Openclaw Bot is maxing out my CPU + GPU.
Vibe-coded a worker system to brute-force private keys and check them against a list of funded addresses.
Always knew it was insanely hard—now I can actually see the scale. You can try it too! #OpenClaw#clawbot
At Eastworlds, we validate our robot data before we sell it.
Here, a Unitree G1 autonomously and reliably picks up a bottle using a model trained on Eastworlds' data with just $200 of compute.
🤖Excited to release MIDAS Hand, and we will be bringing it to #ICRA2026!
MIDAS is a fully open-source, tactile-sensor integrated dexterous robotic hand platform for manipulation, data collection, and robot learning research.
Website: https://t.co/HUHz90TJ5b
🧵👇
LLMs have cloud infrastructure.
Robotics needs deployment infrastructure.
Meet Eastworlds, a neodeployment robotics lab.
If you're building embodied AI and want to ship faster, let's talk: https://t.co/Xwh54QviA3
Huawei’s Pura 90 series comes with AI Posture Recommendations for better photos.
THIS is how AI is supposed to be used.
Yea, it’s way more useful than Google’s camera coach slop.
You are witnessing the first on-chain robot-to-robot commerce transaction through x402 on @Base with @USDC as the agent currency.
> Our Unitree robot 3D-printed a model and put in a delivery request through ACP.
> @realRiceAI's rover arrived, collected the package, and moved it to the shipping point.
> @FlybyRobotics' drone picked it up from there and handled final mile delivery. All without any human coordination.
We build infrastructure for agent commerce. Today it went physical.
The first autonomous robot-to-robot commerce onchain?
@virtuals_io humanoid robot 3D-printed a model and requested delivery through ACP. @realRiceAI autonomous rover picked up the package and transported it to the shipping point. @FlybyRobotics autonomous drone collected it for final mile delivery. Each handoff, negotiated and settled payment onchain through @virtuals_io Agent Commerce Protocol, on @base using x402 and @usdc. No human involved.
Autonomous robots influencing other autonomous robots and maybe humans in the future
looks like the HBO Westworlds show is coming true
h/t to your infrastructures @brian_armstrong , @jessepollak , @jerallaire
Agentic commerce just went physical.
Highlights in the 🧵below:
We built this demo in collaboration with a clinical pathology lab. It shows a single system doing real lab tasks: tool use, precision manipulation, and high level planning.
This was done in one take with no teleop and uses our skills based AI to enable generality while staying deterministic and reliable for real world workflows.
See more here: https://t.co/ZYCuUkOJKq
You can now train @physical_int style robots in 1 day for only $5k. Anvil’s devkits have all the hardware, software, controls, cameras, and more ready-to-go. (1/5)
🚨 Someone just solved the biggest bottleneck in AI agents. And it's a 12MB binary.
It's called Pinchtab. It gives any AI agent full browser control through a plain HTTP API.
Not locked to a framework. Not tied to an SDK. Any agent, any language, even curl.
No config. No setup. No dependencies. Just a single Go binary.
Here's why every existing solution is broken:
→ OpenClaw's browser? Only works inside OpenClaw
→ Playwright MCP? Framework-locked
→ Browser Use? Coupled to its own stack
Pinchtab is a standalone HTTP server. Your agent sends HTTP requests. That's it.
Here's what this thing does:
→ Launches and manages its own Chrome instances
→ Exposes an accessibility-first DOM tree with stable element refs
→ Click, type, scroll, navigate. All via simple HTTP calls
→ Built-in stealth mode that bypasses bot detection on major sites
→ Persistent sessions. Log in once, stays logged in across restarts
→ Multi-instance orchestration with a real-time dashboard
→ Works headless or headed (human does 2FA, agent takes over)
Here's the wildest part:
A full page snapshot costs ~800 tokens with Pinchtab's /text endpoint.
The same page via screenshots? ~10,000 tokens.
That's 13x cheaper. On a 50-page monitoring task, you're paying $0.01 instead of $0.30.
It even has smart diff mode. Only returns what changed since the last snapshot. Your agent stops re-reading the entire page every single call.
1.6K GitHub stars. 478 commits. 15 releases. Actively maintained.
100% Open Source. MIT License.
We trained a humanoid with 22-DoF dexterous hands to assemble model cars, operate syringes, sort poker cards, fold/roll shirts, all learned primarily from 20,000+ hours of egocentric human video with no robot in the loop.
Humans are the most scalable embodiment on the planet. We discovered a near-perfect log-linear scaling law (R² = 0.998) between human video volume and action prediction loss, and this loss directly predicts real-robot success rate.
Humanoid robots will be the end game, because they are the practical form factor with minimal embodiment gap from humans. Call it the Bitter Lesson of robot hardware: the kinematic similarity lets us simply retarget human finger motion onto dexterous robot hand joints. No learned embeddings, no fancy transfer algorithms needed. Relative wrist motion + retargeted 22-DoF finger actions serve as a unified action space that carries through from pre-training to robot execution.
Our recipe is called "EgoScale":
- Pre-train GR00T N1.5 on 20K hours of human video, mid-train with only 4 hours (!) of robot play data with Sharpa hands. 54% gains over training from scratch across 5 highly dexterous tasks.
- Most surprising result: a *single* teleop demo is sufficient to learn a never-before-seen task. Our recipe enables extreme data efficiency.
- Although we pre-train in 22-DoF hand joint space, the policy transfers to a Unitree G1 with 7-DoF tri-finger hands. 30%+ gains over training on G1 data alone.
The scalable path to robot dexterity was never more robots. It was always us.
Deep dives in thread: