Dirgha AI

Open-source Bharat. We're building a sovereign AI OS, a decentralized work/compute marketplace, CLI/IDE, for your computers, robots and rockets. DM ⌨ @salik

[Dirgha AI @DirghaAI]

Lord Rama crossed the ocean by building a bridge stone by stone, with an army of contributors, each doing what they were capable of.

Dirgha is following the same model: we're bringing together people and organizations to build fully indigenous, agentic, sovereign AI computers.

[a16z @a16z]

Mira Murati says human-AI collaboration needs models that can listen while they think:

"The types of models that we work with today, they're very turn-based. You talk, they talk, then they go off and think."

"While they're thinking, it's almost like they're deaf and blind. They cannot perceive anything else about what's going on."

"By contrast, our interactions with each other are very rich. There is a lot of information in our interactions when we are silent, when we're thinking, when we're interrupting one another."

"Interaction models are able to capture all of this nuance. They're not turn-based. They're more like time-based interaction, where they're continuously taking in audio, text, video, and continuously providing output."

"This enables you to catch things like interruptions and simultaneous speech, and really create a rich, high bandwidth interaction between humans and machines."

@miramurati at Bloomberg Tech live with @emilychangtv

[C. Zhang @ChongZzZhang]

Holy crap

https://t.co/Uq2jGHnxkS

[Lukas Ziegler @lukas_m_ziegler]

The robot is used to inspect the walls of ships. 🚢

The Magnet Crawler introduced by the University of Bremen is a lightweight and compact climbing robot for magnetic surfaces.

The robot has two magnetic wheels driven by geared motors and an elastic tail to improve stability. In the front part of the robot, there is a miniature wireless camera.

The robot is remotely teleoperated & the video signal from the camera is displayed to the operator on a portable screen.

Let's wait and perhaps the robot will become more autonomous over time.

Big fan of robots that make the world a safer place for humans.

~~

♻️ Join the weekly robotics newsletter, and never miss any news → https://t.co/GoA3ZuwoPB

[Sakana AI @SakanaAILabs]

Building AI that Builds AI: Introducing the Sakana AI RSI Lab 🚀

https://t.co/AskX3J5oEJ

Today, we are announcing the Sakana AI Recursive Self-Improvement (RSI) Lab: a dedicated research group in Tokyo tasked with redesigning the AI development process itself using AI.

While the industry increasingly speculates about the theoretical potential of self-improving AI, we’ve spent the last two years actively laying the foundations to make it a reality:

▪ LLM²: AI models automating research to invent better preference optimization algorithms.
▪ Darwin Gödel Machine: Agents autonomously rewriting their own codebase to double software-engineering performance.
▪ ShinkaEvolve: Hyper-sample-efficient program evolution that builds novel loss functions for MoE models.
▪ ALE-Agent: Reinforcement agents outperforming hundreds of human experts via self-learning.
▪ Digital Red Queen: Open-ended adversarial coevolution laying the groundwork for RSI in cybersecurity.
▪ The AI Scientist: Towards end-to-end automation of AI research, recently published in Nature.

Now, we are unifying these breakthroughs. The Sakana AI RSI Lab is officially tasked with building open-ended, adaptive architectures that collectively self-improve.

Human intelligence did not emerge from limitless resources; it was forged through the open-ended, compounding process of evolution operating under strict constraints. We are applying this exact principle to AI.

We believe recursive self-improvement is achievable on modest, sample-efficient compute. It shouldn’t be a winner-take-all asset locked inside hyperscale clusters, but a democratized public good.

We’re scaling our team to execute this mission. We are looking for frontier scientists and engineers who are entirely unsatisfied with the brute-force status quo. If you are ready to break away from standard benchmarking and build the self-improving future in Japan, come build with us.

[Unsloth AI @UnslothAI]

Google releases Gemma 4 QAT. ✨
You can now run Gemma 4 at 3x less memory with near original performance.

Quantization-Aware Training (QAT) makes it possible to run Gemma 4 26B-A4B on 16GB RAM.

GGUFs: https://t.co/wQgEocxUId
QAT Guide: https://t.co/Nsm1yeGEHx

[Pamphlets @PamphletsY]

🚨🇨🇳 BREAKING — MADE IN CHINA:

China’s High-Tech 3D Printers Capture Nearly 90% Of Market.

[Kai Han @kaihan_x]

🚀 Speed3R (#CVPR2026): sparse feed-forward 3D reconstruction. 12.4× faster on 1000-view sequences with minimal accuracy loss. Plug-and-play on VGGT, π³, etc.
🔗 https://t.co/FDJt5F8kID
📄 https://t.co/8aBJkz5D2Y
💻 https://t.co/61BKweaWf0

[Jim Fan @DrJimFan]

NitroGen just won CVPR Best Paper Honorable Mention!! We are making strides towards general-purpose embodied agents that master not only the real world physics, but also all possible physics across a multiverse of simulations.

It’s been 4 years since MineDojo, our first embodied agent in Minecraft, won NeurIPS Best Paper. Congrats to everyone on the team!!

[Space and Technology @spaceandtech_]

Figure says it has increased humanoid robot production from one robot per day to one robot per hour in just 120 days.

The robots undergo intensive testing, including repeated squats and jogging movements.

Figure also demonstrated the robot's ability to climb stairs and navigate complex terrain using vision-based control.

