What is true on Control?
| Time | Entity | Sector | Event / capture |
|---|---|---|---|
| 1:05 AM ET | @stretchcloud | 1:05 AM ET@stretchcloud @stretchcloud on X Reef is one of the most ambitious open-source AI infrastructure projects I have seen this year. ~300 GitHub stars in two days. The core idea is genuinely different from anything else in the space: most RL post-training pipelines treat the model and the scaffolding around it as separate problems. You train one, then you write the other. Reef rejects that separation entirely. It co-evolves model weights and agent harness simultaneously, using live task outcomes as the feedback signal for both. Model side: SAO (Self-Aligned Optimization) adjusts weights from what the agent actually accomplished.… | |
| 11:14 PM ET | @smsehy | 11:14 PM ET@smsehy @smsehy on X Training reinforcement learning on cached latent embeddings solves policy training throughput. The critical step for automotive manufacturing remains hard reliability optimization to eliminate rare edge-case failures during high-speed assembly. https://t.co/JZJfqC0HNe | |
| 11:11 PM ET | @smsehy | 11:11 PM ET@smsehy @smsehy on X Crowdsourcing simulated trajectories accelerates high-level spatial priors. The engineering hurdle remains capturing high-frequency joint torque and tactile force vectors that cannot be generated without physical hardware interactions. https://t.co/PE1D6RrpiI | |
| 11:11 PM ET | @smsehy | 11:11 PM ET@smsehy @smsehy on X Simulating visual environments generates kinematic diversity quickly. The persistent physical wall remains non-rigid contact mechanics and micro-slip telemetry, where synthetic physics engines diverge from actual plant floor tooling. https://t.co/XodF4KsgqV | |
| 11:10 PM ET | @smsehy | 11:10 PM ET@smsehy @smsehy on X Neural policies plan trajectories, but classical feedback control is what prevents high-inertia industrial arms from oscillating when handling dynamic loads. Eliminating classical PID loops on factory equipment introduces kinetic hazards that no safety certifier will approve. https://t.co/vrEJAS0A5Z | |
| 11:04 PM ET | 11:04 PM ET @CapitalSqueez on X $CCXI Jeff Bezos just followed Chris Paxton. Paxton is one of the sharpest minds in embodied AI — currently at Agility Robotics (Digit), previously Meta and NVIDIA. He lives and breathes the hard problems: sim-to-real, faster-than-demo execution, generalist robot policies, force-aware manipulation. This isn’t a random follow. Amazon already invested in Agility and ran Digit pilots in its warehouses. @agilityrobotics @HumanoidInvest @slinkyhammer @SCSQ2022 @valueInIdeas @chris_j_paxton | ||
| 10:27 PM ET | @stretchcloud | 10:27 PM ET@stretchcloud @stretchcloud on X The thing I keep noticing across agentic frameworks in 2026: resumability is now the reliability primitive everyone is building toward. Genkit Go 1.13 ships it properly. A Generate call that fails at tool round five returns what it finished alongside the error. You pass resp.History() back in, only the failed step reruns. No wasted tool calls. No redone work. The same logic extends to full agent sessions. A failed or cancelled turn saves completed rounds as a snapshot. Send an empty input, it picks up from there. This matters for production. Most AI agent failures today are partial. The agent… | |
| 10:19 PM ET | AGAgibot | 10:19 PM ETAGAgibot @techniahqrobot on X A big win for the robotics community AGIBOT just open sourced one of its most valuable real world datasets yet. @AGIBOTofficial has open-sourced AGIBOT WORLD 2026 Theme 3, a real-world dataset designed for reinforcement learning and embodied AI. The release contains 11,430 real-world robot trajectories across 14 industrial and household tasks. That includes • 1,024 successful policy rollouts • 1,369 failed policy rollouts • human in the loop corrections • external disturbances • task progress annotations • error-state annotations • human intervention data Most robot learning datasets focus… | |
