Introduction to Artificial Intelligence

Written for the gap between tool manuals that expire and graduate texts that never open — a first course in how artificial intelligence actually works. Fifteen chapters run as one line of argument: concepts and history, search, logic, probability, machine learning, neural networks, deep learning, transformers, LLMs, reinforcement learning, and then ethics, governance, and what comes next.

Its method is arithmetic by hand. Five core computations — pathfinding, Bayesian spam filtering, backpropagation, attention, and Q-value updates — are worked through on paper rather than described. Fast-moving material is quarantined in “Further reading” boxes stamped with the date it was written, so the principles in the main text do not rot with the news. Every chapter closes with a summary, a glossary, and exercises; appendices cover the mathematics, Python labs, a glossary, and worked solutions. Lab code lives in a GitHub repository that is kept current.

Kim writes the bridge between AI in practice and AI in principle — a trilogy running from prompt engineering to enterprise local LLMs to the mechanics underneath.