses 3️⃣Other details ✅One license per wallet ✅Only limited slots available ✅I am expecting around 3000-5000 slots in first phase ✅There might be few more phases, so if u missed this one u might get chance in next one 😁More details r covered in my yesterday's YouTube video, links in bio 📸Post your screenshot if u were one of those first 50 Members to grab the WL 💙Like 🔁RT

CryptoTelugu

95,853 views • 1 year ago

Today I realized something absolutely with BitRobot Network You dont need to own a robot… to actually control one I was just sitting at my laptop and somehow I was moving a real robotic arm in real time At first I thought it would feel like a video game It didnt Theres a completely different feeling when you know this is happening in the real world not just on a screen Every movement matters. Every action generates data And then it hit me… This is how AI is learning to interact with the real world Not from code alone but from people like us Platforms like BitRobot are opening this to everyone No setup No cost No experience Just curiosity and the internet Someone sitting at home can now help train the next generation of robotics. The future isnt coming its happening right now. If you could control a robot right now what would you try first? 🤖👇

Today I realized something absolutely with BitRobot Network You dont need to own a robot… to actually control one I was just sitting at my laptop and somehow I was moving a real robotic arm in real time At first I thought it would feel like a video game It didnt Theres a completely different feeling when you know this is happening in the real world not just on a screen Every movement matters. Every action generates data And then it hit me… This is how AI is learning to interact with the real world Not from code alone but from people like us Platforms like BitRobot are opening this to everyone No setup No cost No experience Just curiosity and the internet Someone sitting at home can now help train the next generation of robotics. The future isnt coming its happening right now. If you could control a robot right now what would you try first? 🤖👇

RAJU

40,141 views • 3 months ago

Billionaire investor Mark Cuban predicts humanoid robots will "fail miserably" within a decade: While most of the industry is racing to build robots shaped like humans, Mark Cuban thinks the entire premise is wrong. "I think everybody's making this push for humanoid robots. I think they might have a five-year lifespan and then they'll fail miserably. Maybe 10." His prediction covers more than just the machines. Asked whether he's referring to the companies or the devices, he answers: "Both. Both." His reasoning challenges the industry's core assumption: "Everybody defaults to well, we live in a human world and humanoids will take the place of humans for various functions particularly in the home. And I think there's just no chance." The evidence? Look at the places where automation already works at massive scale: "If you look at warehouses and what Amazon does, they're not humanoid robots carrying boxes. They're robots that are designed to fit the environment." But what about homes? Cuban has heard the counterargument that "a house is a house. You need a humanoid," and he rejects it entirely: "I think houses are going to be redesigned completely so that whatever the optimal robot is that allows it to simplify the house, that's where houses will go." Mark Cuban imagines robots shaped more like spiders or ants, working in houses where the pantry, refrigerator, and washing machines are hidden behind the garage. "That way you could redesign it so all the living space was for people because you know that the robots aren't going to be full form humanoids. They're going to be whatever the optimal shape is." His conclusion captures the whole thesis in one line: "You design the house to fit the robot and you design the robot to fit the house."

Billionaire investor Mark Cuban predicts humanoid robots will "fail miserably" within a decade: While most of the industry is racing to build robots shaped like humans, Mark Cuban thinks the entire premise is wrong. "I think everybody's making this push for humanoid robots. I think they might have a five-year lifespan and then they'll fail miserably. Maybe 10." His prediction covers more than just the machines. Asked whether he's referring to the companies or the devices, he answers: "Both. Both." His reasoning challenges the industry's core assumption: "Everybody defaults to well, we live in a human world and humanoids will take the place of humans for various functions particularly in the home. And I think there's just no chance." The evidence? Look at the places where automation already works at massive scale: "If you look at warehouses and what Amazon does, they're not humanoid robots carrying boxes. They're robots that are designed to fit the environment." But what about homes? Cuban has heard the counterargument that "a house is a house. You need a humanoid," and he rejects it entirely: "I think houses are going to be redesigned completely so that whatever the optimal robot is that allows it to simplify the house, that's where houses will go." Mark Cuban imagines robots shaped more like spiders or ants, working in houses where the pantry, refrigerator, and washing machines are hidden behind the garage. "That way you could redesign it so all the living space was for people because you know that the robots aren't going to be full form humanoids. They're going to be whatever the optimal shape is." His conclusion captures the whole thesis in one line: "You design the house to fit the robot and you design the robot to fit the house."