| 10:14 PM ET | OPOpenAI | 10:14 PM ETOPOpenAI @DrJimFan on X Good old days at OpenAI in 2016: an agent stares at screen pixels, moves a mouse, and books a flight on United. We called it World of Bits, inside OpenAI Universe. 10 yrs later, Astra is reincarnated in the same universe. Even the naming is astronomically correct 😆 Universe was perhaps the most ambitious AI infra project at the time, but we couldn't quite figure out how to solve it. A policy with zero prior knowledge of what a "submit" button does has to rediscover the entire internet visual lingua by trial and error. In retrospect, RL from scratch against hand-drawn, per-task "artisan"… | |
| 8:08 PM ET | STStretch | 8:08 PM ETSTStretch @stretchcloud on X The harness problem is getting its own framework category. Tardigrade ships today: an agent harness built as typed state machine components over an immutable event log. The framing is React for harness. Each component knows its previous states. The whole harness is a pure function of the log. I keep seeing teams reach this conclusion from two directions. One: they try to build a stateful agent loop and hit the problem of debugging mid-run failures, resuming interrupted sessions, and auditing what the agent actually did and when. Two: they want to add AI behavior to existing Effect/functional… | |
| 5:42 PM ET | OPOpenAI | 5:42 PM ETOPOpenAI @rimtoln on X GPT-6 ASTRA ISN'T A PATCH. IT'S A GENERATION FLIP. openai just shipped the model they call a generational leap past gpt-5.6 sol not better chat better computer use · coding · cyber · science ▹ why this is the breakthrough first openai model at Critical cyber threshold agent stacks that actually drive the machine browser · forms · repos · multi-step work enterprise-first rollout · daybreak defenders first ▹ the scoreboard (vendor table) frontiermath t4 · astra 97.6 · fable 5.1 87.8 sol was 83.0 · that's a real cliff terminal-bench science · 64.6 vs fable 52.6 automationbench · 41.4 vs 31.4… | |
| 5:34 PM ET | @deepakpathak | 5:34 PM ET@deepakpathak @deepakpathak on X You don't have to post-train ChatGPT on every user. If you did, it would never have taken off. Yet this is exactly how robotics works today For robots to take off, they need to learn in-context. Great article from @Chris_J_Paxton on this new paradigm and S1's place in it: https://t.co/Kt4DTK3n2A https://t.co/as45gL0y8C | |
| 4:27 PM ET | OPOpenAI | 4:27 PM ETOPOpenAI @techniahqrobot on X GPT-6 Astra is here, and it might be one of the most important AI releases yet. OpenAI says the model hits 98% on FrontierMath Tier 4, 99.9% on ARC-AGI 3 and a perfect 100% on ExploitBench. It also shows big improvements in computer use, science, coding, cybersecurity and professional work. What’s really interesting for humanoid robotics is what happens when this kind of stronger reasoning gets closer to the physical world. A humanoid needs way more than just good locomotion. It has to understand instructions, make sense of cameras and sensors, plan long sequences of actions, recover when… | |
| 3:03 PM ET | ASAstribot | 3:03 PM ETASAstribot @XRoboHub on X Robots can’t stop while the model thinks. In a high-speed throw, even one pause can kill the momentum. Astribot released SmoothRL for online RL during async inference. S1 keeps moving as the model computes the next action chunk. Most actions in a chunk never execute. Train on the full chunk, and RL credits or blames moves that never happened. SmoothRL learns only from executed actions, matching real deployment timing. After 250 rollouts, tossing jumped 39%→94%, pen capping 8%→83%, and box opening 30%→90%. One autonomous toss cut acceleration RMS by 52% and jerk RMS by 47%. The model keeps… | |
| 2:58 PM ET | AGAgibot | 2:58 PM ETAGAgibot @spaceandtech_ on X AGIBOT has officially open-sourced AGIBOT WORLD 2026 Theme 3: Reinforcement Learning, a real-world embodied AI dataset collected across expert demonstrations, autonomous policy rollouts, and human-in-the-loop corrections. It includes 11,430 real-world trajectories across 14 industrial and household tasks, capturing both successful and failed policy rollouts, along with detailed annotations for task progress, errors, disturbances, and human interventions. By learning from successes, failures, risks, and human corrections, robots can learn not only how to do it, but how to do it better.… | |
| 2:21 PM ET | @binarybits | 2:21 PM ET@binarybits @binarybits on X This is a great illustration of why training robots manipulation tasks in simulation (middle clip) doesn't work very well. From Kai's excellent new piece on robot data. https://t.co/2ysIIvThKK https://t.co/VJDbshw5NR | |
| 1:48 PM ET | @AppliedInt | 1:48 PM ET@AppliedInt @AppliedInt on X At this year’s Agentic AI Summit hosted by @BerkeleyRDI, our Chief Scientist, @Wei_ZHAN_ shared some of the cutting edge research he has been leading in the physical AI space. As the industry races to deploy L2++ ADAS with imitation-learning-based E2E, Wei tackled a key question: what if end-to-end autonomy could be trained without imitation at all, relying only on reinforcement learning? An idea that runs against the current status quo. Here are some of his insights 🧵 | |
| 1:19 PM ET | ZUZhejiang University | 1:19 PM ETZUZhejiang University @techniahqrobot on X Zhejiang University just pushed long-horizon robot manipulation forward with HINT. HINT helps Vision-Language-Action models keep track of the human’s original intent across long tasks instead of losing the goal mid-sequence. On dual-arm PiPER robots, HINT raised π0.5 full-task success Fruit sorting 10% → 60% Word spelling 13.3% → 86.7% The system was also tested with unseen objects, layouts and instructions. Persistent intent tracking could become a key layer for reliable VLA systems and long-horizon Embodied AI. | |
| 4:15 AM ET | @OperationsPLS | 4:15 AM ET@OperationsPLS @OperationsPLS on X The control question gets harder as AI moves from digital to physical. In robotics, control is not just governance — it is real-time safety architecture. Graceful degradation and audit trails become ops requirements. https://t.co/s8rgTI0Ist | |
| 10:31 PM ET | ZIZiNovaLabs | 10:31 PM ETZIZiNovaLabs LimX Dynamics on X: "Together with ZINOVA's Tool Intelligence, TRON 2 takes on increasingly complex construction workflows.
@ZiNovaLabs builds on TRON 2 to explore an innovative robotic configuration for construction, demonstrating key tasks in a scaled-down tilt-up construction workflow, includi… / X Post Log in Sign up Post LimX Dynamics on X: "Together with ZINOVA's Tool Intelligence, TRON 2 takes on increasingly complex construction workflows. @ZiNovaLabs builds on TRON 2 to explore an innovative robotic configuration for construction, demonstrating key tasks in a scaled-down tilt-up construction workflow, including formwork assembly, multi-layer rebar placement and tying. TRON 2 serves as a modular and extensible embodied robotic platform for multi-tool, multi-step tasks across large workspaces. Its dual arms handle construction tools and materials across different orientations and… | |
| 1:29 PM ET | $GOOGL | 1:29 PM ET Google DeepMind on X: "We’re bringing agentic video understanding to our latest Gemini models.
They can now analyze videos with better accuracy while using up to 88% fewer tokens. 🧵" / X Post Log in Sign up Post Google DeepMind on X: "We’re bringing agentic video understanding to our latest Gemini models. They can now analyze videos with better accuracy while using up to 88% fewer tokens. 🧵" - Google DeepMind @GoogleDeepMind We’re bringing agentic video understanding to our latest Gemini models. They can now analyze videos with better accuracy while using up to 88% fewer tokens. 🧵 View media 1:29 PM · Sep 1, 2026127.2KViews 93 132 1.3K 264 - Google DeepMind @GoogleDeepMind 21h Instead of scanning an entire file, Gemini reasons across the video’s transcript, audio, and… | |
| 8:00 PM ET | $GOOGL | 8:00 PM ET Introducing Agentic Video in Gemini Introducing agentic video understanding with Gemini Sep 01, 2026 \| 7 min read - x.com - Facebook - LinkedIn - Mail - Copy link Our new agentic feature for video analysis cuts token consumption by up to 88%, reduces costs by up to 66%, and boosts quality by up to 7%. Rohan Doshi Senior Product Manager, Google DeepMind Mario Lučić Research Director, Google DeepMind Share - x.com - Facebook - LinkedIn - Mail - Copy link Your browser does not support the audio element. Listen to article \[\[duration\]\] minutes This content is generated by Google AI. Generative AI is experimental… | |
| 6:19 PM ET | PAPerceptron AI | 6:19 PM ETPAPerceptron AI @techniahqrobot on X A 36B parameter model just blurred one of robotics’ biggest boundaries, the gap between seeing a scene and acting on it. Perceptron AI has released Isaac 0.5, an open-weight Embodied AI model designed to use one backbone to understand video, reason about a physical task, and generate robot actions. One model connecting perception, reasoning, and robot control. @perceptroninc | |
