27 notes in this chapter

Artificial Intelligence MOC

27 core AI concepts, from classical foundations to modern LLM agents. Each note has: definition, how it works, why it matters, common pitfalls, related terms, and a concrete example.

Foundations

• Turing Test
• Intelligent Agent
• Search Algorithms
• Knowledge Representation
• Expert Systems

Language & NLP

• Natural Language Processing (NLP)
• Tokenization
• Embeddings
• Large Language Model (LLM)
• Transformer Architecture
• Attention Mechanism
• Prompt Engineering
• Retrieval-Augmented Generation (RAG)
• Fine-Tuning
• Hallucination
• Cosine Similarity

Vision & Perception

• Computer Vision
• Object Detection

Agents & Systems

• Multi-Agent System
• Function Calling (Tool Use)
• Model Context Protocol (MCP)
• Agent SDKs and Frameworks
• RLHF (Reinforcement Learning from Human Feedback)

Ethics & Safety

• AI Alignment
• AI Bias and Fairness
• Explainable AI (XAI)
• Artificial General Intelligence (AGI)

---

How to use this

Skim a category before diving into an AI project or reading a paper. For the underlying math/architecture (neural nets, gradients, training), see Machine Learning and Deep Learning Terms MOC.

Suggested order if starting from zero

1. Intelligent Agent → Search Algorithms → Knowledge Representation — the classical AI mindset
2. Tokenization → Embeddings → Transformer Architecture → Attention Mechanism — how modern language models actually work
3. Large Language Model (LLM) → Hallucination → Retrieval-Augmented Generation (RAG) → Prompt Engineering — using LLMs in practice
4. Function Calling (Tool Use) → Model Context Protocol (MCP) → Agent SDKs and Frameworks → Multi-Agent System — building AI agents
5. AI Alignment → AI Bias and Fairness → Explainable AI (XAI) — the responsibility layer