DeepLearning.AI

We're thrilled to announce our new course: Retrieval Augmented Generation (RAG) RAG is a key part of building LLM applications that are grounded, accurate, and adaptable. In this course, taught by AI engineer Zain Hasan and available on Coursera, you’ll learn how to design and deploy production-ready RAG systems. You'll: - Combine retrievers and LLMs using tools like Weaviate, Together AI, and Phoenix - Apply keyword and semantic search methods - Evaluate performance to deploy and optimize production-ready systems You'll apply these techniques using real-world datasets in domains like healthcare, media, and e-commerce, and build the intuition to make informed architectural decisions. 📈 With the global RAG market projected to grow from $1.2B in 2024 to over $11B by 2030, RAG is core to real-world LLM systems. Start building with it today! Enroll now: https://hubs.la/Q03xtjCy0

The European Union published a “General Purpose AI Code of Practice,” explaining voluntary steps developers can take to meet the AI Act’s requirements for general‑use models. The code directs builders of models deemed to pose “systemic risks” to document data sources, log compute and energy use, and report security or safety incidents within two to ten days. Microsoft, Mistral, and all OpenAI opted in, while Meta declined. Read our full breakdown in The Batch to see what the guidelines mean for developers’ costs and timelines: https://hubs.la/Q03zNqSp0

We’re hiring a Product Marketing Manager at DeepLearning.AI! We’re looking for someone who’s excited to: 📣 Craft campaigns that resonate with AI builders and learners 🚀 Launch courses that help people build real-world AI applications 💡 Work cross-functionally to turn great ideas into learner impact 🛠 Build your own apps using AI to assist with core business functions This is a role for a creative doer, someone who loves both strategy and execution, and wants to grow with a fast-moving team shaping AI education. 📍 San Francisco Bay Area | Full time 🔗 Apply here: https://hubs.ly/Q03B1m2l0"

This week’s letter from Andrew Ng in The Batch asks a blunt question: Can surging performance from China’s open-weights models and home-grown chips let it overtake the U.S. in AI? He lays out the data behind China’s momentum, explains why Washington’s new action plan is helpful but not enough, and urges democracies to double down on open science and attracting talent. Read Andrew's letter in The Batch: https://hubs.la/Q03B0k_C0

President Trump released “Winning the Race: America’s AI Action Plan,” along with executive orders directing agencies to favor “ideologically neutral” models, fast-track data-center permits, and promote U.S. AI exports. The roadmap promises federal support for open-weights tools, relaxed environmental reviews for chips and energy projects, and threats to cut funding from states that pass restrictive AI laws. The policies could lower compute costs and unify rules for developers, though the neutrality mandate risks swapping one ideological filter for another. Learn more in The Batch: https://hubs.la/Q03zQkft0

LLMs can make sense of retrieved context because of how transformers work. In one of the lessons from the Retrieval Augmented Generation (RAG) course, we unpack how LLMs process augmented prompts using token embeddings, positional vectors, and multi-head attention. Understanding these internals helps you design more reliable and efficient RAG systems. Watch the breakdown and keep learning how to build production-ready RAG systems in this course, taught by Zain Hasan: https://hubs.la/Q03zPJ--0

This week, in The Batch, Andrew Ng discusses China’s accelerating AI momentum. Plus: 🏛️ White House resets U.S. AI policy 🚀 Alibaba updates Qwen3 family 🔓 U.S. lifts ban on GPUs for China 💭 Study links AI companion use to lower well-being Read The Batch: https://hubs.la/Q03zP3dV0

Beijing‑based Moonshot AI released the Kimi K2 LLM family, providing open‑weights access (under a modified MIT license) to a one trillion‑parameter model. The fine‑tuned Kimi‑K2‑Instruct scores 53 percent on LiveCodeBench and 76.5 percent on AceBench, surpassing other open‑weights non‑reasoning models. The MoE model processes inputs of up to 128k tokens, activates only 32 billion parameters, and supports external tool use through the model context protocol. Learn more in The Batch: https://hubs.la/Q03z8YGk0

In case you missed it, Buildathon, a rapid AI product-building competition is taking place on August 16 in Menlo Park! We teamed up with AI Fund to bring AI builders together for a full day of coding, prototyping, and community. Let's enjoy together: 🎤 Keynote by Andrew Ng 💻 AI-assisted coding challenges 🏆 $3,000+ in prizes 🎁 $100 Claude credits + Replit Core access Apply now: buildathon.ai

OpenAI’s 3 billion dollar bid for AI coding startup Windsurf collapsed. Instead, Google licensed the technology for 2.4 billion dollars while hiring CEO Varun Mohan, co‑founder Douglas Chen, and key engineers. Meanwhile, Cognition AI bought Windsurf’s remaining assets to bolster its Devin agentic coder. The agreement hands Google and Cognition a proven integrated development environment for AI‑assisted programming, Learn more in The Batch: https://hubs.la/Q03z88xh0