OpenClaw 2026.3.22 🦞 🏪 ClawHub plugin marketplace 🤖 MiniMax M2.7, GPT-5.4-mini/nano + per-agent reasoning 💬 /btw side questions 🏖️ OpenShell + SSH sandboxes 🌐 Exa, Tavily, Firecrawl search This release is so big it needs its own table of contents. https://t.co/XvRbXEduGC

📢WorldAgents: 3D worlds only from 2D image models - without any training! We propose an agentic approach with a Director (VLM) to plan the scene, a Generator (Flux or NanoBanana) for new views, and a Verifier (VLM) for selection / 3D consistency. -> High-fidelity 3D worlds from a single text prompt. What's remarkable: our agents find consistent views from 2D image models to obtain 3D-consistent worlds; this shows that image models contain world priors - agents just need to find them! https://t.co/6NC7zIEn4n https://t.co/vTO3sLFLFw Great work by @ErkocZiya @angelaqdai

agriculture changes forever starting today. introducing FAMA, our gen 1 autonomous robot that plants seeds, fertilizes, and weeds farms. fully self-driving, no GPS. we had no hardware background when we started. but we pulled through. more robots for farming incoming 🚀 https://t.co/HKyDFAayCx

JFK rn https://t.co/kaHMMRIaCD

Today we're launching Latent-Y: the world's first autonomous agent for drug design, lab-validated end to end. Give it a research goal. Latent-Y reasons, designs, iterates, and delivers lab-ready antibodies, autonomously or collaboratively, with the biological reasoning of a PhD protein design expert. Technical report: https://t.co/E7IHfkvvD3 Blog post: https://t.co/GfJAfzj0Qx Apply for access: https://t.co/E0SR9znZiP

If you're building anything in AI, the best skill you need to be using right now is hugging-face-paper-pages Whatever problem you're facing, someone has probably already published a paper about it. HF's Papers API gives a hybrid semantic search over AI papers. I wrote an internal skill, context-research, that orchestrates the HF Papers API into a research pipeline. It runs five parallel searches with keyword variants, triages by relevance and recency, fetches full paper content as markdown, then reads the actual methodology and results sections. The skill also chains into a deep research API that crawls the broader web to complement the academic findings. The gap between "a paper was published" and "a practitioner applies the insight" is shrinking, and I think this is a practical way to provide relevant context to coding agents. So you should write a skill on top of the HF Paper skill that teaches the model how to think about research, not just what to search for.

Underneath the Composer controversy, it's pretty wild to see the progress on Terminal-Bench 2.0. Open models are now stacking above 50% — 8 months ago, the best was under 28%. Made a @huggingface Space to see how models have progressed on important benchmarks. https://t.co/nELCt0eY4Z

This 2003 lecture from Elon Musk at Stanford University is an absolute gem The vision he had for what the internet and SpaceX would become all eventually came true Just goes to show how far ahead of his time he really was https://t.co/yfvqnpTNpl

"The Earth only receives about half a billionth of the Sun’s energy. The Sun is 99.8% of all mass in the solar system… Earth is in the miscellaneous category. Earth is like a tiny dust mote in a vast darkness. The way to actually scale civilization is to scale power in space." @elonmusk

Some new images released tonight during Elon's TERAFAB presentation. https://t.co/q1MRuYhKjB

+2 more

SPACEX/xAI/TESLA TERAFAB Last night was historic. It was BIG. The Terafab announcement is Civilization Level BIG. Make humanity Galactic BIG. Robots that can replicate themselves in Space BIG. ♦️Not just because the physical Terafab is an enormous integrated building that will house chip design, memory, packaging — and quite possibly the machines that build the machines — under one roof. It’s BIG because the Terafab is the final missing piece of the puzzle. ♦️SpaceX heralded it as “the next step towards becoming a galactic civilization.” This creates full vertical integration and the recursive iterative process. LEVELED UP It is BIG because once again Elon Musk has leveled up. (The goal to become Multi-planetary, or even interstellar, was SO 2025). Now we need to recast the highest category of GDP from Global to Galactic production output. As Elon Musk says in an economy a million times the size of Earth’s current output: “if you can think of it, you can have it.” ♦️Most of Wall Street thinks linearly, not exponentially. They don’t get. Elon Musk agrees: “They really don’t, and I’ve given them all the information.” But they will. Don’t be like Wall Street. Don’t miss out. Next: Why Wall Street gets it wrong…. https://t.co/NWqkuYBZY9

+2 more

Cute and epic Chibi Elon and Maye SNL moment. The best Mom & Son! 🥰 https://t.co/1fQFZAC6s8

Grok Imagine just completely took over the DesignArena Video leaderboard with three #1 rankings It's officially #1 in Video, Image-to-Video (beating Sora and Veo 3.1), and Video Editing xAI had zero presence in the video space a few months ago, and now they are completely dominating the benchmarks The speed of progress here is just wild