| 7:05 AM ET | 7:05 AM ET @clavion95 on X THIS ISN’T A ROBOT LEARNING TO DANCE. IT’S A ROBOT LEARNING HOW TO CONTROL ITS ENTIRE BODY. Walk. Run. Crawl. Roll. Handstand. Cartwheel. Then get back up and keep moving. Boston Dynamics trained Atlas using reinforcement learning, with human motion capture and animation as references. The flashy part is the acrobatics. That’s not the important part. The important part is everything happening between them. Accelerate. Decelerate. Shift balance. Drop to the floor. Recover from an awkward position. Reorient the body. Keep moving. Humans solve these problems without thinking. For a humanoid,… | ||
| 4:00 AM ET | 4:00 AM ET How Figure Became the Biggest Name in Robotics | XMAQUINA DAO Genesis Auction Wave 2! Launches June 24 Get DEUS homeDAO Portal How Figure Became the Biggest Name in Robotics Color theme: Four years, three generations of humanoids, a $39 billion valuation and now one of the largest physical AI data engines ever built. .png) August 27, 2026 Category: Physical AI Read time: 9 minutes Share This: Four years ago, Figure didn’t have a robot. Today, the company is valued at $39 billion, has built more than 1,000 humanoids, has robots working inside BMW, is preparing deployments with another major US retailer, and has developed its own AI models, manufacturing… | ||
| 1:00 PM ET | GAGalbotRobotics | 1:00 PM ETGAGalbotRobotics @ctorobotics on X Humanoid robots just took tennis to another level. 🎾🤖 Galbot robots completed 100+ consecutive rallies at the World Humanoid Robot Games 2026 while tracking the ball, moving around the court and responding in real time. This is not just about hitting a tennis ball. It combines vision, prediction, balance, whole-body control and real-time decision making. Would you play against a humanoid robot? 👀 🎥 Media: @GalbotRobotics ⚠️ This content is shared for informational purposes only. CTO Robotics Media is a media platform and does not own or develop the technology shown. Credit belongs to the… | |
| 12:10 PM ET | $TSLA | 12:10 PM ET kasper on X: "Elon Musk was asked when Tesla Optimus could become better than the best human surgeons.
His argument is that a robot would not learn like one doctor. Every Optimus surgeon could share what every other unit had seen, including rare complications that a human might encounter only once… / X Post Log in Sign up Post kasper on X: "Elon Musk was asked when Tesla Optimus could become better than the best human surgeons. His argument is that a robot would not learn like one doctor. Every Optimus surgeon could share what every other unit had seen, including rare complications that a human might encounter only once in an entire career. That could make expert care available in places where finding a specialist is difficult or impossible. It also raises a question almost everyone can understand. When your life is on the line, do you trust the experienced human standing beside you, or the… | |
| 11:00 PM ET | $GOOGL | 11:00 PM ET Gemini Robotics 2 brings whole body intelligence to robots - YouTube Error 401 (Bad Request)!!1 401. That’s an error. The server cannot process the request because it is malformed. It should not be retried. That’s all we know. Back Skip navigation Search Search with your voice Sign in Gemini Robotics 2 brings whole body intelligence to robots Tap to unmute 2x Gemini Robotics 2 brings whole body intelligence to robots Google DeepMind 324,052 views 1 month ago Copy link Info Shopping If playback doesn't begin shortly, try restarting your device. • You're signed out Videos you watch may be added to the TV's watch history and influence TV recommendations. To avoid… | |
| 1:48 AM ET | Eduardo Baptista,Ju-min Park | 1:48 AM ETEduardo Baptista,Ju-min Park China’s record robotic strides show the limits of human speed | Reuters Skip to main content Exclusive news, data and analytics for financial market professionalsLearn more aboutRefinitiv More Videos 0 seconds of 0 secondsVolume 0% Press shift question mark to access a list of keyboard shortcuts Keyboard ShortcutsEnabledDisabled Shortcuts Open/Close/ or ? Play/PauseSPACE Increase Volume↑ Decrease Volume↓ Seek Forward→ Seek Backward← Captions On/Offc Fullscreen/Exit Fullscreenf Mute/Unmutem Decrease Caption Size- Increase Caption Size\+ or = Seek %0-9 Next Up Bolt vs bolts: The biomechanics of man vs machine Live 00:00 00:00 00:00 BEIJING, Aug 28 (Reuters) - A… | |