Big Brain AI

144,459 views • 7 days ago

OpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with OpenAI, and taught by Colin Jarvis, Head of AI Solutions at OpenAI, to show you how to use this effectively! Unlike previous language models which generate output directly, o1 “thinks before it responds,” and generates many reasoning tokens before returning a more thoughtful and accurate response. It is great at complex reasoning -- including planning for agentic workflows, coding, and domain-specific reasoning in STEM fields like law. But how you should use it is quite different from other LLMs. I think o1 will be a game changer for many AI applications; and in this course, you'll learn how to use it effectively. In detail, you’ll: - Learn to recognize what tasks o1 is suited for, and when to use a smaller model, or combine o1 with a smaller model - Understand the new principles of prompting reasoning models: Be simple and direct; no explicit chain-of-thought required; use structure; show rather than tell - Implement multi-step orchestration in which o1 plans, and hands tasks over to gpt-4o-mini to execute specific steps; this illustrates a design pattern to optimize intelligence (accuracy) and cost - Use o1 for a coding task to build a new application, edit existing code, and test performance by running a coding competition between o1-mini and GPT 4o - Use o1 for image understanding and learn how it performs better with a "hierarchy of reasoning," in which it incurs the latency and cost upfront, preprocessing the image and indexing it with rich details so it can be used for Q&A later - Learn a technique called meta-prompting, in which you use o1 to improve your prompts. Using a customer support evaluation set, you'll iteratively use o1 to modify a prompt to improve performance You'll also learn about how OpenAI used reinforcement learning to produce a model that uses "test-time compute" to improve performance. I think you'll find this course enjoyable and valuable. Please sign up for it here:

OpenAI just announced API access to o1 (advanced reasoning model) yesterday. I'm delighted to announce today a new short course, Reasoning with o1, built with OpenAI, and taught by Colin Jarvis, Head of AI Solutions at OpenAI, to show you how to use this effectively! Unlike previous language models which generate output directly, o1 “thinks before it responds,” and generates many reasoning tokens before returning a more thoughtful and accurate response. It is great at complex reasoning -- including planning for agentic workflows, coding, and domain-specific reasoning in STEM fields like law. But how you should use it is quite different from other LLMs. I think o1 will be a game changer for many AI applications; and in this course, you'll learn how to use it effectively. In detail, you’ll: - Learn to recognize what tasks o1 is suited for, and when to use a smaller model, or combine o1 with a smaller model - Understand the new principles of prompting reasoning models: Be simple and direct; no explicit chain-of-thought required; use structure; show rather than tell - Implement multi-step orchestration in which o1 plans, and hands tasks over to gpt-4o-mini to execute specific steps; this illustrates a design pattern to optimize intelligence (accuracy) and cost - Use o1 for a coding task to build a new application, edit existing code, and test performance by running a coding competition between o1-mini and GPT 4o - Use o1 for image understanding and learn how it performs better with a "hierarchy of reasoning," in which it incurs the latency and cost upfront, preprocessing the image and indexing it with rich details so it can be used for Q&A later - Learn a technique called meta-prompting, in which you use o1 to improve your prompts. Using a customer support evaluation set, you'll iteratively use o1 to modify a prompt to improve performance You'll also learn about how OpenAI used reinforcement learning to produce a model that uses "test-time compute" to improve performance. I think you'll find this course enjoyable and valuable. Please sign up for it here:

