Image created with gemini-2.5-flash-image with claude-sonnet-4-5. Image prompt: Cinematic wide shot of an orbital space station training arena with autonomous robots of various forms suspended in zero gravity conducting synchronized tactical maneuvers, Earth visible far below through transparent sections, cool blue and green lighting with dramatic rim light, sleek military sci-fi aesthetic, deep blacks and muted neon data streams connecting the machines, Ender’s Game inspired strategic systems atmosphere
One of the biggest use cases for agentic document automation is insurance underwriting ✍️ Underwriting depends on processing *massive* volumes of unstructured documents, from medical reports, scanned forms, and way more. It’s also historically been a massively manual process. https://x.com/jerryjliu0/status/1988394058197184923
Build a document understanding agent for SEC filings that uses a multi-step approach with LlamaClassify and Extract to identify the filing type and hand it off to the right extraction agent. Deployed with LlamaAgents. 🔧 Customize extraction schemas to fit your specific data https://x.com/llama_index/status/1988696219015848401
Super excited to announce SIMA 2! It’s a general agent that can understand & reason about complex instructions and complete tasks in simulated game worlds, even ones it has never seen before. Incredible to see how it can learn just from self-play… a crucial step towards AGI https://x.com/demishassabis/status/1989096784870928721
SIMA 2: A Gemini-Powered AI Agent for 3D Virtual Worlds – Google DeepMind https://deepmind.google/blog/sima-2-an-agent-that-plays-reasons-and-learns-with-you-in-virtual-3d-worlds/
SIMA 2 is our most capable AI agent for virtual 3D worlds. 👾🌐 Powered by Gemini, it goes beyond following basic instructions to think, understand, and take actions in interactive environments – meaning you can talk to it through text, voice, or even images. Here’s how 🧵 https://x.com/GoogleDeepMind/status/1988986218722291877
Our SIMA 2 research offers a strong path towards applications in robotics and another step towards AGI in the real world. Find out more → https://x.com/GoogleDeepMind/status/1988987865401798898
SIMA 2 🤝 Genie 3 We tested SIMA 2’s abilities in simulated 3D worlds created by our world model Genie 3. It demonstrated unprecedented adaptability by navigating its surroundings and took meaningful steps toward goals. https://x.com/GoogleDeepMind/status/1989024090414309622
SIMA 2: A Gemini-Powered AI Agent for 3D Virtual Worlds – Google DeepMind
https://deepmind.google/blog/sima-2-an-agent-that-plays-reasons-and-learns-with-you-in-virtual-3d-worlds/?utm_source=x&utm_medium=social&utm_campaign=&utm_content=
Project Fetch: Can Claude train a robot dog? \ Anthropic https://www.anthropic.com/research/project-fetch-robot-dog
New Anthropic research: Project Fetch. We asked two teams of Anthropic researchers to program a robot dog. Neither team had any robotics expertise–but we let only one team use Claude. How did they do? https://x.com/AnthropicAI/status/1988706380480385470
Morgan Stanley projects Apple could make $133 billion a year on humanoid robots by 2040. https://x.com/TheHumanoidHub/status/1987053990694297952
Marble: A Multimodal World Model | World Labs https://www.worldlabs.ai/blog/marble-world-model
A robot could learn a task just by watching a generated video? PhysWorld connects video generation with real-world robot learning. It turns visual imagination into physical skill. ✅ Takes one image and a task prompt ✅ Generates a video showing how to complete the task ✅ https://x.com/IlirAliu_/status/1988678189527273831
[2511.07416] Robot Learning from a Physical World Model https://arxiv.org/abs/2511.07416
Fei-Fei Li’s World Labs speeds up the world model race with Marble, its first commercial product | TechCrunch https://techcrunch.com/2025/11/12/fei-fei-lis-world-labs-speeds-up-the-world-model-race-with-marble-its-first-commercial-product/
From Words to Worlds: Spatial Intelligence is AI’s Next Frontier https://drfeifei.substack.com/p/from-words-to-worlds-spatial-intelligence
.@drfeifei started her new blog We believe this will be one of the most interesting reads about Spatial Intelligence. She writes, that Spatial Intelligence depends on world models built on 3 core principles: – They must be generative – able to create coherent, https://x.com/TheTuringPost/status/1988727531353305524
