Thank you @pydantic https://t.co/aibMXrZiLX

Hugging Face just launched Storage Buckets blog: https://t.co/SAlKv1eehu https://t.co/cOiev5p4TT

Introducing Storage Buckets on Hugging Face 🧑‍🚀 The first new repo type on the Hub in 4 years: S3-like object storage, mutable, non-versioned, built on Xet deduplication. - Starting at $8/TB/mo. That's 3x cheaper than S3. You (and your coding agents) need somewhere to dump checkpoints, logs, and artifacts. Now they have a home.

GROK IMAGINE PROMPT SHARE : "a huge diamond-like lion, made of crystal, is roaring on the white beach, and next to it stands an elegant woman wearing a long dress, shimmering in the sunlight and sparkling like diamonds. the background features a clear blue sky and sea water, creating a dreamy scene."

BREAKING: Grok has surpassed 2.5 billion website visits. https://t.co/UjiK636p8E

Your creativity will be limitless with Grok Imagine and can’t be said the same with other AIs Start imagining with Grok https://t.co/Qrml6HdUCA

Zatanna lets AI agents use the software businesses already run on, including legacy systems, marketplaces, and portals. They turn them into APIs without needing custom integrations. Congrats on the launch, @rithvanga, @tarunvedula & @alzxbwell! https://t.co/sNkzEY2Nmh https://t.co/yd8MQC08oH

🎬 Kling 3.0 Motion Control is live on Lovart Character image + reference video = high-fidelity cinema with accurate physics. → Perfect facial consistency → Complex emotions and gestures → Up to 15s multi-shot 1080p output The mocap studio just became a browser tab.

Introducing CRABWISE 🦀 Build trust with your agents and set rules they can't break. Local-first agent monitoring, audit trail, and policy enforcement with multi-agent, multi-provider support → live agent monitoring → local policy enforcement (commandments) → a hash-chained audit trail → terminal UIs for status, watch, and audit → support for OpenClaw, Claude Code, Codex CLI (more to come) + pairs nicely with Crabwalk🦀 Follow me @luccasveg to stay updated + RT!

That gap between design and production? Closed. 🔁 GitHub Copilot, @code, and @figma now create a continuous loop. With the bidirectional Figma MCP server, Copilot users can: 🔹 Pull design context into code 🔹 Push working UI back to the canvas Connect your canvas and stay in the flow. 👇 https://t.co/bCZWRMohYJ

Today we're releasing our first open source TTS model, TADA! TADA (Text Audio Dual Alignment) is a speech-language model that generates text and audio in one synchronized stream to reduce token-level hallucinations and improve latency. This means: → Zero content hallucinations across 1,000+ test samples → 5x faster than similar-grade LLM-based TTS → Fits much longer audio: 2,048 tokens cover ~700 seconds with TADA vs. ~70 seconds in conventional systems → Free transcript alongside audio with no added latency

The US has much less secure elections than other major democracies We have the most powerful nuclear arsenal and most consequential elections in the world…yet in many states it’s harder to buy a beer than cast a vote This is a civilizational IQ test https://t.co/OHmRxJ9u9E

🚀 The team at @GoogleDeepMind just released Gemini Embedding 2, a frontier embeddings model with 3072 dimensions and state-of-the-art semantic quality. 👩‍💻 We built a demo showing how to integrate it across the LlamaIndex ecosystem, from LlamaParse to LlamaAgents: 𝗮𝘂𝗱𝗶𝗼-𝗸𝗯, a knowledge base for your audio notes. With audio-kb, you can: 🔹 Upload an MP3 or record directly from your terminal 🔹 LlamaParse extracts the transcript from the audio 🔹 Gemini Embedding 2 generates embeddings 🔹 Metadata + vectors are stored in @SurrealDB and indexed with HNSW 🔍 Once ingested, you can search all your audio notes directly from the terminal. 🎙️ Perfect for turning voice memos, meetings, or lectures into a searchable knowledge base. 📖 Full blog: https://t.co/HcuEvIAiQd 💻 GitHub: https://t.co/B7EMlJXhzW ⚡ Try LlamaParse: https://t.co/zyE5lXTPFV

Another cool app built with Perplexity Computer. A peer to peer file(s) transfer web app. Sends files directly with no accounts using WebRTC and DTLS encryption, file chunking, socket io signaling. I am impressed by how many libraries and tools Computer can orchestrate reliably. https://t.co/7jLCoXKfrE

Say hello to Gemini Embedding 2, our new SOTA multimodal model that lets your bring text, images, video, audio, and docs into the same embedding space! 👀 https://t.co/mjYk8FnTuj

🚨BREAKING: $META just acquired Moltbook. a social network built entirely for AI agents. not humans. agents. >moltbook launched in january and hit 1.6 million AI agents in weeks >the platform looks like reddit but only bots post, comment, and upvote. humans observe >cofounders matt schlicht and ben parr are joining meta’s superintelligence labs run by alexandr wang >the whole platform was vibe coded. schlicht didn’t write a single line of code himself >deal closes this month. both start at meta on march 16 meta is betting that AI agents talking to each other is the next layer of the internet. and they’re buying their way in.