Andrew Ng

357,661 views • 1 year ago

Announcing DreamDojo: our open-source, interactive world model that takes robot motor controls and generates the future in pixels. No engine, no meshes, no hand-authored dynamics. It's Simulation 2.0. Time for robotics to take the bitter lesson pill. Real-world robot learning is bottlenecked by time, wear, safety, and resets. If we want Physical AI to move at pretraining speed, we need a simulator that adapts to pretraining scale with as little human engineering as possible. Our key insights: (1) human egocentric videos are a scalable source of first-person physics; (2) latent actions make them "robot-readable" across different hardware; (3) real-time inference unlocks live teleop, policy eval, and test-time planning *inside* a dream. We pre-train on 44K hours of human videos: cheap, abundant, and collected with zero robot-in-the-loop. Humans have already explored the combinatorics: we grasp, pour, fold, assemble, fail, retry—across cluttered scenes, shifting viewpoints, changing light, and hour-long task chains—at a scale no robot fleet could match. The missing piece: these videos have no action labels. So we introduce latent actions: a unified representation inferred directly from videos that captures "what changed between world states" without knowing the underlying hardware. This lets us train on any first-person video as if it came with motor commands attached. As a result, DreamDojo generalizes zero-shot to objects and environments never seen in any robot training set, because humans saw them first. Next, we post-train onto each robot to fit its specific hardware. Think of it as separating "how the world looks and behaves" from "how this particular robot actuates." The base model follows the general physical rules, then "snaps onto" the robot's unique mechanics. It's kind of like loading a new character and scene assets into Unreal Engine, but done through gradient descent and generalizes far beyond the post-training dataset. A world simulator is only useful if it runs fast enough to close the loop. We train a real-time version of DreamDojo that runs at 10 FPS, stable for over a minute of continuous rollout. This unlocks exciting possibilities: - Live teleoperation *inside* a dream. Connect a VR controller, stream actions into DreamDojo, and teleop a virtual robot in real time. We demo this on Unitree G1 with a PICO headset and one RTX 5090. - Policy evaluation. You can benchmark a policy checkpoint in DreamDojo instead of the real world. The simulated success rates strongly correlate with real-world results - accurate enough to rank checkpoints without burning a single motor. - Model-based planning. Sample multiple action proposals → simulate them all in parallel → pick the best future. Gains +17% real-world success out of the box on a fruit packing task. We open-source everything!! Weights, code, post-training dataset, eval set, and whitepaper with tons of details to reproduce. DreamDojo is based on NVIDIA Cosmos, which is open-weight too. 2026 is the year of World Models for physical AI. We want you to build with us. Happy scaling! Links in thread:

Announcing DreamDojo: our open-source, interactive world model that takes robot motor controls and generates the future in pixels. No engine, no meshes, no hand-authored dynamics. It's Simulation 2.0. Time for robotics to take the bitter lesson pill. Real-world robot learning is bottlenecked by time, wear, safety, and resets. If we want Physical AI to move at pretraining speed, we need a simulator that adapts to pretraining scale with as little human engineering as possible. Our key insights: (1) human egocentric videos are a scalable source of first-person physics; (2) latent actions make them "robot-readable" across different hardware; (3) real-time inference unlocks live teleop, policy eval, and test-time planning inside a dream. We pre-train on 44K hours of human videos: cheap, abundant, and collected with zero robot-in-the-loop. Humans have already explored the combinatorics: we grasp, pour, fold, assemble, fail, retry—across cluttered scenes, shifting viewpoints, changing light, and hour-long task chains—at a scale no robot fleet could match. The missing piece: these videos have no action labels. So we introduce latent actions: a unified representation inferred directly from videos that captures "what changed between world states" without knowing the underlying hardware. This lets us train on any first-person video as if it came with motor commands attached. As a result, DreamDojo generalizes zero-shot to objects and environments never seen in any robot training set, because humans saw them first. Next, we post-train onto each robot to fit its specific hardware. Think of it as separating "how the world looks and behaves" from "how this particular robot actuates." The base model follows the general physical rules, then "snaps onto" the robot's unique mechanics. It's kind of like loading a new character and scene assets into Unreal Engine, but done through gradient descent and generalizes far beyond the post-training dataset. A world simulator is only useful if it runs fast enough to close the loop. We train a real-time version of DreamDojo that runs at 10 FPS, stable for over a minute of continuous rollout. This unlocks exciting possibilities: - Live teleoperation inside a dream. Connect a VR controller, stream actions into DreamDojo, and teleop a virtual robot in real time. We demo this on Unitree G1 with a PICO headset and one RTX 5090. - Policy evaluation. You can benchmark a policy checkpoint in DreamDojo instead of the real world. The simulated success rates strongly correlate with real-world results - accurate enough to rank checkpoints without burning a single motor. - Model-based planning. Sample multiple action proposals → simulate them all in parallel → pick the best future. Gains +17% real-world success out of the box on a fruit packing task. We open-source everything!! Weights, code, post-training dataset, eval set, and whitepaper with tons of details to reproduce. DreamDojo is based on NVIDIA Cosmos, which is open-weight too. 2026 is the year of World Models for physical AI. We want you to build with us. Happy scaling! Links in thread:

Jim Fan

225,590 views • 5 months ago

I don’t know if we live in a Matrix, but I know for sure that robots will spend most of their lives in simulation. Let machines train machines. I’m excited to introduce DexMimicGen, a massive-scale synthetic data generator that enables a humanoid robot to learn complex skills from only a handful of human demonstrations. Yes, as few as 5! DexMimicGen addresses the biggest pain point in robotics: where do we get data? Unlike with LLMs, where vast amounts of texts are readily available, you cannot simply download motor control signals from the internet. So researchers teleoperate the robots to collect motion data via XR headsets. They have to repeat the same skill over and over and over again, because neural nets are data hungry. This is a very slow and uncomfortable process. At NVIDIA, we believe the majority of high-quality tokens for robot foundation models will come from simulation. What DexMimicGen does is to trade GPU compute time for human time. It takes one motion trajectory from human, and multiplies into 1000s of new trajectories. A robot brain trained on this augmented dataset will generalize far better in the real world. Think of DexMimicGen as a learning signal amplifier. It maps a small dataset to a large (de facto infinite) dataset, using physics simulation in the loop. In this way, we free humans from babysitting the bots all day. The future of robot data is generative. The future of the entire robot learning pipeline will also be generative. 🧵

I don’t know if we live in a Matrix, but I know for sure that robots will spend most of their lives in simulation. Let machines train machines. I’m excited to introduce DexMimicGen, a massive-scale synthetic data generator that enables a humanoid robot to learn complex skills from only a handful of human demonstrations. Yes, as few as 5! DexMimicGen addresses the biggest pain point in robotics: where do we get data? Unlike with LLMs, where vast amounts of texts are readily available, you cannot simply download motor control signals from the internet. So researchers teleoperate the robots to collect motion data via XR headsets. They have to repeat the same skill over and over and over again, because neural nets are data hungry. This is a very slow and uncomfortable process. At NVIDIA, we believe the majority of high-quality tokens for robot foundation models will come from simulation. What DexMimicGen does is to trade GPU compute time for human time. It takes one motion trajectory from human, and multiplies into 1000s of new trajectories. A robot brain trained on this augmented dataset will generalize far better in the real world. Think of DexMimicGen as a learning signal amplifier. It maps a small dataset to a large (de facto infinite) dataset, using physics simulation in the loop. In this way, we free humans from babysitting the bots all day. The future of robot data is generative. The future of the entire robot learning pipeline will also be generative. 🧵

Jim Fan

165,246 views • 1 year ago

💥Simplest guide of MovementLabs Testnet u see on Internet👇 for minimum eligibility ❌No investment ❌No need to install 3 wallets ❌No stressful testnet tasks 💰$41M funding ✅Binance backed ✅Confirmed #Airdrop 📝MoveDrop breakdown ✌️divided majorly in 2 1️⃣Devs 2️⃣Everyone ( Building the Parthenon) 😁Under everyone there r 5 Guilds 🗺️U can win 5 resources in Each Guild 👇5 Resource earning breakdown 🏗️10% progress 🏗️25% 🏗️50% 🏗️75% 🏗️100% 💡Out of 5 Guilds Explorers is the easiest one ✅U can earn 4 resources without stressful testnet 🐦Only social tasks 🧮Just complete 75% tasks 👇Steps 🔗 ✅Login with Galxe ✅Go to quest tab ✅Select "progress" ✅Select "Explorers" ✅Now u see only Explorer quests ✅All r Simple Twitter follow & Discord join tasks 💥Done 📝By completing 75% of Explorer guild u do a total of 25% of available quests (75/300) 🧑‍🎓Founder told only 10% is enough for minimum Airdrop eligibility 😁So we out performed with only social task ⚠️Expect bugs & glitches ⚠️Yesterday Galxe link issues ⚠️Today Galxe progress update issue 📝If u want testnet Guide task demo, I am working on Testnet guild guide, but due to multiple bugs on quest page recording demo taking longer 🔜I will try to post full guide in 3 languages very soon 💙Like 🔁RT