Perceptron’s platform is here — built for Physical AI Developers can now use Isaac-0.1 or Qwen3VL 235B via: Perceptron API — fast, reliable multimodal intelligence Python SDK — simple, grounded prompting for vision + language Build apps that see and understand the world. https://x.com/perceptroninc/status/1988713482460750290
One of my proudest moments was sending an autonomous humanoid robot into the real world To this day, I believe Figure is the first and maybe the only company to achieve this”” / X https://x.com/adcock_brett/status/1986861619423682675
We’ve been integrating Isaac across the industry and have realized developers are missing a single platform for Physical AI – prompt engineering, deployment, and integration. Today we are excited to release Perceptron’s Platform – supporting our API – supporting chat”” / X https://x.com/AkshatS07/status/1988713765152649711
Perceptron AI on X: “Perceptron’s platform is here — built for Physical AI Developers can now use Isaac-0.1 or Qwen3VL 235B via: Perceptron API — fast, reliable multimodal intelligence Python SDK — simple, grounded prompting for vision + language Build apps that see and understand the world. https://t.co/5ZyaOGQb1i” / X
https://x.com/perceptroninc/status/1988713482460750290
XPENG’s next-gen IRON robot effectively crossed the uncanny valley, leading many to believe it was a human in a suit. In a follow-up event to prove it was a robot, He Xiaopeng had its leg skin cut open in front of a live audience. The robot then walked off the stage. https://x.com/TheHumanoidHub/status/1986482482460725755
Most diffusion policies can imitate behavior… but they can’t guarantee safe behavior. That’s a problem when robots share space with people. Researchers from TUM and Stanford propose Path-Consistent Safety Filtering (PACS), a method that gives formal safety guarantees to https://x.com/IlirAliu_/status/1988550922444693898
Robots are great at following instructions. But what happens when those instructions fail? Most Vision-Language-Action models freeze or repeat the same mistake. A new approach called FailSafe shows how robots can detect and fix their own failures. The method uses a companion https://x.com/IlirAliu_/status/1986353266322538634
My conversation with Nikhil Kamath is now live We talked about why humanoid robots and flying taxis are possible today, and the journey of building Archer Aviation and Figure Here’s the full video: https://x.com/adcock_brett/status/1986477765055615118
APPROVED: Elon’s 2025 CEO Performance Award To earn 12% of outstanding $TSLA shares, he requires to grow Tesla to an $8.5T market cap by hitting 12 tranches of market cap and operational goals (1M bots, 20M cars, 1M robotaxis) https://x.com/TheHumanoidHub/status/1986555258886402190
Elon Musk says Tesla will build a 1-million-unit Optimus production line in Fremont and then a 10-million-unit production line in Austin. https://x.com/TheHumanoidHub/status/1986563829925617806
OpenReal2Sim Demo https://hesic73.github.io/OpenReal2Sim_demo/
On the road to AGI: deadmau5 + robots tonight https://x.com/adcock_brett/status/1986912861986398519
Humanoid robots at deadmau5 🔥🔥🔥 https://x.com/adcock_brett/status/1987333320334844305
Brett Adcock on Agility Robotics: “”Because of poor engineering choices and almost no progress over the past 10 years, they’re unlikely to survive as things stand.”” https://x.com/TheHumanoidHub/status/1986917249245225104
Heading to the first humanoid robot concert, let’s gooooo https://x.com/adcock_brett/status/1987323462504133060
Never leaving this app Agility Robotics took a shot at Brett. Brett brings the clapback catapult. https://x.com/TheHumanoidHub/status/1986907347487445292
Figure and Optimus are making astounding progress.”” / X https://x.com/chamath/status/1986814442114601175
We’re back with another update to Veo 3.1: Rolling out now on mobile and desktop, you can upload multiple reference images alongside your video prompts, to create entirely new worlds and more nuanced videos that are true to your vision.”” / X https://x.com/GeminiApp/status/1989440642179801192
It doesn’t look like much. But this small robot from Stanford might do for robotics what the Apple II did for computing… make it personal. 🧵 [📍 Save this… you’ll want to remember his name.] https://x.com/IlirAliu_/status/1987531734468927691
Ben Bolte talks about why he failed to raise funding for K-Scale, which led to the winding down of the startup. – Bolte says his number-one misstep was “”overindexing on how easy I thought it would be to get funding.”” He was having trouble raising money, and a VC he admired https://x.com/TheHumanoidHub/status/1986854109614645572
