Feb 27
Loading video player...
Full code in google colab: https://colab.research.google.com/drive/1Jy-DKEnwq8dn_Pk8yEeqkForFDhf76LP In this video, we build a Smart RAG (Retrieval-Augmented Generation) Agent from scratch using a fully educational Jupyter Notebook 🧠📓 You’ll learn how modern AI agents actually work, step by step, by creating a real system that can answer questions about AI alignment techniques — one of the most important topics in AI safety today. We use three classic AI alignment articles from the Alignment Forum and LessWrong as our knowledge base, and turn them into a working retrieval system that reasons over trusted sources instead of hallucinating. What you’ll learn: - What RAG is and why it matters for AI safety - How retrieval + generation work together - How to build an alignment-focused AI agent - How to structure an educational Jupyter Notebook for learning agents - How modern AI systems ground answers in real sources Business contact: mpyanev@gmail.com