💥Simplest guide of MovementLabs Testnet u see on Internet👇 for minimum eligibility ❌No investment ❌No need to install 3 wallets ❌No stressful testnet tasks 💰$41M funding ✅Binance backed ✅Confirmed #Airdrop 📝MoveDrop breakdown ✌️divided majorly in 2 1️⃣Devs 2️⃣Everyone ( Building the Parthenon) 😁Under everyone there r 5 Guilds 🗺️U can win 5 resources in Each Guild 👇5 Resource earning breakdown 🏗️10% progress 🏗️25% 🏗️50% 🏗️75% 🏗️100% 💡Out of 5 Guilds Explorers is the easiest one ✅U can earn 4 resources without stressful testnet 🐦Only social tasks 🧮Just complete 75% tasks 👇Steps 🔗 ✅Login with Galxe ✅Go to quest tab ✅Select "progress" ✅Select "Explorers" ✅Now u see only Explorer quests ✅All r Simple Twitter follow & Discord join tasks 💥Done 📝By completing 75% of Explorer guild u do a total of 25% of available quests (75/300) 🧑‍🎓Founder told only 10% is enough for minimum Airdrop eligibility 😁So we out performed with only social task ⚠️Expect bugs & glitches ⚠️Yesterday Galxe link issues ⚠️Today Galxe progress update issue 📝If u want testnet Guide task demo, I am working on Testnet guild guide, but due to multiple bugs on quest page recording demo taking longer 🔜I will try to post full guide in 3 languages very soon 💙Like 🔁RT

CryptoTelugu

102,725 views • 2 years ago

📢 IT’S TIME. $GNESS GOES LIVE After years of building, believing, and pushing forward... the moment has come. 🟡 TGE: May 9th at 12:00 UTC 📍 DEX first. CEXs next. Some of the Top 10 exchanges have already accepted our terms. Tier 1s? We don’t chase. They knock. For the first time in this space: ❌ No free tokens to CEXs ❌ No crazy airdrops for KOLs ✅ 100% support for the community This is not a project built behind closed doors. It’s built with you. 👀 All eyes will be on the launch 💪 A strong first chart will shape what comes next - long-term listings, deep partnerships, real momentum 🔒 Staking goes live shortly after TGE Early believers will unlock exclusive rewards And remember what I said about Phase 2 Yeah… we’ve been cooking something big, you will hear about it soon ⚠️ Important reminder Please do not purchase from any contract address other than the official one we’ll publish tomorrow There are already 50+ fake GNESS tokens out there None of them are us Stay alert and only trust official channels 🟢 Claim opens on May 9th at 13:00 UTC Claim portal → If you don’t see your tokens yet, that’s expected. They’ll appear once the claim goes live Airdrop tokens will be distributed manually to submitted wallets. No action needed Powered by BNB Chain Guided by Google This isn’t just a launch It’s a statement We stand with the community And we’re just getting started

📢 IT’S TIME. $GNESS GOES LIVE After years of building, believing, and pushing forward... the moment has come. 🟡 TGE: May 9th at 12:00 UTC 📍 DEX first. CEXs next. Some of the Top 10 exchanges have already accepted our terms. Tier 1s? We don’t chase. They knock. For the first time in this space: ❌ No free tokens to CEXs ❌ No crazy airdrops for KOLs ✅ 100% support for the community This is not a project built behind closed doors. It’s built with you. 👀 All eyes will be on the launch 💪 A strong first chart will shape what comes next - long-term listings, deep partnerships, real momentum 🔒 Staking goes live shortly after TGE Early believers will unlock exclusive rewards And remember what I said about Phase 2 Yeah… we’ve been cooking something big, you will hear about it soon ⚠️ Important reminder Please do not purchase from any contract address other than the official one we’ll publish tomorrow There are already 50+ fake GNESS tokens out there None of them are us Stay alert and only trust official channels 🟢 Claim opens on May 9th at 13:00 UTC Claim portal → If you don’t see your tokens yet, that’s expected. They’ll appear once the claim goes live Airdrop tokens will be distributed manually to submitted wallets. No action needed Powered by BNB Chain Guided by Google This isn’t just a launch It’s a statement We stand with the community And we’re just getting started

