Image created with Flux Pro v1.1 Ultra. Image prompt: Giant “100” as pure white negative‑space cutout dominating the frame; minimalist poster style; headset reflection showing a minimal 3D grid receding through the cutout; neon‑teal on charcoal backdrop; high contrast, crisp edges, soft studio light, no other text, no logos
Chan Zuckerberg Initiative’s rBio uses virtual cells to train AI, bypassing lab work https://venturebeat.com/ai/chan-zuckerberg-initiatives-rbio-uses-virtual-cells-to-train-ai-bypassing-lab-work
Google DeepMind researchers say world models allow robotic agents to learn and plan in simulated environments, cutting real-world risks and costs. Robots can also potentially experience and learn rare scenarios in a world model, like Halloween crowds. https://x.com/TheHumanoidHub/status/1958614223178932474
Your robot moves fast… but objects slide off the tray? This system hears the sliding and learns how to stop it: Researchers at CMU developed a new method that uses sound to model real-world friction in motion planning. It enables time-optimized, high-speed transport without https://x.com/IlirAliu_/status/1960244593007423664
Field AI raised $405M in funding and introduced Field Foundation Models These models are designed to grapple with uncertainties and the physical constraints of the real world, enabling safe robot behaviors when navigating in new environments https://x.com/adcock_brett/status/1959647835650953310
This is so dope! Fun play on simulation theory”” / X https://x.com/bilawalsidhu/status/1959441928233340932
FOCI: Trajectory Optimization on Gaussian Splats”” TL;DR: 1) collision measure between GS based on the overlap integral between Gaussians. 2) trajectory planning algorithm –> fully differentiable trajectories. 3) GPU implementation of the trajectory optimization. https://x.com/Almorgand/status/1960732725368766801
GenAI is accelerating spatial intelligence. Here are some results from a Large World Model developed by World Labs. The model generates permanently consistent 3D worlds from a single image that can be smoothly navigated in a web browser. https://x.com/TheHumanoidHub/status/1960385533923749924
Generative AI is really cool but sometimes you want to drift your car through an intersection in a super specific way. Draw a spline then get 3D onion skinning so you can adjust the curves with a clear spatial reference. iCars plugin for Blender: https://x.com/bilawalsidhu/status/1958982936172134410
MindJourney enables AI to explore simulated 3D worlds to improve spatial interpretation – Microsoft Research https://www.microsoft.com/en-us/research/blog/mindjourney-enables-ai-to-explore-simulated-3d-worlds-to-improve-spatial-interpretation/
TINKER: Diffusion’s Gift to 3D–Multi-View Consistent Editing From Sparse Inputs without Per-Scene Optimization”” TLDR: TINKER achieves generalizable 3D editing with one or few inputs without per-scene optimization. https://x.com/Almorgand/status/1958884433643065510
Using 3d gaussian splats to stabilize shaky handheld footage — giving you the ultimate hyper lapse. New ai research (LongSplat) https://x.com/bilawalsidhu/status/1960088618174648341
World labs image-to-3D scene tech. Consistent AI environments are about to get a lot easier. Been playing w/ the private beta and it’s getting really good. Since it’s 3DGS, you can connect multiple scenes together and bring it to life in any 3D tool. https://x.com/bilawalsidhu/status/1958908913631600718
Harvard dropouts to launch ‘always on’ AI smart glasses that listen and record every conversation | TechCrunch https://techcrunch.com/2025/08/20/harvard-dropouts-to-launch-always-on-ai-smart-glasses-that-listen-and-record-every-conversation/
Today, we’re launching the Runway Game Worlds Beta. Over the last few months, we have been working on research and products that are moving us closer toward a future where you will be able to explore any character, story or world in real time. While generating the pixels of https://x.com/runwayml/status/1958516860149997672