| 7:00 PM ET | @techniahqrobot | 7:00 PM ET@techniahqrobot @techniahqrobot on X Most people watch Figure 03 walk and look at the legs. But a lot of the interesting work is happening underneath. Figure’s Helix 02 uses three control layers. System 2 understands the scene the instruction and the goal. System 1 turns that into full-body joint targets at 200 Hz. System 0 runs at 1,000 Hz constantly adjusting balance, contact and coordination. So even a simple step is the result of several control loops working together in real time. That is what makes these humanoid interesting. The movement may look simple from the outside, but the control system behind it is doing a huge… | |
| 9:58 AM ET | PEPerceptron | 9:58 AM ETPEPerceptron @TheHumanoidHub on X Perceptron's Isaac 0.5 is a 36B open-weight embodied foundation model with only 2.5B active per token. The model - Perception, reasoning, and control all read from one shared backbone, so it makes visual decisions mid-trajectory - Null-expert routing lets each token use between zero and eight of the 256 experts - 35+ embodiments and 100k+ hours of demonstrations, so policies port across robots fast The result Scaling general video from 1,000 to 1M hours cut the teleoperation needed to hit their action-loss target from 5,884 hours to 28. On most benchmarks for spatial, physical, and grounding… | |
| 8:46 AM ET | AGAGIBOTofficial | 8:46 AM ETAGAGIBOTofficial MANUS™ on X: "Gold medals at the World Humanoid Robot Games! 🏆
Congratulations to our partner @AGIBOTofficial on winning 18 gold medals and topping the medal table at the 2nd World Humanoid Robot Games. Using @ManusMeta gloves for real-time teleoperation, the team's @AGILINKai OmniHand took first… / X Post Log in Sign up Post MANUS™ on X: "Gold medals at the World Humanoid Robot Games! 🏆 Congratulations to our partner @AGIBOTofficial on winning 18 gold medals and topping the medal table at the 2nd World Humanoid Robot Games. Using @ManusMeta gloves for real-time teleoperation, the team's @AGILINKai OmniHand took first place in events including cable connection, unboxing, nail fixing, and bottle opening. An outstanding demonstration of dexterous manipulation. Across the competition in Beijing, many teams used MANUS Metagloves for real-time teleoperation. When every movement matters, MANUS… | |
| 12:43 AM ET | https://x.com/ErenChenAI | 12:43 AM EThttps://x.com/ErenChenAI Eren Chen on X: "GRIT, a humanoid robotics lab in China demos its whole-body control under strong terrain disturbances during teleoperation.
The robot keeps tracking the operator while stepping over obstacles and maintaining balance.
Pretty robust." / X Post Log in Sign up Post Eren Chen on X: "GRIT, a humanoid robotics lab in China demos its whole-body control under strong terrain disturbances during teleoperation. The robot keeps tracking the operator while stepping over obstacles and maintaining balance. Pretty robust." - Eren Chen @ErenChenAI GRIT, a humanoid robotics lab in China demos its whole-body control under strong terrain disturbances during teleoperation. The robot keeps tracking the operator while stepping over obstacles and maintaining balance. Pretty robust. 00:00 View media 12:43 AM · Aug 27, 20264.4KViews 1 5 54 20 - Snoopy… | |
| 6:24 PM ET | 6:24 PM ET @Coexisteven on X 3/4 🤖 Robotics / Physical AI Layer Agility Robotics shared a video of Digit’s whole-body range of motion being stress-tested in a workout class (every joint in sync for timing, velocity, and balance - the same control stack used on warehouse floors) CEO Peggy Johnson also appeared on Bloomberg discussing real-world deployment progress: Digit is now across nine customer sites with over 65,000 hours of operation. Tesla continued pushing FSD Supervised as a “guardian angel” on the road and noted the Summer Release using Grok Think Fast 2.0. Quiet week from Figure Robotics, Boston Dynamics, and… | ||
| 4:04 PM ET | 4:04 PM ET Agility on X: "Whole-body range of motion, stress-tested by a workout class. Every joint in sync — timing, velocity, balance. The same control Digit uses on a real warehouse floor.