[Li Zexin 李泽欣 @XH_Lee23]

🚀🏖️China launched a rocket sending a new satellite group to constitute the Chinese Starlink from Hainan Island this afternoon.

This coastal rocket launch site is, in my opinion, the most beautiful one in China. Watching launches here has become a very popular tourist attraction.

[asim ᯅ @asimahmed]

for physical AI to work in the real world, it needs high-fidelity 3D models of the real world.

not just for visualization — for localization, spatial reasoning, and real-to-sim workflows that close the gap between training and deployment.

at @nianticspatial, we're turning @spexigeospatial drone imagery into exactly that — city-scale Gaussian splats grounded in the environments that matter most.

[Yoonseok Pyo @passionvirus]

Build your humanoid in 7 days! 🦾

By combining NVIDIA Isaac Sim with @ROBOTIS DYNAMIXEL-Q (QDD) actuators, you get high backdrivability and precise torque control that seamlessly bridges the Sim2Real gap.

See how we fast-tracked physical deployment:
https://t.co/mnY0ipBQCd

[altan tutar @altantutar]

10 startups collecting the data to train robots (all launched or funded in the last year):

1/ Config (Korea): 100K+ hrs of human-motion data, raised $35M led by Samsung Ventures
2/ Shift App (Global): Spun out of MicroAGI, promises to clean the houses for free if data collection is allowed.
3/ Human Archive (US/India): workers wear camera caps on the job, raised $8.2M from Wing, NVP, YC
4/ Build AI: dropped the Egocentric-100K dataset (an open dataset of 100K+ hrs of factory-worker POV), raised $5M from Abstract and Pear
5/ Asimov (YC W26): pays 5,000+ households to record their daily routines on hardware Asimov builds itself.
6/ Mecka AI (NYC): captures human movement with body sensors and iphones, raised $60M led by Framework Ventures
7/ Tutor Intelligence (MA, USA): runs a 100-robot "Data Factory" generating real-world manipulation data, raised $34M led by Union Square Ventures
8/ Cortex AI (SF, YC F25): first-person workplace video plus robot trajectories, raised $6M
9/ Mimic (Zurich): workers wear capture rigs on the factory floor so robots learn by imitation, raised $16M as an ETH spinout.
10/ Lightwheel (EgoSuite): 300K+ hrs of first-person human data from homes, factories, and warehouses

Did I miss any? Add it below.

[Google Research @GoogleResearch]

Introducing D4RT: A unified AI model for 4D scene reconstruction and tracking across space and time. 🎯 Catch the demo with Skanda Koppula at 12 pm at our #CVPR2026 Google booth kiosk! https://t.co/p6SclNe1zi @GoogleDeepMind

[Li Zexin 李泽欣 @XH_Lee23]

China is building a huge solar farm in the desert. Robots are deployed to install solar panels.

China’s green energy commitment is no kidding.

[Hardeep Singh Puri @HardeepSPuri]

Congratulations @OilIndiaLimited !

An ocean of energy opportunities reinforced in the Andaman Sea!
Very happy to report the presence of natural gas in Sri Vijayapuram-3 an exploratory well drilled by Oil India Ltd. 15 km off the east coast of the Andaman Islands at a water depth of 355 meters.
Initial production testing of the well at the depth of 1900 plus meters in the Eocene formation has established the presence of natural gas through continuous flaring.
Oil India is carrying out gas sampling to assess the composition & calorific value of gas and to carry out isotope studies to understand the genesis of the gas.
Under the Samudra Manthan Mission (National Deep Water Exploration Mission) announced by Hon’ble PM @narendramodi Ji on Independence Day 2025, large number of deepwater & Ultra deepwater exploration wells are planned in our offshore basins to fully exploit our hydrocarbon reserves.
Presence of hydrocarbon is now reported in 2(Two) wells out of 3 (Three) exploratory wells drilled by OIL in current exploratory campaign off Andaman Basin.
This presence of natural gas will help us in taking forward our exploration ambitions in coordination with global deepwater exploration experts like @petrobras, @TotalEnergies, @bp_india, @Shell, @exxonmobil and will be a significant milestone in our journey through Amrit Kaal!
@PMOIndia @PIB_India @PetroleumMin @mygovindia
#OilIndia #EnergySecurity #Exploration #NaturalGas #Andaman

[Mario Nawfal @MarioNawfal]

🇨🇳 China is all-in on AI

They're now using it in 100 autonomous trucks in one of their largest coal mines in Inner Mongolia

[Ye Yuan @_ye_yuan]

Humanoid robotics is hitting a data wall. Teleop and mocap took us far, but they don’t scale to every object, terrain, and behavior.

We’re releasing GRAIL: https://t.co/LxTKtMPtw0 — a fully digital pipeline for generating loco-manipulation data before the robot moves. 🧵(1/8)

[Eren Chen @ErenChenAI]

Mocap actress @xixiyusakana Real Time Teleoperating 6 robots in an AMD sponsored event.

[Unsloth AI @UnslothAI]

Gemma 4 12B can now run locally on just 8GB RAM via Dynamic GGUFs.

Google's new model, Gemma 4 12B Unified supports image, audio and 256K context.

You can run and train the model via Unsloth Studio.

GGUF: https://t.co/8cL321pVDh
Guide: https://t.co/odRo9WjRpA