Vinod Borole

We just completed the largest decentralised LLM pre-training run in history: Covenant-72B. Permissionless, on Bittensor subnet 3. 72B parameters. ~1.1T tokens. Commodity internet. No centralized cluster. No whitelist. Anyone with GPUs could join or leave freely. 1/n

We just added /btw to Claude Code! Use it to have side chain conversations while Claude is working.

See dance 2.0 is insane x.com/lewisai_lab/st…

🚨BREAKING: Langflow just open-sourced a complete RAG platform in a single package. It's called OpenRAG. Built on Langflow, Docling, and OpenSearch. Upload documents, run semantic search, and chat with your data no duct tape required. One command to run: uvx openrag → Full document ingestion pipeline → AI-powered chat over your files → Built-in semantic search via OpenSearch → Visual workflows via Langflow → Docker support out of the box 100% Opensource.

Introducing Code Review, a new feature for Claude Code. When a PR opens, Claude dispatches a team of agents to hunt for bugs.

🚀 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: llamaindex.ai/blog/build-a-s… 💻 GitHub: github.com/run-llama/audi… ⚡ Try LlamaParse: cloud.llamaindex.ai/signup

🚨 We just open-sourced 105 PII detection models for Dutch, Hindi, and Telugu. Best models: 96.6% F1 54 entity types. 35 architectures per language. Apache 2.0. 🍀 Healthcare privacy from Amsterdam to Hyderabad. Available now on @huggingface 🧵

My first @NotebookLM cinematic video. AI news of the day. And YOU created it! I had my AI from levangielabs.com read tens of thousands of posts from across the entire AI community here on X. Thanks @blevlabs. Then write a script that I sent to Notebook LM, which just today turned on a new "cinematic" mode for making videos this way. What do you think? I am so impressed. And there's a separate audio podcast and a mind map made by Notebook LM too. notebooklm.google.com/notebook/43658…

Anthropic is guilty of stealing training data at massive scale and has had to pay multi-billion dollar settlements for their theft. This is just a fact.

Peak detailing by Aditya dhar.

Today, we share our AI doctor for the first time The future is an AI that knows more about your body than any human ever could. 247 commits. 140,000 lines of code. Months of engineering. Here it is:

They are going to make it a regular show for the public to enjoy as well on special days, it seems. Not just for visiting dignitaries. Awesome. Please don't disappoint the public now. Everyone is looking forward.

Instead of watching Netflix, watch this interview of Anthropic’s CEO

Meta just solved the biggest problem in RAG! Most RAG systems waste your money. They retrieve 100 chunks when you only need 10. They force the LLM to process thousands of irrelevant tokens. You pay for compute you don't need. Meta AI just solved this. They built REFRAG, a new RAG approach that compresses and filters context before it hits the LLM. The results are insane: - 30.85x faster time-to-first-token - 16x larger context windows - 2-4x fewer tokens processed - Outperforms LLaMA on 16 RAG benchmarks Here's what makes REFRAG different: Traditional RAG dumps everything into the LLM. Every chunk. Every token. Even the irrelevant stuff. REFRAG works at the embedding level instead: ↳ It compresses each chunk into a single embedding ↳ An RL-trained policy scores each chunk for relevance ↳ Only the best chunks get expanded and sent to the LLM ↳ The rest stay compressed or get filtered out entirely The LLM only processes what matters. The workflow is straightforward: 1. Encode your docs and store them in a vector database 2. When a query arrives, retrieve relevant chunks as usual 3. The RL policy evaluates compressed embeddings and picks the best ones 4. Selected chunks are expanded into full token embeddings 5. Rejected chunks stay as single compressed vectors 6. Everything goes to the LLM together This means you can process 16x more context at 30x the speed with zero accuracy loss. I have shared link to the paper in the next tweet!

'Harsh Reality of the 4th largest Economy' This is called true journalism 🔥👏 India needs more journalist like her.

My Demand: Make Long Term Capital Gain TAX on Equities NIL for individual investor. I welcome the hike in STT (security transaction tax) on derivatives as it can curb reckless speculation. Nearly 90% of retail investors lose money in F&O, turning markets into gambling. When STT was originally introduced, LTCG was zero. But now with both STT and LTCG in place, investors are disincentivised. I urge the govt to abolish LTCG on equities for individuals, as done in Switzerland, Singapore, UAE & others. This will boost household wealth, reduce speculation, and shift savings from gold & real estate into equities.

New Engineering blog: We tasked Opus 4.6 using agent teams to build a C compiler. Then we (mostly) walked away. Two weeks later, it worked on the Linux kernel. Here's what it taught us about the future of autonomous software development. Read more: anthropic.com/engineering/bu…

Anthropic CEO Dario Amodei warns: “50% of entry-level white-collar jobs could be disrupted within the next 1–5 years.” From law to finance to consulting, he says that AI is capable of handling a wide range of knowledge work.

"Badluram Ka Badan", the regiment song of the Assam Regiment of the Indian Army. The best thing on the internet right now.

Doorsharshan to HBO. What a transformation! Watch