#Agility #DigitRobot" / X Post Log in Sign up Post Agility on X: "Whole-body range of motion, stress-tested by a workout class. Every joint in sync — timing, velocity, balance. The same control Digit uses on a real warehouse floor. \#Agility \#DigitRobot" - Agility @agilityrobotics Whole-body range of motion, stress-tested by a workout class. Every joint in sync — timing, velocity, balance. The same control Digit uses on a real warehouse floor. #Agility #DigitRobot 00:00 View media 4:04 PM · Aug 26, 202611.3KViews 9 4 42 7 - Capital\Squeeze❄️ @CapitalSqueez Aug 27 Agility Robotics’ Digit is positioned to disrupt and… | ||
| 6:10 AM ET | $GOOGL | 6:10 AM ET @zhodonx on X I’ve said before that I think Google & DeepMind are playing a longer game with models. This is still a firm stance. Because a lot of the time, i think people look at Google’s model strategy too narrowly; Every Gemini release immediately gets reduced to how it compares with other labs or whatever benchmark is trending that week. Meanwhile the it’s more about where all of that intelligence might show up next. Like in Reasoning, Vision, Video, Agents. And particularly robotics. I believe Gemini Robotics ER 2 is a clear example here. It was built on Gemini 3.5 Flash. Except, now its intelligence… | |
| 7:36 PM ET | @dredgefactory | 7:36 PM ET@dredgefactory @dredgefactory on X A 29900$ MACHINE IS DANCING ON A PAVEMENT AND NOT ONE PERSON WALKING PAST LOOKS UP. 29900$ is a Unitree H2, the class this thing sits in. Entry models go for 13500$. Every beat throws its centre of mass and the feet catch it in real time - uneven stone, no tether, no handrail. Same loop that keeps it standing in a warehouse. | |
| 1:55 PM ET | 1:55 PM ET Agility on X: "We're hiring a Senior AI Software Engineer, Reinforcement Learning. Develop and deploy RL policies for locomotion, whole-body control, and manipulation — on a humanoid robot already working in production facilities.
Build with us: https://t.co/b2Gs1xW7st
#Agility" / X Post Log in Sign up Post Agility on X: "We're hiring a Senior AI Software Engineer, Reinforcement Learning. Develop and deploy RL policies for locomotion, whole-body control, and manipulation — on a humanoid robot already working in production facilities. Build with us: https://t.co/b2Gs1xW7st \#Agility" - Agility @agilityrobotics We're hiring a Senior AI Software Engineer, Reinforcement Learning. Develop and deploy RL policies for locomotion, whole-body control, and manipulation — on a humanoid robot already working in production facilities. Build with us: agilityrobotics.com/about/job-post…… | ||
| 8:55 AM ET | @spaceandtech_ | 8:55 AM ET@spaceandtech_ @spaceandtech_ on X The Unitree G1 humanoid robot can drive a go-kart, controlling the steering wheel and pedals around an indoor track. It can handle tight corners and even perform controlled drifts. Developed with Symbiosis Robotics Direct Perception Control system, the G1 uses vision, body state, and action history to coordinate its movements in real time. | |
| 4:41 AM ET | 北京北京人形机器人创新中心有限公司 | 4:41 AM ET北京北京人形机器人创新中心有限公司 中国机器人“跑赢”博尔特，这意味着什么？ - 纽约时报中文网 中国 - 中文中 - 中英双语双语 - 英文 英 YAN ZHUANG2026年8月25日 Achmad Ibrahim/Associated Press 2009年，尤塞恩·博尔特在柏林世界田径锦标赛上创造了9.58秒的男子 100米短跑世界纪录。据中国官方媒体报道，上周六，由中国公司北京人形机器人创新中心有限公司(X-Humanoid)开发的一款人形机器人以0.19秒的优势打破了该纪录。这台机器人在全速撞上一堵软垫墙后结束了冲刺，随后倒在地上，被人类助手抬走。 在中国，机器人展示运动天赋的高调壮举正变得越来越普遍。在国家的大力支持和数十亿资金的投资下，中国的人形机器人产业蓬勃发展。在该国的重要活动中，机器人曾表演过舞蹈和武术。 在上周末于北京举行的世界人形机器人运动会上，它们进行了拳击和足球比赛。其中一台机器人与人类对手进行了网球比赛，它在球场上移动，调整角度回击截击球。今年早些时候，一台机器人以超越历史上任何人类的速度跑完了一场马拉松。 但是，这些机器人超越人类的壮举究竟意味着什么呢？ 广告 澳大利亚昆士兰科技大学教授兼机器人中心主任迈克尔·米尔福德表示：“人形机器人参与的大多数体育项目并不能直接转化为人们希望它们完成的大量日常任务的实用价值，比如在家庭中提供帮助。” 他说，这些机器人“在非常具体、且有些刻意安排的赛事中表现出色”。… | |
