# News: March 17, 2026

> 8 curated links on AI agents, Python, security, Google Cloud, and engineering, selected by Médéric Hurier.

- Author: [Médéric Hurier (Fmind)](https://www.fmind.dev/)
- Published: 2026-03-17
- URL: https://www.fmind.dev/news/2026-03-17/

- [NVIDIA GTC Keynote 2026](<https://www.youtube.com/watch?v=jw_o0xr8MWU>): Jensen Huang presents the future of accelerated computing, including the Rubin GPU architecture, new physical AI breakthroughs, and agentic systems for industries.
- [NVIDIA Rubin at GTC 2026: Full Technical Breakdown for ML Engineers](<https://blog.barrack.ai/nvidia-rubin-specs-architecture-2026/>): NVIDIA's next-generation Rubin architecture features 336 billion transistors, HBM4 memory with 22 TB/s bandwidth, and a 10x inference cost reduction for MoE models.
- [NVIDIA Expands Open Model Families to Power the Next Wave of Agentic, Physical and Healthcare AI](<https://nvidianews.nvidia.com/news/nvidia-expands-open-model-families-to-power-the-next-wave-of-agentic-physical-and-healthcare-ai>): NVIDIA introduced new open model families including Nemotron 3 for agentic systems, Cosmos 3 for physical AI, and specialized models for robotics and drug discovery.
- [Introducing Mistral Small 4](<https://mistral.ai/news/mistral-small-4>): Mistral Small 4 is a new Apache 2.0 licensed multimodal model that unifies reasoning, coding, and general-purpose capabilities with 119B total parameters and a 256k context window.
- [Engineering the next generation of LinkedIn’s Feed](<https://www.linkedin.com/blog/engineering/feed/engineering-the-next-generation-of-linkedins-feed>): LinkedIn detailed its transition to a unified retrieval and ranking system powered by fine-tuned LLMs and sequential transformers for personalized content delivery at scale.
- [BigQuery Studio is more useful than ever, with enhanced Gemini assistant](<https://cloud.google.com/blog/products/data-analytics/gemini-supercharges-the-bigquery-studio-assistant/>): Google Cloud updated BigQuery Studio's Gemini assistant with context-aware interoperability, intelligent resource discovery via Dataplex, and instant job troubleshooting capabilities.
- [Meta planning sweeping layoffs as AI costs mount](<https://www.reuters.com/business/world-at-work/meta-planning-sweeping-layoffs-ai-costs-mount-2026-03-14/>): Meta is reportedly planning layoffs affecting up to 20% of its workforce as it refocuses resources on heavy generative AI infrastructure investments.
- [Introducing AI Works for Europe](<https://blog.google/company-news/inside-google/around-the-globe/google-europe/ai-works-for-europe/>): Google launched a new initiative including a $30 million fund and professional certificates to help European workers and students develop essential AI skills.