Mario Filho it is up to all of us to bring. it. back. say no to professional sponsored influencers with hyper-optimized content. say yes to boutique anons from the internet speaking their mind on their little corner of the internet.

NoProp: Training Neural Networks without Back-propagation or Forward-propagation ift.tt/UjN04y6

debates aside, you need Deepseek Wukong BYD in India to signal a new reality

NVIDIA adding native Python support to CUDA seems like a big deal. Python is the de facto coding language of AI.

📣 This just in... The *open source* and official GitHub MCP Server is here 🤩 github.blog/changelog/2025…

Today is the start of a new era of natively multimodal AI innovation. Today, we’re introducing the first Llama 4 models: Llama 4 Scout and Llama 4 Maverick — our most advanced models yet and the best in their class for multimodality. Llama 4 Scout • 17B-active-parameter model

AI at Meta We just dropped the ultimate guide to run llama-4 models (maverick and scout) on serverless gpus on your own AWS. Check out the guide here. tensorfuse.io/docs/guides/mo…

Existing Graph RAG (GraphRAG) methods struggle because they represent knowledge using only binary relations (linking two entities), missing complex real-world connections involving more than two entities. This paper introduces HyperGraphRAG, which uses hypergraphs to model these

"pip install" is dead. "uv add" is the new king.

‘Agent’ might be the most misused term in tech right now Here's what separates real agents from glorified chatbots: At their core, AI agents are LLMs with a specific role and task that have access to memory and external tools. They use reasoning capabilities to plan steps and

You can code an AI agent with MCP in 25 lines of code. Pydantic quietly dropped the simplest framework to build AI Agents.

How to use LLMs without giving your private data away? 𝗪𝗵𝗮𝘁'𝘀 𝘁𝗵𝗲 𝗽𝗿𝗼𝗯𝗹𝗲𝗺? Behind every LLM app, no matter how complex, agentic or ragish, there is at least one LLM that acts as either/and > the planer > tool decider > output synthetizer. And let's face it.

𝗖𝗵𝗲𝗰𝗸𝗼𝘂𝘁 𝘁𝗵𝗶𝘀 𝗜𝗡𝗖𝗥𝗘𝗗𝗜𝗕𝗟𝗘 𝘃𝗶𝘀𝘂𝗮𝗹 𝗿𝗲𝗽𝗿𝗲𝘀𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗼𝗳 𝘄𝗵𝗮𝘁 𝗵𝗮𝗽𝗽𝗲𝗻𝘀 𝗶𝗻𝘀𝗶𝗱𝗲 𝗟𝗟𝗠𝘀 𝗹𝗶𝗸𝗲 𝗖𝗵𝗮𝘁𝗚𝗣𝗧! ⬇️ Try it here: bbycroft.net/llm The visualization explained by Andrej Karpathy: youtube.com/watch?v=7xTGNN…

I still can't believe it, I'm incredibly humbled 🤩 When I created FastAPI, I never imagined it would get even close to Flask and Django. Let alone at the top across languages 🌍️ Thanks to those who inspired, helped, and adopted it 🙌 ...and I have big news tomorrow 😎

Opensource alternative to $200/month Manus AI agent. It runs locally on your computer using DeepSeek R1 or Qwen 3 to browse the web, write code, and execute tasks while keeping all your data private. 100% free and without internet.

Andrej Karpathy calls large language models the new computing paradigm: CPU -> LLM bytes -> tokens RAM -> context window this is the large language model OS (LMOS)

Coder 🤔

Can anyone convincingly argue that ai agents are anything more than just adding tool calling to an LLM

Wow. This is one of the best interactive sites I’ve seen for learning how LLMs work! 🔥 It starts w/ a clear intro and guides you through every core component: from Embedding, Layer Norm, and Self-Attention to MLPs, Transformer blocks, Softmax, and Output layers. link in 🧵↓

🚨HUGE. Windows OS is now using MCP!!!! Andrej Karpathy Microsoft beat you to LLMOS!!