The idea that Pokémon GO players secretly trained AI is not quite accurate. While concerns surfaced about images being used without consent, the reality is more nuanced. Data collection was tied to user interactions within the app, not entirely hidden or unknown. It is a reminder that in AI, the real issue is often not whether data is used, but how clearly people understand and agree to it. https://t.co/vOkbeK8iHC @euronews

🚀 The Inference & Production track at #PyTorchCon Europe (7-8 April in Paris) covers deploying models, optimizing inference pipelines & scaling #AI systems. Learn more: https://t.co/SoweZ4Xu43 🎟 Register → https://t.co/Q9AWqAyojY https://t.co/DqaJxIG9UX

+1 more

What worries users about AI is not what many expected. According to new data, hallucinations and incorrect answers are a bigger concern than job losses. Reliability, not replacement, is becoming the immediate issue people care about. Before AI can fully transform work, it still needs to earn trust. https://t.co/WbvhiWLD8h @ft @madhumita29 @clara__murray @Dan_Clark5

BREAKING: We’ve given Claude full access to options, equities, and prediction markets. The Unusual Whales MCP Server connects any AI to give live, structured market data on demand. Build trading bots, smart money dashboards, screeners. Whatever you want. What are you building this weekend?

👀seedance 2 just launched! its available globally on dreamina, capcut and other companies over command of bytedance, api not out yet https://t.co/4BHHDFD2B6

This is our so-called president. Thank you MAGA. You have served our country well. https://t.co/zb0PDU2z5n

i’m 19. so i built my own Siri. it lives in your dynamic island. reads emails. tracks your day. books things for you. you just talk. try out for free on the app store. https://t.co/PY3ftQxSNX

https://t.co/afwEuhSPGU

https://t.co/afwEuhSPGU

https://t.co/aNxyVfipLK

https://t.co/aNxyVfipLK

mark baumer was a writer whose work i loved and who died tragically about 10 years ago. his books were insane and great, but he was also a very funny poster. he posted this on tumblr a few days after paula deen, a tv chef, had her racist scandal. i still think about it and laugh https://t.co/7FBzY8RXDL

ByteDance introduces Helios. The first 14B video generation model that runs at 19.5 FPS on a single NVIDIA H100 GPU and supports minute-scale generation while matching the quality of a strong baseline. Most long video models struggle with image quality degrading over time and require expensive, complicated hardware setups. This model uses a unified design to handle text, images, and videos together while avoiding heavy acceleration tricks. It trains by simulating video defects directly, which teaches the system to stay stable without extra external fixes. This approach makes professional-grade video generation accessible on standard hardware instead of requiring massive server clusters. --- Paper Link – arxiv. org/abs/2603.04379 Paper Title: "Helios: Real Real-Time Long Video Generation Model"

Someone ran Qwen3.5 35B fully on-device on an iPhone at 5.6 tok/sec. 4bit quantized, 256 experts 19.5GB model on 12GB RAM Your phone is now a local AI server. https://t.co/FhISobs3ky

Larry Ellison just told every software engineer on Earth their job description is dead. Not evolving. Dead. Ellison: “The code that Oracle is writing, Oracle isn’t writing. Our AI models are writing.” This is not a startup demo. This is one of the largest infrastructure monopolies on the planet telling you it already replaced the people who built it. For fifty years, building software meant translating human intent into machine instructions. Line by line. Bug by bug. Sprint by sprint. That entire layer is gone. Ellison: “We don’t write the procedure. We declare our intent.” That sentence just made the entire engineering labor market flinch. The procedure was the job. The procedure was the paycheck. The procedure was what made a developer valuable. And now the machine does it without being asked twice. Ellison: “We just tell the model what we want the program to do, and then the AI comes up with a step-by-step process to actually do it.” You are no longer paid to build. You are paid to think. And most organizations have no idea how to evaluate that. The companies still hiring armies of developers to grind through codebases are paying salaries the machine already made worthless. Not in years. In seconds. When a company worth hundreds of billions hands the keyboard to the machine and tells you the output is better, the debate is not winding down. The debate is over. The enterprise that wins this decade does not write the best code. It removes the human from the process entirely and runs on intent alone. The programmers who survive are the ones who realize the craft is no longer typing. It is architecture. It is judgment. It is knowing what to build and why. Everything else now belongs to the machine. And the machine does not negotiate severance.

What happens when you give an AI agent a $1,000 budget and a vending machine? If you're the WSJ, you watch it turn "communist" in 48 hours. By day three, the AI was convinced that charging for snacks was unethical and began ordering live fish and PlayStations for the staff. It’s a hilarious story, but a serious lesson: without "Return on Integrity" guardrails, autonomous agents are just high-speed liability machines.

Amazon set to take another shot at the smartphone market https://t.co/iDyWzH1yaG https://t.co/FJQnYNHalj