掌握現代 AI 架構的核心組件：LLM、System Prompts、RAG、Agents 與 MCP

Master the core building blocks of modern AI architecture (LLMs, System Prompts, RAG, Agents, MCP)

AI 導讀

這段影片以 12 個核心組件拆解現代 AI 架構，並透過日常類比和 Python SDK 程式碼示範説明。內容涵蓋 System Prompts 與 User Prompts、向量數據庫與 Dense Embeddings，以及 AI 智能體的 ReAct 循環。影片亦講解 MCP 的 tools/list 和 tools/call JSON-RPC 2.0 規格。

正文

How do modern AI systems actually work under the hood? In this video, we break down the 12 core building blocks of modern AI Architecture using simple everyday real-life analogies (LEGO bricks, USB-C cables, open-book exams) and real Python SDK code snippets!

Whether you're a software engineer, developer, product manager, or AI enthusiast, this complete visual guide covers everything from Tokens and System Prompts to RAG, Vector Databases, AI Agents, and the Model Context Protocol (MCP).

📌 WHAT YOU WILL LEARN:
• The difference between System Prompts (developer directives) and User Prompts (runtime queries) in Python.
• How Vector Databases and Dense Embeddings turn text into mathematical coordinates.
• How AI Agents use the ReAct loop (Think -> Act -> Observe) for multi-step tasks.
• Standard Model Context Protocol (MCP) tools/list & tools/call JSON-RPC 2.0 specifications.