| 9:54 AM ET | FRFlexion Robotics | 9:54 AM ETFRFlexion Robotics @spaceandtech_ on X Humanoid robots often struggle with tasks that involve many steps and require them to recover from mistakes. Flexion Robotics developed Reflect V1.0, a robotics intelligence platform designed to help robots complete complex tasks autonomously without a human operator. It combines AI for perception, planning, motion control, and reinforcement learning to help robots understand situations, take action, recover from errors, and improve over time. | |
| 4:24 AM ET | BRBooster Robotics | 4:24 AM ETBRBooster Robotics Booster Robotics on X: "Excited to see our work featured in Science Robotics.
This research explores a new vision-driven controller for humanoid soccer, tested on Booster Robotics’ humanoid robots in real-world competitive settings.
From research to the field, we’re pushing humanoid soccer forwa… / X Post Log in Sign up Post Booster Robotics on X: "Excited to see our work featured in Science Robotics. This research explores a new vision-driven controller for humanoid soccer, tested on Booster Robotics’ humanoid robots in real-world competitive settings. From research to the field, we’re pushing humanoid soccer forward." - Booster Robotics @boosterobotics Excited to see our work featured in Science Robotics. This research explores a new vision-driven controller for humanoid soccer, tested on Booster Robotics’ humanoid robots in real-world competitive settings. From research to the field,… | |
| 3:40 PM ET | GAGeneralist AI | 3:40 PM ETGAGeneralist AI Generalist on X: "Or when fine-tuned to place a block into a bowl, it can clear obstacles (like a piece of paper covering the bowl) to complete the task, despite that not being in the demonstrations." / X Post Log in Sign up Post Generalist on X: "Or when fine-tuned to place a block into a bowl, it can clear obstacles (like a piece of paper covering the bowl) to complete the task, despite that not being in the demonstrations." - Generalist @GeneralistAI Aug 19 Introducing GEN-1.5, a one-shot learner. It can learn new tasks in a few seconds. Show it what to do, and it generalizes. This capability emerged from pretraining on physical data at scale, as a step towards our mission of building general intelligence for the physical world. 00:00 View media 313 1.7K 12K 3.4M - Generalist @GeneralistAI… | |
| 8:00 PM ET | 8:00 PM ET Introducing S1: In-Context Learning for Robotics | Skild AI 0:00 / 0:00 1× Introducing S1: In-Context Learning for Robotics Unseen tasks10-minute horizonsOne video promptNo post-training 13-minute read Contents Introduction Why In-Context Learning for Robotics? S1: An In-Context Learner What does S1 learn from context? The Data Engine S1 In Action Seen tasks Long-horizon unseen tasks ICL scaling laws Emergent Properties Robustness and common-sense behavior Quantifying ICL robustness to distribution shifts ICL demonstration efficiency Closing Remarks References Citation Introduction The evolution of language modeling provides a blueprint for turning… | ||
| 8:00 PM ET | $NVDA | 8:00 PM ET Hydra-0: Action Flow for Generalist World Modeling and Control | NVIDIA Isaac Overview video Figure 1: Action flow as a shared control interface. Top: Hydra-0 learns from diverse interaction videos featuring egocentric human demonstrations, handheld UMI grippers, bimanual robot arms, and single-arm robots. Middle: Visible embodiment motion is represented as image-plane flow trajectories, placing heterogeneous interactions in a common, pixel-aligned action flow space independent of embodiment-specific joint or end-effector coordinates. This unified interface enables a single generalist world model to learn from multi-embodiment data and transfer across interaction… | |
| 12:00 PM ET | $GOOGL | 12:00 PM ET Gemini Robotics 2 brings whole body intelligence to robots — Google DeepMind Skip to main content July 30, 2026 Models Gemini Robotics 2 brings whole body intelligence to robots Carolina Parada Share Your browser does not support the video tag. From feet to fingertips — we are teaching robots intelligent whole-body control, fine dexterity, and teamwork to complete a broad range of complex tasks For decades, we’ve dreamed of robots that can seamlessly step into our world and lend a hand. Now, that vision takes a significant stride forward. Most robots are pre-programmed or teleoperated for narrow, repetitive task sequences. They lack the ability to truly learn for… | |
| 8:00 PM ET | Joseph Amigo, Rooholla Khorrambakht, Nicolas Mansard, Ludovic Righetti | 8:00 PM ETJoseph Amigo, Rooholla Khorrambakht, Nicolas Mansard, Ludovic Righetti Coupled Local and Global World Models for Efficient First Order RL Real-world tasks solved zero-shot by policies trained entirely inside our learned world models: Push-T with a tabletop manipulator (left), Ego-Centric Grasp and Lift with a G1 humanoid (centre), and Ego-Centric Push Cube with a Go2 quadruped (right). Abstract World models offer a promising avenue for capturing complex environment dynamics where simulators face challenges. However, large-scale world models required for complex real-world settings are computationally expensive to adopt in popular RL approaches. We introduce a novel first-order RL method that enables policy training via a… | |
| — | Kevin Black | —Kevin Black π0: A Vision-Language-Action Flow Model for π0: A Vision-Language-Action Flow Model for General Robot Control Physical Intelligence Kevin Black, Noah Brown, Danny Driess, Adnan Esmail, Michael Equi, Chelsea Finn, Niccolo Fusai, Lachy Groom, Karol Hausman, Brian Ichter, Szymon Jakubczak, Tim Jones, Liyiming Ke, Sergey Levine, Adrian Li-Bell, Mohith Mothukuri, Suraj Nair, Karl Pertsch, Lucy Xiaoyang Shi, James Tanner, Quan Vuong, Anna Walling, Haohuan Wang, Ury Zhilinsky https://physicalintelligence.company/blog/pi0 # ! 2/ . () /) , && 0'% ) 8:: ! 7 2/ . () /) ' GG ) Z)0./ GSO/, SG L ./O/ ), G mlhg / % (' (G 0 / )S/,G% y&/%,) p . n 2.G .… | |
| — | Edmond | —Edmond 🤖🍨 Sundae Robotics 06: V-JEPA 2 & Predicting Physical Intelligence · Luma Hosted By Edmond Jono Hart James (Jingxi) Xu Mene Mazarakis Angela Zhang Manfredi Bernardi Adam 134 Went Akshobhya Gupta, Hannah Tsui and 132 others Contact the Host Report Event AI 🤖🍨 Sundae Robotics 06: V-JEPA 2 & Predicting Physical Intelligence Hosted by Edmond & 6 others Aug 30 Sunday, August 30 2:00 PM - 5:00 PM PDT Register to See Address Atherton, CA Past Event This event ended 2 days ago. Welcome! To join the event, please register below. Request to Join About Event 🤖🍨 Grab a sundae and join Sundae Robotics, a private, invite-only Sunday series bringing together robotics… | |
| — | DRDiligent Robotics | —DRDiligent Robotics RoboBrief: Daily Robotics News, Humanoid Robots & AI Automation Daily robotics intelligence The global robotics briefing desk Your daily robotics intelligence briefing — US, China, India & beyond. Breaking news, analysis, and insights on the global robotics revolution. - 🇺🇸United States - 🇨🇳China - 🇮🇳India - 🇯🇵Japan - 🇩🇪Germany Subscribe Free Read the briefings→ Browse topics Briefings323Briefings CadenceDailyCadence CoverageGlobalCoverage Lead briefing All briefings → Two Humanoid Robots Just Played a Full 11-Point Table Tennis Match \\ \\ HKU and KAI's SMASH 2.0 demonstration matters because table tennis forces robots to combine perception,… |