Open-Source AI
Open models — Llama, Mistral, DeepSeek, Qwen and others — are catching up to closed labs and let you run AI on your own hardware, vendor-free. That shifts the balance of power, the safety debate and the economics of the field. Here is everything on open-source AI: open-weight releases, licenses, local deployment and the race against closed models.
Meta to Begin Manufacturing Own AI Chips in September 2026 to Reduce Nvidia Dependence
Meta expects to begin production of its own AI chip in September 2026—part of a strategy to increase computing power to 14 gigawatts by 2027, reports Reuters.
AWS Explains How It Safely Launches Frontier AI Models in Bedrock Service
Amazon Web Services has shared its approach to safely launching frontier AI models for customers through a managed cloud service Amazon Bedrock.
Meta Launches AI Mode Search in Facebook for Posts, Groups, and Reels
Meta adds a new search mode called AI Mode to Facebook that generates answers based on analyzing open posts, group publications, and Reels videos on the platform.
AI makes us faster, but wiser? Gas Town automates code
Mamba-3: transformer alternative with linear complexity
Running GLM-5.1 Locally: Complete Guide to llama.cpp and GGUF
Memory OS: Open 6-Layer Memory Stack for Hermes Agent
Nous Research presented Hermes Agent profile builder
New Hermes Agent interface replaces multi-step CLI configuration — users now assemble agent profiles in a single dashboard flow.
ECB Requires Banks to Plan Protection Against AI Cyberattacks by October
European Central Bank warns largest eurozone banks that frontier AI can be used to hack financial systems. By October 2026 banks must develop protection plans.
Mistral Launches Muse Spark 1.1 — A Model for Finding Code Errors
Positron Startup Raises $750M to Compete with NVIDIA in AI Chip Development
Ant Group Presented LingBot-VLA 2.0 — Open Model for Controlling Robots of Any Design
Tencent Open-Sourced LLM Hy3 with 295 Billion Parameters
The New York Times demands sanctions against OpenAI for evidence concealment
Publishers The New York Times and Daily News are asking a federal judge to impose sanctions on OpenAI, accusing the company of hiding key evidence in a copyright…
AI Agent Orchestration Proved More Important Than Model Selection: arXiv Research
Research showed that optimizing the AI agent orchestration layer reduces costs more than switching to another model: 41% savings while maintaining quality.