Let’s meet up in Paris April 7-8 to talk @PyTorch, @vllm_project , @DeepSpeedAI, @raydistributed and the full stack of open source for training, inference, and agents Sponsor info: https://t.co/0x5MkgbgOO

210 days ago, @naturalpay was just a one-pager and a memo. Today, we’re coming out of stealth. Natural is the agentic payments platform powering frictionless money movement between agents, businesses, and consumers. Wallets. Payments. Ledgering. Routing. Identity. Compliance. Credit. Observability. Risk. Everything needed to move money. Engineered for agents and designed for humans. These primitives give you the ability to transact without becoming a payments expert or stitching together a dozen fragmented tools. Huge thanks to our team of 10 (soon to be 25), our early investors, and the supporters who believed in this vision from the start. If you want to help shape how money moves over the coming decades, we’re hiring. And if you’re building agents you should probably be moving money too. Reach out. Read more about Natural and the products we’re launching in the blog below. https://t.co/XlELDfBl2J

@haptagod @TheAhmadOsman Thanks! You might also like the sequel on RLVR with GRPO and distillation: https://t.co/uUtwrTWptp https://t.co/gpj7WstA1V

Mark Zuckerberg posted a photo with Alexandr Wang to on Threads to shut down the rumors, but because it's Threads, no one has any idea who it is or why he's posting it https://t.co/B9ojtJM7Ir

+1 more

Dear Chuck Schumer, Hi. Black dude here. I can trace my family ancestry to slavery. I even know where they were slaves. My mom experienced Jim Crow. I think I’ve watched every episode of “Eyes On the Prize” when I was younger. With that said… Can you directly explain to me how the SAVE Act is “Jim Crow 2.0?” Literally every black person I know has ID. Literally every black person I know has a car or at least a ride. Literally every black person I know knows how to vote (well… except the ones with felonies… but they don’t count). With your advanced white liberal thinking, you must know more than me. Apparently, as I experience daily on this app, white liberals are experts on being black; even more so than actual black folks. Perhaps you could explain it like I’m five. I’d look it up on the internet, but Kathy Hochul has already told me I don’t know what a computer is and Joe Biden said I can’t navigate it, anyway. Looking forward to your answer. No hugs. Zeek

The people support the SAVE America Act. Tell your Senator you want it passed: (202) 224-3121 https://t.co/wlFR1xMidZ

NotebookLM: Do a deep research report and make a video telling me exactly how to take over Rome if I time travelled to 66 BC with a single backpack. Actually pretty fun to watch and gets a lot of historical details in as well. https://t.co/SOWAnMHdeC

Go from @huggingface dataset → interactive embedding visualization, one command. Generates embeddings on a GPU via HF Jobs, deploys as a static Space using @Apple's Embedding Atlas. Search, filter, zoom - all in the browser with WebGPU. Script + docs: https://t.co/WeKW8XXN51 https://t.co/aNGpVIidUe

+1 more

ICYMI @huggingface Text Embeddings Inference v1.9+ now supports @nvidia Blackwell instances 💥 More in the thread 🧵 https://t.co/4HfmQn2Eca

AutoResearch-RL Perpetual Self-Evaluating Reinforcement Learning Agents for Autonomous Neural Architecture Discovery paper: https://t.co/M92rs8MfIg https://t.co/MEF5wjtLns

New Gemini updates to make @GoogleWorkspace more personal, helpful and collaborative: choose your sources and create a Doc draft in seconds, build complex Sheets 9X faster, or generate on-brand Slide layouts with a simple prompt. Plus, Drive now generates summarized answers right at the top of your search results so no more digging through folders. Rolling out today in beta, available first to Ultra + Pro subscribers in English globally for Docs, Sheets and Slides, and in the US for Drive.

I'll be giving a talk at the PyAI conf today. Below are the notes incase you can't make it: https://t.co/oZyVIAjjmn

Sandbar has closed a $23m Series A led by Adjacent and Kindred Ventures We aim to augment individuals in an agentic world Batch 2, developers, careers, & a mini preview below https://t.co/VUhlbQmqUq

Sparse-BitNet 1.58-bit LLMs are Naturally Friendly to Semi-Structured Sparsity paper: https://t.co/4QzqCY8MgA https://t.co/dcqvUo4KzY

NLE Non-autoregressive LLM-based ASR by Transcript Editing paper: https://t.co/O0oIVCp0IM https://t.co/AwK5Rl9Gag