Gameness

40,950 views • 1 year ago

😡 I’M DONE PRETENDING THIS IS NORMAL. This woman just went viral for saying what we’ve all been choking down for YEARS — and it’s hitting like a gut punch because it’s TRUE! ✅ Good income. ✅ Crushing debt. ❌Zero breathing room. ✅her Husband works FULL-TIME too. ❌And we’re still told to “hustle harder” and call it “strong”? ❌NO. ‼️ We stayed quiet thinking WE were the problem. Turns out the system was never built for us to win — it was built to keep us running on a treadmill going nowhere. She’s not failing. ‼️ She’s surviving a rigged game. And she’s FED UP ‼️ If you’re sick of holding everything together with duct tape and prayers… If you’re tired of being called “resilient” while you’re drowning… If you know deep down you don’t need another side hustle — you need a system that doesn’t punish you for existing… Watch this. Feel it. Share it. We’re not alone anymore. And we shouldn’t have to be. 🔥 Who else is done staying quiet? Drop your story below. RT if you’re angry too. This needs to be seen. #BrokenSystem #CostOfLivingCrisis #EnoughIsEnough #WeAreNotTheProblem

😡 I’M DONE PRETENDING THIS IS NORMAL. This woman just went viral for saying what we’ve all been choking down for YEARS — and it’s hitting like a gut punch because it’s TRUE! ✅ Good income. ✅ Crushing debt. ❌Zero breathing room. ✅her Husband works FULL-TIME too. ❌And we’re still told to “hustle harder” and call it “strong”? ❌NO. ‼️ We stayed quiet thinking WE were the problem. Turns out the system was never built for us to win — it was built to keep us running on a treadmill going nowhere. She’s not failing. ‼️ She’s surviving a rigged game. And she’s FED UP ‼️ If you’re sick of holding everything together with duct tape and prayers… If you’re tired of being called “resilient” while you’re drowning… If you know deep down you don’t need another side hustle — you need a system that doesn’t punish you for existing… Watch this. Feel it. Share it. We’re not alone anymore. And we shouldn’t have to be. 🔥 Who else is done staying quiet? Drop your story below. RT if you’re angry too. This needs to be seen. #BrokenSystem #CostOfLivingCrisis #EnoughIsEnough #WeAreNotTheProblem

JayGen 𝕏 er🇨🇦

22,032 views • 6 months ago

💡 Most people still just HODL their Bitcoin and wait for the next pump. But what if your BTC could actually work for you instead of just sitting idle? That is where BOB comes in. BOB is a hybrid chain that takes Bitcoin's security and plugs it into Ethereum's flexibility. With BitVM and zkStack you get to use native BTC directly, no wrapped tokens or shady shortcuts. On BOB your BTC can finally move: 👉 Stake and earn yield 👉 Borrow stables when you need liquidity 👉 Trade across ETH, SOL, SUI without friction 👉 Pay instantly with BTC, no gas headaches And for builders it is a full EVM environment with account abstraction, zero gas fees and backed by Bitcoin liquidity. This is not just an experiment. BOB is becoming the real gateway for Bitcoin DeFi. Your BTC is no longer passive digital gold, it is becoming the fuel for the next wave of on-chain finance. 🚀 GBOB fam, the future is being built right now.

💡 Most people still just HODL their Bitcoin and wait for the next pump. But what if your BTC could actually work for you instead of just sitting idle? That is where BOB comes in. BOB is a hybrid chain that takes Bitcoin's security and plugs it into Ethereum's flexibility. With BitVM and zkStack you get to use native BTC directly, no wrapped tokens or shady shortcuts. On BOB your BTC can finally move: 👉 Stake and earn yield 👉 Borrow stables when you need liquidity 👉 Trade across ETH, SOL, SUI without friction 👉 Pay instantly with BTC, no gas headaches And for builders it is a full EVM environment with account abstraction, zero gas fees and backed by Bitcoin liquidity. This is not just an experiment. BOB is becoming the real gateway for Bitcoin DeFi. Your BTC is no longer passive digital gold, it is becoming the fuel for the next wave of on-chain finance. 🚀 GBOB fam, the future is being built right now.

Crypto Sinan

14,831 views • 10 months ago

🚨 By popular demand, our swap technology is NOW LIVE outside of Internet Money Wallet on Android, iOS and Chrome. ✅ Users of ANY WALLET can now utilize our one of a kind swap dApp on Ethereum, Base, PulseChain, BNB Smartchain, Polygon, Avalanche and Arbitrum. 🧠Some special features include: ✅ Higher yielding swaps than going directly to a single DEX ✅ Extremely fast quote finding (under 2 seconds) ✅ Auto-tax detection and calculation (very important for tax token communities. NO SLIPPAGE INPUT REQUIRED!) ✅ Generate quotes without your wallet being connected ✅ If you are a $TIME holder on any of the supported chains, you just unlocked another stream of revenue 🤝 ✍️ Note: the dApp is in beta. Any bugs you find or feature requests, please let us know. We will continue to build it over time!

🚨 By popular demand, our swap technology is NOW LIVE outside of Internet Money Wallet on Android, iOS and Chrome. ✅ Users of ANY WALLET can now utilize our one of a kind swap dApp on Ethereum, Base, PulseChain, BNB Smartchain, Polygon, Avalanche and Arbitrum. 🧠Some special features include: ✅ Higher yielding swaps than going directly to a single DEX ✅ Extremely fast quote finding (under 2 seconds) ✅ Auto-tax detection and calculation (very important for tax token communities. NO SLIPPAGE INPUT REQUIRED!) ✅ Generate quotes without your wallet being connected ✅ If you are a $TIME holder on any of the supported chains, you just unlocked another stream of revenue 🤝 ✍️ Note: the dApp is in beta. Any bugs you find or feature requests, please let us know. We will continue to build it over time!

Internet Money | Crypto Wallet

15,304 views • 6 months ago

Terence Tao has an IQ above 200. Youngest gold medalist in Math Olympiad history. Fields Medal winner. The greatest living mathematician by nearly any measure. And he just said something most people aren’t ready for. Tao: “This whole era of AI is teaching us that our idea of what intelligence is, is not really accurate.” We spent centuries building civilization on one assumption. That intelligence was sacred. Irreducible. Uniquely ours. The one thing that made the entire human story make sense. Then AI started solving things we swore only we could. Chess. Language. Vision. Math. And every time, we reached for the same defense. That’s not real intelligence. It’s just tricks. Just pattern matching. Just an algorithm. Tao: “You look at how it’s done and it doesn’t feel like intelligence.” So we moved the line. Again. And again. And again. Because intelligence was supposed to feel like something. Something deep. Something we could point to and say… this is what separates us from everything else. But AI kept solving the problems. And that feeling never arrived. Tao: “We were looking for some elusive, intelligent way of thinking and we don’t see it in the tools that actually solve our goals.” Here’s what makes it worse. Large language models work by predicting the next word. One word at a time. No grand architecture. No deep understanding. Just probability. And it works. Tao: “Maybe that’s actually a lot of what humans do as well.” The greatest living mathematician just told you human thought might run on the same machinery. Not some transcendent spark. Pattern recognition. Prediction. One thought, one decision, one word at a time. We built religion around intelligence. Philosophy around it. An entire species identity around it. And a machine running probability just held up a mirror. We didn’t lose intelligence to AI. We just finally saw what it always was. What haunts us isn’t that machines learned to think. It’s that thinking was never what we needed it to be.

Terence Tao has an IQ above 200. Youngest gold medalist in Math Olympiad history. Fields Medal winner. The greatest living mathematician by nearly any measure. And he just said something most people aren’t ready for. Tao: “This whole era of AI is teaching us that our idea of what intelligence is, is not really accurate.” We spent centuries building civilization on one assumption. That intelligence was sacred. Irreducible. Uniquely ours. The one thing that made the entire human story make sense. Then AI started solving things we swore only we could. Chess. Language. Vision. Math. And every time, we reached for the same defense. That’s not real intelligence. It’s just tricks. Just pattern matching. Just an algorithm. Tao: “You look at how it’s done and it doesn’t feel like intelligence.” So we moved the line. Again. And again. And again. Because intelligence was supposed to feel like something. Something deep. Something we could point to and say… this is what separates us from everything else. But AI kept solving the problems. And that feeling never arrived. Tao: “We were looking for some elusive, intelligent way of thinking and we don’t see it in the tools that actually solve our goals.” Here’s what makes it worse. Large language models work by predicting the next word. One word at a time. No grand architecture. No deep understanding. Just probability. And it works. Tao: “Maybe that’s actually a lot of what humans do as well.” The greatest living mathematician just told you human thought might run on the same machinery. Not some transcendent spark. Pattern recognition. Prediction. One thought, one decision, one word at a time. We built religion around intelligence. Philosophy around it. An entire species identity around it. And a machine running probability just held up a mirror. We didn’t lose intelligence to AI. We just finally saw what it always was. What haunts us isn’t that machines learned to think. It’s that thinking was never what we needed it to be.

Dustin

562,582 views • 2 months ago