Vision-based teleoperation 👀🤖 Open-source robotic hand. Try it yourself… [📍GitHub] Aero Hand Open is a low-cost, open-source, lightweight (< 400 grams), anthropomorphic robotic hand designed and developed by @TetherIA_ai for affordable dexterous manipulation research. It https://x.com/IlirAliu_/status/1986712187306377395
UBTech claims to have completed the world’s first mass delivery of humanoid robots. The Shenzhen-based company has secured over 800 million yuan (US$112M) in Walker S2 orders this year, including a 159M yuan deal for Zigong’s data collection project and another 250M yuan https://x.com/TheHumanoidHub/status/1988666839384445132
The IRON robot, in its full bare metal glory! XPeng’s CEO shares an inside look at the humanoid in the lab. https://x.com/TheHumanoidHub/status/1987049101192765782
today, we’re open sourcing the largest egocentric dataset in history. – 10,000 hours – 2,153 factory workers – 1,080,000,000 frames the era of data scaling in robotics is here. (thread) https://x.com/eddybuild/status/1987951619804414416
It’s easy to miss the big picture: the Unitree launched G1 with only basic locomotion just 18 months ago. The whole field is accelerating, and Unitree has been a key catalyst by getting robots into the hands of thousands of researchers. New video: Unitree’s whole-body teleop https://x.com/TheHumanoidHub/status/1986495352997093433
OpenArm: 7DOF bimanual humanoid arms for physical AI development in contact-rich tasks. Human-scale (63 cm reach), 6kg peak payload, 1kHz CAN-FD control, all HW/SW open-source. https://x.com/TheHumanoidHub/status/1986518488022982706
🎙️ @ax_pey, co-founder & CEO of @innate_bot, shows you how to teach real robots with language and quick demos without being a roboticist. We talk about Axel’s path from France to Stanford and why he is betting on personal robotics you can program with prompts, code, and https://x.com/IlirAliu_/status/1986435149119234515
Most robots fail at the last millimeter. That’s where precision, not power, decides everything. This system aims to close that gap. It executes tolerance-critical motions with steady, millimeter-level control, without manual tuning or retraining. ✅ Adapts instantly to new https://x.com/IlirAliu_/status/1988167649784705392
It’s what’s inside that counts. Next-gen XPENG IRON https://x.com/TheHumanoidHub/status/1986508587292893528
We will transcend space & time by teleoperation human robots anywhere in the world using VR headsets. It’s wild how far we’ve come from iPads on wheels. https://x.com/bilawalsidhu/status/1986336092111708459
Tesla vehicles and Optimus share core technology: Actuators Power electronics Battery Manufacturing Data communication Audio system Cameras A14/A15 chips Training cluster Neural simulation Real-world AI https://x.com/TheHumanoidHub/status/1986565979841962377
America and China build robots. Europe builds committees. Tesla just showed Optimus in pilot production. Humanoids being assembled like cars. Meanwhile, in Europe, we’re still arguing over regulation, ethics boards, and frameworks that nobody in the real world reads. This https://x.com/IlirAliu_/status/1986869259226456142
Most delivery robots can roll. Very few can fly. Mercury, built by @Mercurius_Tech in San Francisco, is a multi-modal delivery robot-drone that can both drive and take off… carrying up to 1 kg of payload. The team built the first prototype during the Blueprint program at https://x.com/IlirAliu_/status/1987806479491965173
“Maybe specifically in this robotics climate, robots shouldn’t have boobs, like that was kind of like a criticism, I just want to give you an opportunity to talk about that specifically.” https://x.com/TheHumanoidHub/status/1987181250935988635
Tesla: “”Optimus pilot production line is currently running in our Fremont Factory. Significantly larger Gen 3 production line coming in 2026. Our goal is $20k COGS per robot at scale”” https://x.com/TheHumanoidHub/status/1986564958692515900
Optimus hand on display at Tesla Annual Shareholder Meeting https://x.com/TheHumanoidHub/status/1986546190151000269
The beginning of a new era for Tesla. The Annual Shareholder Meeting scheduled to start in 5 minutes. https://x.com/TheHumanoidHub/status/1986537866546323553
The work for the 10-million-unit Optimus facility is already underway. Elon at a shareholder meeting last week said: “”We’re gonna launch the fastest production ramp of any product, of any large complex manufactured product ever. And starting with building a million-unit https://x.com/TheHumanoidHub/status/1988650633843617942
Leave a Reply