Simulations are the future, & one of the main tools we’ll ultimately use to understand and predict things about the universe. This is why I’m so excited about Genie 3, our latest interactive world simulator – here are some insanely cool things you might have missed about it 🧵: https://x.com/demishassabis/status/1958696882105995312
AI world models & why they matter… in 20 minutes. The holodeck is closer than you think, and the implications for robotics, AR/VR and content creation are staggering. In this video, I’ll breakdown everything from Google Genie 3 and NVIDIA Cosmos to startups powering https://x.com/bilawalsidhu/status/1958320179714150824
AI is getting REALLY meta. Genie 3 builds a simulation of reality by digesting YouTube. SIMA agent learns inside of it. Repeat. Your robots will be dreaming at night, reliving their mistakes, and figuring out how to do a better job next time. This world model tech really https://x.com/bilawalsidhu/status/1959295692515541143
Boston Dynamics and Toyota Research Institute are advancing humanoid robotics with Large Behavior Models (LBMs) for Atlas https://x.com/sushilpokhrel/status/1958299471621607673
NVIDIA’s new humanoid robot brain is officially on sale for $3,499. NVIDIA has announced the general availability of the Jetson AGX Thor developer kit and production modules. Powered by an NVIDIA Blackwell GPU and featuring 128GB of memory, it delivers up to 2,070 FP4 teraflops https://x.com/TheHumanoidHub/status/1960008007540199871
Robots don’t wait. So why should your model? Large VLAs run in real time… with no training-time changes. ❗️Worth reading if you’re working on real-world deployment of large models in robotics. I found this write-up on Real-Time Action Chunking (RTC) from Physical Intelligence, https://x.com/IlirAliu_/status/1960760040773611752
The Large Behavior Model (LBM), co-developed by Boston Dynamics and TRI, can perform various manipulation tasks on Atlas MTS (the Atlas upper-body robot). The training process is the same whether it’s manipulating a 22 lb car tire, tying a rope, or flipping a barstool. https://x.com/TheHumanoidHub/status/1958356134076768312
The production process… of a 𝗟𝗘𝗚𝗢 𝗳𝗶𝗴𝘂𝗿𝗲: Lego assembly is more than just toys—it’s a feat of precision engineering. What goes on behind the scenes? ✅ The machines are really smart. They use things like vibrating bowls to sort pieces and robot arms to put them https://x.com/IlirAliu_/status/1959881016434139194
What if robots could learn new skills in 20 minutes, without humans? [📍 Save code & videos] How do we teach robots new motor skills directly in the real world? A framework called Robot-Trains-Robot (RTR) explores a teacher/student setup: a robot arm trains a humanoid by https://x.com/IlirAliu_/status/1959158235786510474
World’s first demonstration of a robot able to tie shoelaces & hang t-shirts… 𝗮𝘂𝘁𝗼𝗻𝗼𝗺𝗼𝘂𝘀𝗹𝘆𝗅 ALOHA Unleashed 🌋: A Simple Recipe for Robot Dexterity They trained a diffusion policy at scale: 26,000 demonstrations over 5 tasks on Aloha 2 robot! Recent research https://x.com/IlirAliu_/status/1960399401618141551
What if robots could learn new tasks as easily as ChatGPT picks up a new prompt? [📍 Save the project for later] Robots don’t yet have the in-context learning powers of LLMs like ChatGPT or Gemini. Most Vision-Language-Action (VLA) models need retraining or fine-tuning to https://x.com/IlirAliu_/status/1958795845681226024
Elon could start by making X the best matchmaker on the planet. Put Grok to work to build connections in the real world. Not pull us away from it into the arms of an AI companion.”” / X https://x.com/bilawalsidhu/status/1958570141119037766
Runway has a new game engine thing that creates interactive comics with choices in any style you want. This is what it did when I gave it the prompt “”explore noses.”” Similar, but more graphical than previous AI Dungeon-like work, with some decent systems to keep things on rails. https://x.com/emollick/status/1959268961964417027
